Introduction

            Quantum chemistry is a branch of theoretical chemistry, which applies quantum mechanics and quantum field theory to address issues and problems in chemistry. The description of the electronic behavior of atoms and molecules as pertaining to their reactivity is one of the applications of quantum chemistry. Quantum chemistry lies on the border between chemistry and physics, and significant contributions have been made by scientists from both fields. It has a strong and active overlap with the field of atomic physics and molecular physics, as well as physical chemivistry. Quantum chemistry mathematically describes the fundamental behavior of matter at the molecular scale.

            Computational chemistry is a new discipline. Its advent and popularity have paralleled improvements in computing power during the last several decades. As with other disciplines in chemistry, computational chemistry uses tools to understand chemical reactions and processes. Scientists use computer software to gain insight into chemical processes. Although computational chemists frequently develop and refine software tools, their primary interest is in applying software tools to enhance chemical knowledge.

            It is, in principle, possible to describe all chemical systems using this theory. In practice, only the simplest chemical systems may realistically be investigated in purely quantum mechanical terms, and approximations must be made for most practical purposes (e.g., Hartree-Fock, post Hartry-Fock or Density functional theory, see computional chemistry for more details). Hence a detailed understanding of quantum mechanics is not necessary for most chemistry, as the important implications of the theory (principally the orbital approximation) can be understood and applied in simpler terms.

            The book is written for students of the universities which are studying quantum chemistry in English; in doing so we used our lectures which have been read for many years of working at various universities.   as well as our experience in scientific investigations. Experience of scientific working in the area of quantum chemistry also was useful to us.

            The given book cannot capture all aspects of computer chemistry. However, we would like to create the textbook which would be useful to students who attempt to begin the quantum calculations of simple molecules. Therefore, the book involves the basis of the quantum theory, methods of quantum calculations and to end it compete a brief manual of using of computing program HyperChem. is resulted with a brief manual of using of computing program HyperChem.

            Authors with gratitude will accept all useful remarks.

Authors

            Chapter 1. Bases of the quantum chemistry

            1.1. History of Quantum Chemistry

Quantum mechanics (QM, or quantum theory) is a physical science dealing with the behaviour of matter and energy on the scale of atoms and subatomic particles / waves. QM also forms the basis for the contemporary understanding of how very large objects such as stars and galaxies, and cosmological events such as the Big Bang, can be analyzed and explained. Quantum mechanics is the foundation of several related disciplines including nanotechnology, condensed matter physics, quantum chemistry, structural biology, particle physics, and electronics.

            The term "quantum mechanics" was first coined by Max Born in 1924. The acceptance by the general physics community of quantum mechanics is due to its accurate prediction of the physical behaviour of systems, including systems where Newtonian mechanics fails. Even general relativity is limited—in ways quantum mechanics is not—for describing systems at the atomic scale or smaller, at very low or very high energies, or at the lowest temperatures. Through a century of experimentation and applied science, quantum mechanical theory has proven to be very successful and‎‎‏ِ practical. The foundations of quantum mechanics date from the early 1800s, but the real beginnings of QM date from the work of Max Planck in 1900. Albert Einstein and Niels Bohr soon made important contributions to what is now called the "old quantum theory." However, it was not until 1924 that a more complete picture emerged with Louis de Broglie's matter-wave hypothesis and the true importance of quantum mechanics became clear. Some of the most prominent scientists to subsequently contribute in the mid-1920s to what is now called the "new quantum mechanics" or "new physics" were Max Born, Paul Dirac, Werner Heisenberg, Wolfgang Pauli, and Erwin Schrödinger. Later, the field was further expanded with work by Julian Schwinger, Sin-Itiro Tomonaga and Richard Feynman for the development of Quantum Electrodynamics in 1947 and by Murray Gell-Mann in particular for the development of Quantum Chromodynamics.

In establishment quantum chemistry the important role has played spectroscopic investigations of the termic radiation of solids.  Early researchers differed in their explanations of the fundamental nature of what we now call electromagnetic radiation. Some maintained that light and other frequencies of electromagnetic radiation are composed of particles, while others asserted that electromagnetic radiation is a wave phenomenon. In classical physics these ideas are mutually contradictory. Ever since the early days of QM scientists have acknowledged that neither idea by itself can explain electromagnetic radiation.

In 1690, Christian Huygens explained the laws of reflection and refraction on the basis of a wave theory. Sir Isaac Newton believed that light consisted of infinitesimally small particles which he designated "corpuscles". In 1827 Thomas Young and Augustin Fresnel made experiments on interference that showed that a corpuscular theory of light was inadequate. Then in 1873 James Clerk Maxwell showed that by making an electrical circuit oscillate it should be possible to produce electromagnetic waves. His theory made it possible to compute the speed of electromagnetic radiation purely on the basis of electrical and magnetic measurements, and the computed value corresponded very closely to the empirically measured speed of light. In 1888, Heinrich Hertz made an electrical device that actually produced what we would now call microwaves — essentially radiation at a lower frequency than visible light. Everything up to that point suggested that Newton had been entirely wrong to regard light as corpuscular.

            The history of quantum mechanics began essentially with the 1838 discovery of catode rays by Michael Faraday,  the 1859 statement of the black body radiation problem by Gustav Kirchhof, the 1877 suggestion by Ludwig Boltzmann that the energy states of a physical system could be discrete, and the 1900 quantum hypothesis by Max Planck that any energy radiating atomic system can theoretically be divided into a number of discrete ‘energy elements’ ε such that each of these energy elements is proportional to the frequency ν with which they each individually radiate energy, as defined by the following formula:

E=hν                                                                      (1)

where h is a numerical value called Planck’s Constant. According Plank’s theory the density of the  termical radiation energy of the solids submits to following Plank’s  law

                                                         (2)

It is fairly easy to see a spectrum produced by white light when it passes through a prism, the bevelled edge of a mirror or a tapered pane of glass, or through drops of rain to form a rainbow. When samples of single elements are caused to emit light they may emit light at several characteristic frequencies. The frequency profile produced is characteristic of that element. Instead of there being a wide band filled with colors from violet to red, there will be isolated bands of single colors separated by darkness. Such a display is called a line spectrum. Some lines go beyond the visible frequencies and can only be detected by special photographic film or other such devices. Scientists hypothesized that an atom could radiate light the way the string on a fine violin radiates sound – not only with a fundamental frequency (in which the entire string moves the same way at once) but with several higher harmonics (formed when the string divides itself into halves and other divisions that vibrate in coordination with each other as when one half of the string is going one way as the other half of the string is going the opposite way). For a long time nobody could find a mathematical way to relate the frequencies of the line spectrum of any element.

NASA photo of the bright-line spectrum of hydrogen

Wavelength

Fig.1. The photo of the bright-line spectrum of hydrogen

 

Photo of the bright-line spectrum of nitrogen

Fig.2. Photo of the bright-line spectrum of nitrogen

 

In 1885, Johann Jakob Balmer (1825-1898) figured out how the frequencies of atomic hydrogen are related to each other. The formula is a simple one:

                                                             (3)

where λ is wavelength, R is the Rydberg constant and n is an integer (n =3, 4,...) This formula can be generalized to apply to atoms that are more complicated than hydrogen, but we will stay with hydrogen for this general exposition. (That is the reason that the denominator in the first fraction is expressed as a square.)

            The bond order concept is used e.g. in molecular dynamics bond order potentials. The next development was the discovery of the Zeeman effect, named after Pieter Zeeman (1865-1943). The physical explanation of the Zeeman effect was worked out by Hendrik Antoon Lorentz (1853-1928). Lorentz hypothesized that the light emitted by hydrogen was produced by vibrating electrons. It was possible to get feedback on what goes on within the atom because moving electrons create a magnetic field and so can be influenced by the imposition of an external magnetic field in a manner analogous to the way that one iron magnet will attract or repel another magnet.

            The Zeeman effect could be interpreted to mean that light waves are originated by electrons vibrating in their orbits, but classical physics could not explain why electrons should not fall out of their orbits and into the nucleus of their atoms, nor could classical physics explain why their orbits would be such as to produce the series of frequencies derived by Balmer’s formula and displayed in the line spectra. Why did the electrons not produce a continuous spectrum?

            Then, in 1905, to explain the photoelectric effect (1839), i.e. that shining light on certain materials can function to eject electrons from the material, Albert Einstein postulated, as based on Planck’s quantum hypothesis, that light itself consists of individual quantum particles, which later came to be called photons (1926). In 1905, Albert Einstein used Planck's constant to postulate that the energy in a beam of light occurs in concentrations that he called photons. According to that account, a single photon of a given frequency delivers an invariant amount of energy. In other words, individual photons can deliver more or less energy, but only depending on their frequencies. Although the description that stemmed from Planck's research sounds like Newton's corpuscular account, Einstein's photon was still said to have a frequency, and the energy of the photon was accounted proportional to that frequency. The particle account had been compromised once again.

            Both the idea of a wave and the idea of a particle are models derived from our everyday experience. We cannot see individual photons. We can only investigate their properties indirectly. We look at some phenomena, such as the rainbow of colors that we see when a thin film of oil rests on the surface of a puddle of water, and we can explain that phenomenon to ourselves by comparing light to waves. We look at other phenomena, such as the way a photoelectric meter in our camera works, and we explain it by analogy to particles colliding with the detection screen in the meter. In both cases we take concepts from our everyday experience and apply them to a world we have never seen. Neither form of explanation is entirely satisfactory. In general any model can only approximate that which it models. A model is useful only within the range of conditions where it is able to predict the real thing with accuracy. Newtonian physics is still a good predictor of many of the phenomena in our everyday life. To remind us that both "wave" and "particle" are concepts imported from our macro world to explain the world of atomic-scale phenomena, some physicists such as George Gamow have used the term "wavicle" to refer to whatever it is that is really there. In the following discussion, "wave" and "particle" may both be used depending on which aspect of quantum mechanical phenomena is under discussion.

            1.2. Bohr atom

            In 1897 the particle called the electron was discovered. By means of the gold foil experiment physicists discovered that matter is, volume for volume, largely space. Once that was clear, it was hypothesized that negative charge entities called electrons surround positively charged nuclei. So at first all scientists believed that the atom must be like a miniature solar system. But that simple analogy predicted that electrons would, within about one hundredth of a microsecond, crash into the nucleus of the atom. The great question of the early 20th century was, "Why do electrons normally maintain a stable orbit around the nucleus?" In 1913, Niels Bohr removed this substantial problem by applying the idea of discrete (non-continuous) quanta to the orbits of electrons. This account became known as the Bohr model of the atom. Bohr basically theorized that electrons can only inhabit certain orbits around the atom. These orbits could be derived by looking at the spectral lines produced by atoms. Bohr explained the orbits that electrons can take by relating the angular momentum of electrons in each "permitted" orbit to the value of h, Planck's constant. He held that an electron in the lowest orbital has an angular momentum equal to h/2π. Each orbit after the initial orbit must provide for an electron's angular momentum being an integer multiple of that lowest value. He depicted electrons in atoms as being analogous to planets in a solar orbit. However, he took Planck's constant to be a fundamental quantity that introduces special requirements at this subatomic level and that explains the spacing of those "planetary" orbits.

Fig.3. The Bohr model of the atom, showing electron quantum jumping to ground state n=1

 

            Bohr considered one revolution in orbit to be equivalent to one cycle in an oscillator (as in Planck's initial measurements to define the constant h) which is in turn similar to one cycle in a wave. The number of revolutions per second is (or defines) what we call the frequency of that electron or that orbital. Specifying that the frequency of each orbit must be an integer multiple of Planck's constant h would only permit certain orbits, and would also fix their size.

Bohr generalized Balmer's formula for hydrogen by replacing denominator in the term 1/4 with an explicit squared variable:

                                                       (4)

m=1,2,3,4,5,..., and n > k, where λ is the wavelength of the light and RH is the Rydberg constant for hydrogen. This generalization predicted many more line spectra than had been previously detected, and experimental confirmation of this prediction followed.

It follows almost immediately that if λ is quantized as the formula above indicates, then the momentum of any photon must be quantized. The frequency of light, ν, at a given wavelength λ is given by the relationship

ν=c/λ,   and: λ=c/ν  and multiplying by h/h = 1,

λ=hc/hν, and we know that E = hν so λ=hc/E which we can rewrite as:

λ=h/(Ec), and E/c = p (momentum) so λ=h/p or p=h/λ                                (5)

Beginning with line spectra, physicists were able to deduce empirically the rules according to which the orbits of electrons are determined and to discover something vital about the momentums involved--that they are quantized.

Bohr next realized how the angular momentum of an electron in its orbit, L, is quantized, i.e., he determined that there is some constant value K such that when it is multiplied by Planck’s constant, h, it will yield the angular momentum that pertains to the lowest orbital. When it is multiplied by successive integers it will then give the values of other possible orbitals. He later determined that K = 1/2π . Bohr's theory represented electrons as orbiting the nucleus of an atom in a way that was amazingly different from what we see in the world of our everyday experience. He showed that when an electron changed orbits it did not move in a continuous trajectory from one orbit around the nucleus to another. Instead, it suddenly disappeared from its original orbit and reappeared in another orbit. Each distance at which an electron can orbit is a function of a quantized amount of energy. The closer to the nucleus an electron orbits, the less energy it takes to remain in that orbital. Electrons that absorb a photon gain a quantum of energy, so they jump to an orbit that is farther from the nucleus, while electrons that emit a photon lose a quantum of energy and so jump to an inner orbital. Electrons cannot gain or lose a fractional quantum of energy, and so, it is argued, they cannot have a position that is at a fractional distance between allowed orbitals. Allowed orbitals were designated as whole numbers using the letter n with the innermost orbital being designated n = 1, the next out being n = 2, and so on. Any orbital with the same value of n is called an electron shell. Bohr's model of the atom was essentially two-dimensional because it depicts electrons as particles in circular orbits. In this context, two-dimensional means something that can be described on the surface of a plane. One-dimensional means something that can be described by a line. Because circles can be described by their radius, which is a line segment, sometimes Bohr's model of the atom is described as one-dimensional.

            1.3. Wave-particle duality. De Broglie's hypothesis. De Broglie’s waves

            Niels Bohr determined that it is impossible to describe light adequately by the sole use of either the wave analogy or of the particle analogy. Therefore he enunciated the principle of complementarity, which is a theory of pairs, such as the pairing of wave and particle or the pairing of position and momentum. Louis de Broglie worked out the mathematical consequences of these findings. In quantum mechanics, it was found that electromagnetic waves could react in certain experiments as though they were particles and in other experiments as though they were waves. It was also discovered that subatomic particles could sometimes be described as particles and sometimes as waves. This discovery led to the theory of wave-particle duality by Louis-Victor de Broglie in 1924, which states that subatomic entities have properties of both waves and particles at the same time.

            The Bohr atom model was enlarged upon with the discovery by de Broglie that the electron has wave-like properties. In accord with de Broglie's conclusions, electrons can only appear under conditions that permit a standing wave. A standing wave can be made if a string is fixed on both ends and made to vibrate (as it would in a stringed instrument). That illustration shows that the only standing waves that can occur are those with zero amplitude at the two fixed ends. The waves created by a stringed instrument appear to oscillate in place, simply changing crest for trough in an up-and-down motion. A standing wave can only be formed when the wave's length fits the available vibrating entity. In other words, no partial fragments of wave crests or troughs are allowed. In a round vibrating medium, the wave must be a continuous formation of crests and troughs all around the circle. Each electron must be its own standing wave in its own discrete orbital.

Probability distribution of the Bohr atom

Fig.4. Bohr atom after wave-particle duality and the Uncertainty Principle

            In 1924, Louis-Victor de Broglie formulated the de Broglie hypothesis, claiming that all matter, not just light, has a wave-like nature; he related wavelength (denoted as λ), and momentum (denoted as p):

λ=h/p                                                                     (6)

            This is a generalization of Einstein's equation above, since the momentum of a photon is given by p=E/c and the wavelength by λ=c/f, where c is the speed of light in vacuum.

            De Broglie's formula was confirmed three years later for electrons (which differ from photons in having a rest mass) with the observation of electron diffraction in two independent experiments. At the University of Aberdeen, George Paget Thomson passed a beam of electrons through a thin metal film and observed the predicted interference patterns. At Bell Labs Clinton Joseph Davisson and Lester Halbert Germer guided their beam through a crystalline grid.

            De Broglie was awarded the Nobel Prize for Physics in 1929 for his hypothesis. Thomson and Davisson shared the Nobel Prize for Physics in 1937 for their experimental work.

            1.4. Heisenberg's uncertainty principle

            In his work on formulating quantum mechanics, Werner Heisenberg postulated his uncertainty principle, which states:

ΔxΔp≥ž/2                                                                   (7)

where Δ here indicates standard deviation, a measure of spread or uncertainty;  x and p are a particle's position and linear momentum respectively. ž is the reduced Planck's constant (Planck's constant divided by 2π). Heisenberg originally explained this as a consequence of the process of measuring: Measuring position accurately would disturb momentum and vice-versa, offering an example (the "gamma-ray microscope") that depended crucially on the de Broglie hypothesis. It is now understood, however, that this only partly explains the phenomenon: the uncertainty also exists in the particle itself, even before the measurement is made. In fact, the modern explanation of the uncertainty principle, extending the Copenhagen Interpretation first put forward by Bohr and Heisenberg, depends even more centrally on the wave nature of a particle: Just as it is nonsensical to discuss the precise location of a wave on a string, particles do not have perfectly precise positions; likewise, just as it is nonsensical to discuss the wavelength of a "pulse" wave traveling down a string, particles do not have perfectly precise momenta (which corresponds to the inverse of wavelength). Moreover, when position is relatively well-defined, the wave is pulse-like and has a very ill-defined wavelength (and thus momentum). And conversely, when momentum (and thus wavelength) is relatively well-defined, the wave looks long and sinusoidal, and therefore it has a very ill-defined position.

            De Broglie himself had proposed a pilot wave construct to explain the observed wave–particle duality. In this view, each particle has a well-defined position and momentum, but is guided by a wave function derived from Schrödinger's equation. The pilot wave theory was initially rejected because it generated non-local effects when applied to systems involving more than one particle. Non-locality, however, soon became established as an integral feature of Quantum Theory (see EPR paradox), and David Bohm extended de Broglie's model to explicitly include it. In Bohmian mechanics, the wave–particle duality is not a property of matter itself, but an appearance generated by the particle's motion subject to a guiding equation or quantum potential.

            Since the demonstrations of wave-like properties in photons and electrons, similar experiments have been conducted with neutrons and protons. Among the most famous experiments are those of Estermann and Otto Stern in 1929. Authors of similar recent experiments with atoms and molecules, described below, claim that these larger particles also act like waves.

            A dramatic series of experiments emphasizing the action of gravity in relation to wave–particle duality were conducted in the 1970s using the neutron interferometer. Neutrons, one of the components of the atomic nucleus, provide much of the mass of a nucleus and thus of ordinary matter. In the neutron interferometer, they act as quantum-mechanical waves directly subject to the force of gravity. While the results were not surprising since gravity was known to act on everything, including light (see tests of general relativity and the Pound-Rebka falling photon experiment), the self-interference of the quantum mechanical wave of a massive fermion in a gravitational field had never been experimentally confirmed before.

            In 1999, the diffraction of C60 fullerenes by researchers from the University of Vienna was reported. Fullerenes are comparatively large and massive objects, having an atomic mass of about 720u. The de Broglie wavelength is 2.5 pm, whereas the diameter of the molecule is about 1 nm, i.e. about 400 times larger. As of 2005, this is the largest object for which quantum-mechanical wave-like properties have been directly observed in far-field diffraction.

            In 2003 the Vienna group also demonstrated the wave nature of tetraphenylporphyrin – a flat biodye with an extension of about 2 nm and a mass of 614u. For this demonstration they employed a near-field Talbot Lau interferometer. In the same interferometer they also found interference fringes for C60F48., a fluorinated buckyball with a mass of about 1600u, composed of 108 atoms[6] Large molecules are already so complex that they give experimental access to some aspects of the quantum-classical interface, i.e. to certain decoherence mechanisms.

            Whether objects heavier than the Planck mass (about the weight of a large bacterium) have a de Broglie wavelength is theoretically unclear and experimentally unreachable; above the Planck mass a particle's Compton wavelength would be smaller than the Planck length and its own Schwarzschild radius, a scale at which current theories of physics may break down or need to be replaced by more general ones.

            1.5. Wave model. Wave function

            The foundation of quantum mechanics and quantum chemistry is the wave model, in which the atom is a small, dense, positively charged nucleus surrounded by electrons. Unlike the earlier Bohr model of the atom, however, the wave model describes electrons as "clouds" moving in orbitals, and their positions are represented by probability distributions rather than discrete points. The strength of this model lies in its predictive power. Specifically, it predicts the pattern of chemically similar elements found in the periodic table. The wave model is so named because electrons exhibit properties (such as interference) traditionally associated with waves.

            Electron cloud is a term used, if not originally coined, by the Nobel Prize laureate and acclaimed educator Richard Feynman in The Feynman Lectures on Physics for discussing "exactly what is an electron?". This intuitive model provides a simplified way of visualizing an electron as a solution of the Schrödinger equation. In the electron cloud analogy, the probability density of an electron, or wavefunction, is described as a small cloud moving around the atomic or molecular nucleus, with the thickness of the cloud proportional to the probability density.

            The model evolved from the earlier Bohr model, which likened an electron orbiting an atomic nucleus to a planet orbiting the sun. The electron cloud formulation better explains many observed phenomena, including the double slit experiment, the periodic table and molecular bonding, and atomic interactions with light. Although lacking in certain details, the intuitive model roughly allows for wave-particle duality, for electron behaviour that is both "wavy" as per the left side of E=mc2 and "lumpy" as per the right side.

            Experimental evidence suggests that the probability density is not just a theoretical model for the uncertainty in the location of the electron, but rather that it reflects the actual state of the electron. This carries an enormous philosophical implication, that the universe's evolution is fundamentally uncertain. What drives this randomness in the motion of an electron is a question that has seen little advancement since its discovery in the early 20th century, although the answer, written in the fine print of the universe, has the prospect of being among the greatest breakthroughs of all time. Certainly the physicists of the time appreciated its significance, readily seen in the Copenhagen Interpretation.

            In quantum mechanics, a probability amplitude is a complex-valued function that describes an uncertain or unknown quantity. For example, each particle has a probability amplitude describing its position. This amplitude is the wave function, expressed as a function of position. The wave function is a complex-valued function of a continuous variable.

            For a state ψ, the associated probability density function is ψ*ψ, which is equal to |ψ|2. This is sometimes called just probability density, and may be found and used without normalization.

Probability amplitude:                                     ψ(x)

Probability density:                                |ψ(x)|2 = ψ(x)· ψ(x)*

            If |ψ|2 has a finite integral over the whole of three-dimensional space, then it is possible to choose a normalising constant, c, so that by replacing ψ by cψ the integral becomes 1. Then the probability that a particle is within a particular region V is the integral over V of |ψ|2. Which means, according to the Copenhagen interpretation of quantum mechanics, that, if some observer tries to measure the quantity associated with this probability amplitude, the result of the measurement will lie within ε with a probability P(ε) given by

P(ε)= ∫ε|ψ(x)|2dx                                                              (8)

            Probability amplitudes which are not square integrable are usually interpreted as the limit of a series of functions which are square integrable. For instance the probability amplitude corresponding to a plane wave corresponds to the 'non physical' limit of a monochromatic source of particles. Another example: The Siegert wave functions describing a resonance are the limit for of a time-dependent wave packet scattered at an energy close to a resonance. In these cases, the definition of P(ε) given above is still valid.

            The change over time of this probability (in our example, this corresponds to a description of how the particle moves) is expressed in terms of ψ itself, not just the probability function |ψ|2. See Schrödinger equation.

            In order to describe the change over time of the probability density it is acceptable to define the probability flux (also called probability current). The probability flux j is defined as

                                     (9)

and measured in units of (probability)/(area × time) = r−2t−1.

            The probability flux satisfies a quantum continuity equation, i.e.:

                                                        (10)

where P(x, t) is the probability density and measured in units of (probability)/(volume) = r−3. This equation is the mathematical equivalent of probability conservation law.

It is easy to show that for a plane wave function,

                                                (11)

the probability flux is given by              

                                                          (12)

The bi-linear form of the axiom has interesting consequences as well.

            The first step in solving a quantum chemical problem is usually solving the Schrödinger equation (or Dirac equation in relativistic quantum chemistry) with the electronic molecular Hamiltonian. This is called determining the electronic structure of the molecule. It can be said that the electronic structure of a molecule or crystal implies essentially its chemical properties.

            Schrödinger's wave equation with its unique wavefunction for a single electron is also spread out in a probability distribution like Heisenberg's quantized particle-like electron. This is because a wave is naturally a widespread disturbance and not a point particle. Therefore, Schrödinger's wave equation has the same predictions made by the uncertainty principle because uncertainty of location is built into the definition of a widespread disturbance like a wave. Uncertainty only needed to be defined from Heisenberg's matrix mechanics because the treatment was from the particle-like aspects of the electron. Schrödinger's wave equation shows that the electron is in the probability cloud at all times in its probability distribution as a wave that is spread out. Max Born discovered in 1928 that when you compute the square of Schrödinger's wavefunction (psi-squared), you get the electron's location as a probability distribution. Therefore, to the extent that a measurement of the position of an electron can be made as an exact location in space instead of as a probability distribution, the electron appears to momentarily cease to have wave-like properties. Without wave-like properties, none of Schrödinger's definitions of the electron being wave-like make sense. The measurement of the position of the particle nullifies the simple wave-like properties and the one body form of Schrödinger's equation then fails. Because the electron can no longer be described by its separate wavefunction when measured, due to its wave length becoming much shorter and its becoming entangled with the particles of the measuring apparatus, this is called wavefunction collapse.

            1.6. Electron cloud model

            In the electron cloud model, rather than following fixed orbits, electrons bound to an atom are observed more frequently in certain areas around the nucleus called orbitals. The electron cloud can transition between electron orbital states, and each state has a characteristic shape and energy, all predicted as the Schrödinger equation has infinitely many solutions. Experimental results motivated this conceptual refinement of the Bohr model. The famous double slit experiment demonstrates the random behaviour of electrons, as free electrons shot through a double slit are seen to materialize at random locations with wavelike interference. Heisenberg's uncertainty principle accounts for this and, taken together with the double slit experiment, implies that an electron behaves more like a smear of infinitesimal pieces, or "cloud", each piece moving somewhat independently as in a churning cloud. These pieces can be forced to coincide at an isolated point in time, but then they all must move relative to each other at an increased spread of rates to conserve the "uncertainty". Certain physical interactions of this wave-like electron, such as observing which slit an electron passes through in the double-slit experiment, require this coincidence of pieces into a lump-like particle. In such an interaction the electron "materializes", "lumps", or "is observed" at the location of one of the infinitesimal pieces, apparently randomly chosen. Although the cloud shrinks to the accuracy of the observation (if observed by light for example the wavelength of the light limits the accuracy), its momentum spread increases so that Heisenberg's uncertainty principle is still valid.

            Unlike the fixed orbit conceptualization, the electron cloud is not predicted to collapse into a proton while emitting a photon to minimize the sum of electric potential and kinetic energies, since the "cloud" would gain too much kinetic energy, as required to conserve uncertainty. The smear obeys Schrödinger's equation (see also Erwin Schrödinger), which has discrete solutions at differing energy levels. Each of these solutions can be depicted in gray scale, loosely resembling a cloud. This predicts light interactions with an atom, as electrons transition between these cloud states by absorbing or emitting photons equivalent to the difference, or quantum, in their energy. Also, the periodic table is predicted as an electron is added to the lowest unoccupied energy orbital in progressing from hydrogen to helium, and to subsequent elements, with properties that match those predicted by the orbital solutions to Schrödinger's equation.

            The term "electron cloud" carries a connotation that simple language facilitates progress especially in areas such as small scale physics where everyday experience does not extrapolate well. Additional experiments, such as the behaviour of electrons in high speed accelerators, have resulted in more sophisticated models including quantum electrodynamics and superstring theories, although the most exciting discoveries are certainly ahead of us in, as Feynman put it, "the greatest adventure that the human mind has ever begun".

            1.7. Physical mean’s of the wave function

            There is no quantum world. There is only an abstract physical description. It is wrong to think that the task of physics is to find out how nature is. Physics concerns what we can say about nature.

            There is no definitive statement of the Copenhagen Interpretation since it consists of the views developed by a number of scientists and philosophers at the turn of the 20th Century. Thus, there are a number of ideas that have been associated with the Copenhagen interpretation.

            Asher Peres remarked that very different, sometimes opposite, views are presented as the Copenhagen interpretation by different authors.

            The Copenhagen Interpretation involve the following statements:

  1. A system is completely described by a wave function ψ, which represents an observer's knowledge of the system. (Heisenberg)
  2. The description of nature is essentially probabilistic. The probability of an event is related to the square of the amplitude of the wave function. (Max Born)
  3. Heisenberg's uncertainty principle ensures that it is not possible to know the values of all of the properties of the system at the same time; those properties that are not known with precision must be described by probabilities.
  4. (Complementary Principle) Matter exhibits a wave-particle duality. An experiment can show the particle-like properties of matter, or wave-like properties, but not both at the same time.(Niels Bohr)
  5. Measuring devices are essentially classical devices, and measure classical properties such as position and momentum.
  6. The Correspondence Principle of Bohr and Heisenberg, saying that the quantum mechanical description of large systems should closely approximate to the classical description.

            The Copenhagen Interpretation denies that the wave function is real, or is at least non-committal about its reality. There are some who say that there are objective variants of the Copenhagen Interpretation that allow for a "real" wave function, but it is questionable whether that view is really consistent with positivism and some of Bohr's statements. Niels Bohr emphasized that Science is concerned with the predictions of experiments, additional questions are not scientific but rather meta-physical. Bohr was heavily influenced by positivism. On the other hand, Bohr and Heisenberg were not in complete agreement, and took different views at different times. Heisenberg in particular was prompted to move towards realism.

            Even if the wave function is not regarded as real, there is still a divide between those who treat it as definitely and entirely subjective, and those who are non-committal or agnostic about the subject.

            An example of the agnostic view is given by von Weiszacker, who, while participating in a colloquium at Cambridge, denied that the CI asserted: "What cannot be observed does not exist". He suggested instead that the CI follows the principle: "What is observed certainly exists; about what is not observed we are still free to make suitable assumptions. We use that freedom to avoid paradoxes."

The subjective view, that the wave function is merely a mathematical tool for calculating probabilities of specific experiment, is a similar approach to the Ensemble interpretation.

            All versions of the Copenhagen interpretation include at least a formal or methodological version of wave function collapse, in which unobserved eigenvalues are removed from further consideration. (In other words, Copenhagenists have never rejected collapse, even in the early days of quantum physics, in the way that many worlds adherents do.)

            An adherent of the subjective view, that the wave function represents nothing but knowledge, would take an equally subjective view of "collapse", as nothing more than an observer becoming informed about something that was previously ambiguous. The existence of collapse as an objective process, with obvious implications about the reality of the wave function, is more contentious.

            Some argue that the concept of collapse of a "real" wave function was introduced by John Von Neumann and was not part of the original formulation of the Copenhagen Interpretation.

            According to a poll at a Quantum Mechanics workshop in 1997, the Copenhagen interpretation is the most widely-accepted specific interpretation of quantum mechanics, followed by the many-worlds interpretation. Although current trends show substantial competition from alternative interpretations, throughout much of the twentieth century the Copenhagen interpretation had strong acceptance among physicists. John Gribbin describes it as having fallen from primacy after the 1980's.

            1.8. Eigenstates and eigenvalues

            The term eigenstate is derived from the German/Dutch word "eigen," which means "inherent" or "characteristic." The word eigenstate is descriptive of the measured state of some object that possesses quantifiable characteristics such as position, momentum, etc. The state being measured and described must be an "observable" (i.e. something that can be experimentally measured either directly or indirectly like position or momentum), and must have a definite value. In the everyday world, it is natural and intuitive to think of everything being in its own eigenstate. Everything appears to have a definite position, a definite momentum, a definite value of measure, and a definite time of occurrence. However, quantum mechanics affirms that it is impossible to pinpoint exact values for the momentum of a certain particle like an electron in a given location at a particular moment in time, or, alternatively, that it is impossible to give an exact location for such an object when the momentum has been measured. Due to the uncertainty principle, statements regarding both the position and momentum of particles can only be given in terms of a range of probabilities, a "probability distribution". Eliminating uncertainty in one term maximizes uncertainty in regard to the second parameter.

Therefore it became necessary to have a way to clearly formulate the difference between the state of something that is uncertain in the way just described, such as an electron in a probability cloud, and effectively contrast it to the state of something that is not uncertain, something that has a definite value. When something is in the condition of being definitely "pinned-down" in some regard, it is said to possess an eigenstate. For example, if the position of an electron has been made definite, it is said to have an eigenstate of position.

            A definite value, such as the position of an electron that has been successfully located, is called the eigenvalue of the eigenstate of position. The German word "eigen" was first used in this context by the mathematician David Hilbert in 1904. Schrödinger's wave equation gives wavefunction solutions, meaning the possibilities where the electron might be, just as does Heisenberg's probability distribution. As stated above, when a wavefunction collapse occurs because something has been done to locate the position of an electron, the electron's state becomes an eigenstate of position, meaning that the position has a known value.

            1.9. The Pauli Exclusion Principle

            There was a doublet, meaning a pair of lines, in the spectrum of a hydrogen atom that was unaccounted for. This meant that there was more energy in the electron orbital from magnetic moment than had previously been described. Wolfgang Pauli, when studying alkali metals, had introduced what he called a "two-valued quantum degree of freedom" associated with the electron in the outermost shell. "Degrees of freedom" simply means the number of possible independent ways a particle may move. This led to the Pauli Exclusion Principle, which predicted that no more than two electrons can inhabit the same orbital. It also predicted that any neutron, electron, or proton (types of fermions) could not exist in the same quantum state within an atom. According to Schrödinger's equation, there are three quantum states of the electron, but if two electrons can be in the same orbital, there has to be another quantum number to distinguish those two electrons from each other and to describe the extra magnetic moment shown in the atomic spectrum. In early 1925, the young physicists Uhlenbeck and Goudsmit introduced a theory that the electron rotates in space in the same way that the earth rotates on its axis. This would account for the missing magnetic moment and allow for two electrons in the same orbital to be different if their spin was in opposite directions to each other, thus satisfying the Exclusion Principle.

            The Pauli Exclusion Principle states that no electron (or other fermion) can be in the same quantum state as another within an atom. This has an effect on the probability distribution of the electron further defining the number of cells in its phase space. The minimum limit is the limit of the Uncertainty Principle and the Exclusion Principle states that no two electrons can be within this same minimum space defined by the Uncertainty Principle.

            Therefore, a single electron in its orbital when defined by its quantum state which is its wavefunction which is defined by its four quantum numbers cannot have the same four quantum numbers of another electron in that atom. Where two electrons are in the same n-sphere and therefore share the same principal quantum number, they must then have some other unique quantum number of shape l, magnetic moment m or spin s. Even in the formation of degenerate gases where the electrons are not in an orbital around the nucleus of an atom, they must still follow the Pauli Exclusion Principle when in a confined space.

            1.10. Schrödinger Wave Equation

            In physics, the Schrödinger equation, proposed by the Austrian physicist Erwin Schrödinger in 1926, describes the space- and time-dependence of quantum mechanical systems. It is of central importance in non-relativistic quantum mechanics, playing a role for microscopic particles analogous to Newton's second law in classical mechanics for macroscopic particles. Microscopic particles include elementary particles, such as electrons, as well as systems of particles, such as atomic nuclei. Macroscopic particles vary in mass from cells to the galactic superclusters.

            Schrödinger's great insight, late in 1925, was to express the phase of a plane wave as a complex factor:

ψ≈exp(i(kx-ωt))                                                              (13)

and to realize that since

then

                                                   (14)

and similarly since:

then

                                              (15)

and hence:

                                                     (16)

so that, again for a plane wave, he got:

       (17)

And by inserting these expressions into the Newtonian formula for a particle with total energy E, mass m, moving in a potential V:

                                                            (18)

(simply the sum of the kinetic energy and potential energy; the plane wave model assumed V = 0) he got his famed equation for a single particle in the 3-dimensional case in the presence of a potential:

                                                   (19)

            Using this equation, Schrödinger computed the spectral lines for hydrogen by treating a hydrogen atom's single negatively charged electron as a wave, , moving in a potential well, V, created by the positively charged proton. This computation tallied with experiment, the Bohr model and also the results of Werner Heisenberg's matrix mechanics - but without having to introduce Heisenberg's concept of non-commuting observables. Schrödinger published his wave equation and the spectral analysis of hydrogen in a series of four papers in 1926.

            The Schrödinger equation defines the behaviour of ψ, but does not interpret what is. Schrödinger tried unsuccessfully to interpret it as a charge density. In 1926 Max Born, just a few days after Schrödinger fourth and final paper was published, successfully interpreted as a probability amplitude, although Schrödinger was never reconciled to this statistical or probabilistic approach.

Model of the Schrödinger atom, showing the nucleus with two protons (blue) and two neutrons (red), orbited by two electrons (waves)

Fig.5. Model of the Schrödinger atom show the nucleus with two protons and two neutrons,

orbited by two electrons (waves)

 

            Because particles could be described as waves, later in 1925 Erwin Schrödinger analyzed what an electron would look like as a wave around the nucleus of the atom. Using this model, he formulated his equation for particle waves. Rather than explaining the atom by analogy to satellites in planetary orbits, he treated everything as waves whereby each electron has its own unique wavefunction.

            1.11. Time-independent Schröginger equation

For many real-world problems the energy operator (H), does not depend on time. Then it can be shown that the time-dependent Schrödinger equation simplifies to the time-independent Schrödinger equation, which has the well-known appearance HΨ=EΨ.

An example of a simple one-dimensional time-independent Schrödinger equation for a particle of mass m, moving in a potential U(x) is:

The analogous 3-dimensional time-independent equation is:

or

                                                    (20)

where Ñis the del operator.

            For every time-independent Hamiltonian, H, there exists a set of quantum states, , known as energy eigenstates, and corresponding real numbers En satisfying the eigenvalue equation,

.

            Such a state possesses a definite total energy, whose value En is the eigenvalue of the Hamiltonian. The corresponding eigenvector yn is normalizable to unity. This eigenvalue equation is referred to as the time-independent Schrödinger equation. We purposely left out the variable(s) on which the wavefunction yn depends. In the first example above it depends on the single variable x and in the second on x, y, and z—the components of the vector r. In both cases the Schrödinger equation has the same appearance, but its Hamilton operator is defined on different function (state, Hilbert) spaces. In the first example the function space consists of functions of one variable and in the second example the function space consists of functions of three variables.

Self-adjoint operators, such as the Hamiltonian, have the property that their eigenvalues are always real numbers, as we would expect, since the energy is a physically observable quantity. Sometimes more than one linearly independent state vector corresponds to the same energy En. If the maximum number of linearly independent eigenvectors corresponding to En equals k, we say that the energy level En is k-fold degenerate. When k=1 the energy level is called non-degenerate.

            On inserting a solution of the time-independent Schrödinger equation into the full Schrödinger equation, we get

                                                           .
            It is relatively easy to solve this equation. One finds that the energy eigenstates (i.e., solutions of the time-independent Schrödinger equation) change as a function of time only trivially, namely, only by a complex phase:

.

It immediately follows that the probability amplitude,

is time-independent. Because of a similar cancellation of phase factors in bra and ket, all average (expectation) values of time-independent observables (physical quantities) computed from y(t) are time-independent.

            Energy eigenstates are convenient to work with because they form a complete set of states. That is, the eigenvectors {|nñ}form a basis for the state space. We introduced here the short-hand notation |nñ=yn . Then any state vector that is a solution of the time-dependent Schrödinger equation (with a time-independent H) |y(t)ñ can be written as a linear superposition of energy eigenstates:

,       ,  .

            (The last equation enforces the requirement that |y(t)ñ, like all state vectors, may be normalized to a unit vector.) Applying the Hamiltonian operator to each side of the first equation, the time-dependent Schrödinger equation in the left-hand side and using the fact that the energy basis vectors are by definition linearly independent, we readily obtain

            Therefore, if we know the decomposition of |y(t)ñ into the energy basis at time t = 0, its value at any subsequent time is given simply by

            Note that when some values cn(0) are not equal to zero for differing energy values En, the left-hand side is not an eigenvector of the energy operator H. The left-hand is an eigenvector when the only cn(0)-values not equal to zero belong the same energy, so that can be factored out. In many real-world application this is the case and the state vector (containing time only in its phase factor) is then a solution of the time-independent Schrödinger equation.

            Analytical solutions of the time-independent Schrödinger equation can be obtained for a variety of relatively simple conditions. These solutions provide insight into the nature of quantum phenomena and sometimes provide a reasonable approximation of the behavior of more complex systems (e.g., in statistical mechanics, molecular vibrations are often approximated as harmonic oscillators). Several of the more common analytical solutions can be found in the list of quantum mechanical systems with analytical solutions.

            For many systems, however, there is no analytic solution to the Schrödinger equation. In these cases, one must resort to approximate solutions. Some of the common techniques are:

- Perturbation theory

- The variational principle underpins many approximate methods (like the popular Hartree-Fock method which is the basis of the post Hartree-Fock methods);

- Quantum Monte Carlo methods;

- Density functional theory;

- The WKB approximation;

- discrete delta-potential method.

Example

Let |1ñ and |2ñ be degenerate eigenstates of the time-independent Hamiltonian H:

Suppose a solution y(t) of the full (time-dependent) Schrödinger equation of has the form at t = 0: |y(0)ñ=c1|1ñ+c2|2ñ.

Hence, because of the discussion above, at t > 0 :

which shows that y(t) only depends on time in a trivial way (through its phase), also in the case of degeneracy.

Apply now H:

Conclusion: The wavefunction y(t) with the given initial condition (its form at t = 0), remains a solution of the time-independent Schrödinger equation Hψ(t) = Eψ(t) for all times t > 0.

            1.12. The Time-Dependent Schrödinger Equation

            We are now ready to consider the time-dependent Schrödinger equation. Although we were able to derive the single-particle time-independent Schrödinger equation starting from the classical wave equation and the de Broglie relation, the time-dependent Schrödinger equation cannot be derived using elementary methods and is generally given as a postulate of quantum mechanics. It is possible to show that the time-dependent equation is at least reasonable if not derivable, but the arguments are rather involved.

            The single-particle three-dimensional time-dependent Schrödinger equation is

                            (21)

where $V$is assumed to be a real function and represents the potential energy of the system (a complex function $V$will act as a source or sink for probability, as shown in Merzbacher, problem 4.1). Wave Mechanics is the branch of quantum mechanics with equation (21) as its dynamical law. Note that equation (21) does not yet account for spin or relativistic effects.

Of course the time-dependent equation can be used to derive the time-independent equation. If we write the wavefunction as a product of spatial and temporal terms, ψ(r,t)=ψ(r)f(t), then equation (21) becomes

                          (22)

or

                            (23)

Since the left-hand side is a function of  t only and the right hand side is a function of r only, the two sides must equal a constant. If we tentatively designate this constant E (since the right-hand side clearly must have the dimensions of energy), then we extract two ordinary differential equations, namely

                                                   (24)

and

                                 (25)                        

       The latter equation is once again the time-independent Schrödinger equation. The former equation is easily solved to yield

                                                    (26)

            The Hamiltonian in equation (25) is a Hermitian operator, and the eigenvalues of a Hermitian operator must be real, so $E$is real. This means that the solutions f(t) are purely oscillatory, since f(t) never changes in magnitude (recall Euler's formula ).

            Thus if                                                                                        (27)

then the total wave function  ψ(r,t) differs from ψ(r) only by a phase factor of constant magnitude. There are some interesting consequences of this. First of all, the quantity | ψ(r,t)|2 is time independent, as we can easily show:

           (28)

Secondly, the expectation value for any time-independent operator is also time-independent, if ψ(r,t) satisfies equation (27). By the same reasoning applied above,

                                 (29)

            For these reasons, wave functions of the form (27) are called stationary states. The state ψ(r,t) is ``stationary,'' but the particle it describes is not!

            Of course equation (27) represents a particular solution to equation (21). The general solution to equation (21) will be a linear combination of these particular solutions, i.e.

                                               (30)

1.13. Dirac Wave Equation

            In 1928, Paul Dirac worked out a variation of the Schrödinger equation that accounted for a fourth property of the electron in its orbital. Paul Dirac introduced the fourth quantum number called the spin quantum number designated by the letter s to the new Dirac equation of the wavefunction of the electron. In 1930, Dirac combined Heisenberg's matrix mechanics with Schrödinger's wave mechanics into a single quantum mechanical representation in his Principles of Quantum Mechanics. The quantum picture of the electron was now complete.

            All of the above development of quantum theory was based mainly on the atomic spectrum of the hydrogen atom. This is due to the fact that each atom of each element produces a unique pattern of spectral lines when light from each different kind of element is passed through a prism. Scientists could not study the electron and nucleus of the atom itself because they cannot be seen. Even today with High-resolution Scanning Tunneling Electron Microscopes we can only get images of the atom as a blurry fuzzball. However, the spectral lines of the atom reveal the orbits of electrons and the energy that can be expected. It was basically this study of the spectroscopic analysis of first the hydrogen atom and then the helium atom that led to quantum theory. Therefore, the mathematical formulas were made to fit the picture of the atomic spectrum. That is why quantum mechanics is sometimes referred to as a form of mathematical physics.

            Chapter 2. Quantum Theory of the atom

            2.1. Hydrogen-like atom

            A hydrogen-like atom is an atom with one electron. Except for the hydrogen atom itself (which is neutral) these atoms carry the positive charge e(Z-1), where Z is the atomic number of the atom. Because hydrogen-like atoms are two-particle systems with an interaction depending only on the distance between the two particles, their (non-relativistic) Schrödinger equation can be solved in analytic form. The solutions are one-electron functions and are referred to as hydrogen-like atomic orbitals.

            The atomic orbitals of hydrogen-like atoms are solutions to the Schrödinger equation in a spherically symmetric potential. In this case, the potential term is the potential given by Coulomb's law:

V(r) = -\frac{1}{4 \pi \epsilon_0} \frac{Ze^2}{r}                                                             (31)

where ε0 is the permittivity of the vacuum, Z is the atomic number (charge of the nucleus), e is the elementary charge (charge of an electron),  r is the distance of the electron from the nucleus.

            If to accept that a atom nucleus motionless the full energy of atom will be equal to the sum kinetic and potential energies. Then the operator of Hamilton will have a following appearance:

                                      (32)

Here (ž2/2m)∆ is kinetic energy operator and U(r) or –Ze/r 2is potential energy operator (in this expression for simplicity assume that 4πε=1).

The Schrödinger equation for a single particle of mass m moving in a central potential (for example hydrogen atom) (one that depends only on the radial coordinate r) can be written as

                                              (33)

This equation is not separable in cartesian coordinates (x,y,z) because of the way x,y, and z appear together in the square root. However, it is separable in spherical coordinates. Because of the decision of these equation can be written as a product of two separate functions. By introducing the spherical variables we get following wave function

y(x,y,z)=y(r, J, j)= R(r)Y(J, j)                                    (34)

where r, J, j  are spherical coordinates of electron and R(r) is a radial wave function and Y(q, j) is an  angular wave function.

            After writing the wave function as a product of functions (in spherical coordinates), we arrive at the following Schrödinger equation:

                (35)

where μ is, approximately, the mass of the electron. More accurately, it is the reduced mass of the system consisting of the electron and the nucleus.

            Different values of l give solutions with different angular momentum, where l (a non-negative integer) is the quantum number of the orbital angular momentum. The magnetic quantum number m (satisfying  - l£ m £ l) is the (quantized) projection of the orbital angular momentum on the z-axis.

Angular wave function can is received by the decision of the following equation

                        (36)

            2.2. Non-relativistic wave function and energy

In addition to l and m, a third integer n > 0, emerges from the boundary conditions placed on R. The functions R and Y that solve the equations above depend on the values of these integers, called quantum numbers. It is customary to subscript the wave functions with the values of the quantum numbers they depend on. The final expression for the normalized radial wave function is:

               (37)

where:

Note that aμ approximately equal to a0 (the Bohr radius). If the mass of the nucleus is infinite then μ=me and aμ=a0.

             Energy eigenvalues of the atom is equal to

                                                         (38)

which depend only on n, not on l or m. Latest formula corresponds to results received by N.Bohr for hydrogen atom.

The decision of angular equation can be received using a technique of separation of variables, i.e.:

Ylm(J, j)=Θlm(θ)Φm(φ)

Using this substitution we get folloving equation for an angular wave function:

                         (39)

Applying that  method give us two independent angular equation, i.e. :

                                        (40)

and

                                                      (41)

After solving each equations we have receive the following common decision:

                                      (42)

here  is associated Legendre functions.

            These functions are denoted. We put the superscript in parentheses to avoid confusing it with an exponent. Their most straightforward definition is in terms of derivatives of ordinary Legendre polynomials (m ≥ 0)

P_\ell^{(m)}(x) = (-1)^m\ (1-x^2)^{m/2}\ \frac{d^m}{dx^m}\left(P_\ell(x)\right)\,                    (43)

The ( − 1)m factor in this formula is known as the Condon-Shortley phase. Some authors omit it.

            Since, by Rodrigues' formula,

P_\ell(x) = \frac{1}{2^\ell\,\ell!} \  \frac{d^\ell}{dx^\ell}\left([x^2-1]^\ell\right),                                                  (44)

one obtains

P_\ell^{(m)}(x) = \frac{(-1)^m}{2^\ell \ell!} (1-x^2)^{m/2}\  \frac{d^{\ell+m}}{dx^{\ell+m}}(x^2-1)^\ell.                                (45)

This equation allows extension of the range of m to: -l ≤ m ≤ l.

The spherical coordinates used in this article are consistent with those used by physicists, but differ from those employed by mathematicians (see spherical coordinates). In particular, the colatitude θ, or polar angle, ranges from 0 ≤ θ ≤ π and the longitude φ, or azimuth, ranges from 0 ≤ φ ≤ 2π. Thus, θ is 0 at the North Pole, π/2 at the Equator, and π at the South Pole.

            When Laplace's equation is solved on the surface of the sphere, the periodic boundary conditions in φ, as well as regularity conditions at both the north and south poles, ensure that the degree and order m are integers that satisfy ≥ 0 and |m| ≤ . In contrast, if the function f were only to have been defined for θ ≤ θ0, then the resulting spherical cap harmonics would have been defined for integer order, but non-integer degree. The general solution to Laplace's equation is a linear combination of the spherical harmonic functions multiplied by the solutions of R(r).

 

In atoms with a single electron (essentially the hydrogen atom), the energy of an orbital (and, consequently, of any electrons in the orbital) is determined exclusively by n. The n = 1 orbital has the lowest possible energy in the atom. Each successively higher value of n has a higher level of energy, but the difference decreases as n increases. For high n, the level of energy becomes so high that the electron can easily escape from the atom.

In atoms with multiple electrons, the energy of an electron depends not only on the intrinsic properties of its orbital, but also on its interactions with the other electrons. These interactions depend on the detail of its spatial probability distribution, and so the energy levels of orbitals depend not only on n but also on \ell. Higher values of \ell are associated with higher values of energy; for instance, the 2p state is higher than the 2s state. When \ell= 2, the increase in energy of the orbital becomes so large as to push the energy of orbital above the energy of the s-orbital in the next higher shell; when \ell= 3 the energy is pushed into the shell two steps higher.

The energy order of the first 24 subshells is given in the following table. Each cell represents a subshell with n and \ellgiven by its row and column indices, respectively. The number in the cell is the subshell's position in the sequence. Empty cells represent subshells that do not exist.

            2.3. The quantum numbers

An atomic orbital is uniquely identified by the values of the three quantum numbers, and each set of the three quantum numbers corresponds to exactly one orbital, but the quantum numbers only occur in certain combinations of values. The rules governing the possible values of the quantum numbers are as follows:

·        The principal quantum number n is always a positive integer. In fact, it can be any positive integer, but for reasons discussed below, large numbers are seldom encountered. Each atom has, in general, many orbitals associated with each value of n; these orbitals together are sometimes called a shell.

·        The azimuthal quantum number \ell is a non-negative integer. Within a shell where n is some integer n0, \ell ranges across all (integer) values satisfying the relation 0 \le \ell \le n_0-1. For instance, the n = 1 shell has only orbitals with \ell=0, and the n = 2 shell has only orbitals with \ell=0, and \ell=1. The set of orbitals associated with a particular value of \ellare sometimes collectively called a subshell.

·        The magnetic quantum number m_\ellis also always an integer. Within a subshell where \ellis some integer \ell_0, m_\ell  ranges thus: -\ell_0 \le m_\ell \le \ell_0.

The above results may be summarized in the following table. Each cell represents a subshell, and lists the values of m_\ellavailable in that subshell. Empty cells represent subshells that do not exist.

 

\ell = 0

1

2

3

4

...

n = 1

ml = 0

 

 

 

 

 

2

0

-1, 0, 1

 

 

 

 

3

0

-1, 0, 1

-2, -1, 0, 1, 2

 

 

 

4

0

-1, 0, 1

-2, -1, 0, 1, 2

-3, -2, -1, 0, 1, 2, 3

 

 

5

0

-1, 0, 1

-2, -1, 0, 1, 2

-3, -2, -1, 0, 1, 2, 3

-4, -3, -2 -1, 0, 1, 2, 3, 4

 

...

...

...

...

...

...

...

Subshells are usually identified by their n- and \ell-values. n is represented by its numerical value, but \ellis represented by a letter as follows: 0 is represented by 's', 1 by 'p', 2 by 'd', 3 by 'f', and 4 by 'g'. For instance, one may speak of the subshell with n = 2 and \ell=0 as a '2s subshell'.

            2.4. Atomic orbital

            An atomic orbital is a mathematical description of the region in which an electron may be found around a single atom. Specifically, atomic orbitals are the possible quantum states of the individual electrons in the electron cloud around a single atom.

            Classically, the electrons were thought to orbit the atomic nucleus, much like the planets around the Sun (or more accurately, a moth orbiting very quickly around a lamp). Explaining the behavior of the electrons that "orbit" an atom was one of the driving forces behind the development of quantum mechanics. In quantum mechanics, atomic orbitals are described as wave functions over space, indexed by the n, l, and m quantum numbers of the orbital or by the names as used in electron configurations, as shown on the right. As electrons cannot be described as solid particles (as a planet or a moth) in this way, a more accurate analogy would be that of a huge atmosphere, the spatially distributed electron, around a tiny planet which is the atomic nucleus. Hence the term "orbit" was substituted with something else: orbital.

            The orbital names (s, p, d, f, g, h,...) are derived from the quality of their spectroscopic lines: sharp, principal, diffuse and fundamental, the rest being named in alphabetical order. For mnemonic reasons, some call them spherical & peripheral.

Orbitals are given names in the form:            Xtypey

where X is the energy level corresponding to the principal quantum number n, type is a lower-case letter denoting the shape or subshell of the orbital and it corresponds to the angular quantum number l, and y is the number of electrons in that orbital.

            For example, the orbital 1s2 (pronounced "one ess two") has two electrons and is the lowest energy level (n = 1) and has an angular quantum number of l = 0. In some cases, the principal quantum number is given a letter associated with it. For n = 1, 2, 3, 4, 5 ....., the letters associated with those numbers are K, L, M, N, O .... respectively.

In quantum mechanics, the state of an atom, i.e. the eigenstates of the atomic Hamiltonian, is expanded (see configuration interaction expansion and basis (linear algebra)) into linear combinations of anti-symmetrized products (Slater determinants) of one-electron functions. The spatial components of these one-electron functions are called atomic orbitals. When one considers also their spin component, one speaks of atomic spin orbitals.

            In atomic physics, the atomic spectral lines correspond to transitions (quantum leaps) between quantum states of an atom. These states are labelled by a set of quantum numbers summarized in the term symbol and usually associated to particular electron configurations, i.e. by occupations schemes of atomic orbitals (e.g. 1s2 2s2 2p6 for the ground state of neon -- term symbol: 1S0).

            This notation means that the corresponding Slater determinants have a clear higher weight in the configuration interaction expansion. The atomic orbital concept is therefore a key concept for visualizing the excitation process associated to a given transition. For example, one can say for a given transition that it corresponds to the excitation of an electron from an occupied orbital to a given unoccupied orbital. Nevertheless one has to keep in mind that electrons are fermions ruled by Pauli exclusion principle and cannot be distinguished from the other electrons in the atom. Moreover, it sometimes happens that the configuration interaction expansion converges very slowly and that one cannot speak about simple one-determinantal wave function at all. This is the case when electron correlation is large.

Fundamentally, an atomic orbital is a one-electron wavefunction. Don't forget that when thinking about orbitals, we are often bombarded (even if we don't know it) by the Hartree-Fock vision of molecular orbital theory.

            2.5. The shapes of orbitals

            Here it is shown the shapes of the first five atomic orbitals. The colors show the wavefunction phase. Any discussion of the shapes of electron orbitals is necessarily imprecise, because a given electron, regardless of which orbital it occupies, can at any moment be found at any distance from the nucleus and in any direction due to the uncertainty principle.

            However, the electron is much more likely to be found in certain regions of the atom than in others. Given this, a boundary surface can be drawn so that the electron has a high probability to be found anywhere within the surface, and all regions outside the surface have low values. The precise placement of the surface is arbitrary, but any reasonably compact determination must follow a pattern specified by the behavior of Y2, the square of the wavefunction. This boundary surface is what is meant when the "shape" of an orbital is mentioned.

Generally speaking, the number n determines the size and energy of the orbital: as n increases, the size of the orbital increases.

Also in general terms, \ell determines an orbital's shape, and  ml its orientation. However, since some orbitals are described by equations in complex number, the shape sometimes depends on ml also. The single s-orbitals (\ell=0 ) are shaped like spheres. For n=1 the sphere is "solid" (it is most dense at the center and fades exponentially outwardly), but for n=2 or more, each single s-orbital is composed of spherically symmetric surfaces which are nested shells (i.e., the "wave-structure" is radial, following a sinusoidal radial component as well). The s-orbitals for all n numbers are the only orbitals with an anti-node (a region of high wave function density) at the center of the nucleus. All other orbitals (p, d, f, etc.) have angular momentum, and thus avoid the nucleus (having a wave node at the nucleus).

The three p-orbitals have the form of two ellipsoids with a point of tangency at the nucleus (sometimes referred to as a dumbbell). The three p-orbitals in each shell are oriented at right angles to each other, as determined by their respective values of ml.

Four of the five d-orbitals look similar, each with four pear-shaped balls, each ball tangent to two others, and the centers of all four lying in one plane, between a pair of axes. Three of these planes are the xy-, xz-, and yz-planes, and the fourth has the centres on the x and y axes. The fifth and final d-orbital consists of three regions of high probability density: a torus with two pear-shaped regions placed symmetrically on its z axis.

There are seven f-orbitals, each with shapes more complex than those of the d-orbitals.

The shapes of atomic orbitals in one-electron atom are related to 3-dimensional spherical harmonics. Following orbitals table  shows all orbital configurations up to 7s, therefore it covers the simple electronic configuration for all elements from the periodic table up to Ununbium (element 112) with the exception of Lawrencium (element 103), which would require a 7p orbital

            2.6. Hund's rules

            In atomic physics, Hund's rules, (occasionally called the "bus seat rule") refer to a simple set of rules used to determine which is the term symbol that corresponds to the ground state of a multi-electron atom. They are named in honour of Friedrich Hund who contributed Hund's Rule.

The four rules are:

Rule #1

            It can be shown that for full orbitals and suborbitals both the residual electrostatic term (repulsion between electrons) and the spin-orbit interaction can only shift all the energy levels together. Thus when determining the ordering of energy levels in general only the outer valence electrons need to be considered.

Rule #2

            Due to the Pauli exclusion principle, two electrons cannot share the same set of quantum numbers within the same system. Therefore, there is room for only two electrons in each spatial orbital. One of these electrons must have (for some chosen direction z), Sz =1/2, and the other must have Sz =1/2. Hund's second rule states that the lowest energy atomic state is the one which maximizes the sum of the S values for all of the electrons in the system, maximizing the number of unpaired electrons.

            It is often stated that this is the lowest energy atomic state because it forces the unpaired electrons to reside in different spatial orbitals, and this results in a larger average distance between the two electrons, reducing electron-electron repulsion energy. But, in fact, careful calculations have shown that this explanation can be wrong, at least for light systems.

Rule #3

            This rule deals again with reducing the repulsion between electrons. It can be understood from the classical picture that if all electrons are orbiting in the same direction


 

s (l=0)

p (l=1)

d (l=2)

f (l=3)

n=1

 

 

 

 

n=2

 

 

 

n=3

Fig.6. The

 

n=4

shapes

n=5

of  orbitals

. . .

n=6

 

. . .

. . .

n=7

 

. . .

. . .

. . .

 


(higher orbital angular momentum) they meet less often than if some of them orbit in opposite directions. In that last case the repulsive force increases, which separates electrons. This adds potential energy to them, so their energy level is higher.

Rule #4

            This rule considers the energy shifts due to spin-orbit coupling. In the case where the spin-orbit coupling is weak compared to the residual electrostatic, where L and S are still good quantum numbers the splitting is given by:

ΔE=ζ(L,S){L,S}= ζ(L,S){J(J+1)-L(L+1)-S(S+1)}

The value of  ζ(L,S) changes from plus to minus for shells greater than half full. The second term gives the dependence of the ground state on the magnitude of J.

Example

            Hund's rules applied to Si. The up arrows signify electrons with up-spin. The boxes represent different magnetic quantum numbers. As an example, consider the ground state of silicon. The electronic configuration of Si is 1s2, 2s2, 2p6, 3s2, 3p2. Applying the first rule, only the outer 3s2electrons need be considered.

            Magnetic quantum levels of  Si  

           -1           0           +1

 

 

            The possible multiplets are 1S, 3S, 1P, 3P, 1D, 3D; of those, 3S and 3D are not allowed because of the exclusion principle. The second rule now states that the triplet state 3P with S=1 has the lowest energy. There is no choice of triplets (S = 1) states, so the third rule is not required. [If the 3D(L=2) state were allowed, then the third rule would come into force and state that it was more favourable than the 3P (L=1) state.] The triplet 3P consists of three states, J = 2, 1, 0. With only two of six possible electrons in the shell, it is less than half-full and thus 3P0 is the ground state.

            2.7. Electron placement and the periodic table

            Several rules govern the placement of electrons in orbitals (electron configuration). The first dictates that no two electrons in an atom may have the same set of values of quantum numbers (this is the Pauli exclusion principle). These quantum numbers include the three that define orbitals, as well as s, or spin quantum number. Thus, two electrons may occupy a single orbital, so long as they have different values of s. However, only two electrons, because of their spin, can be associated with each orbital.

            Additionally, an electron always tries to occupy the lowest possible energy state. It is possible for it to occupy any orbital so long as it does not violate the Pauli exclusion principle, but if lower-energy orbitals are available, this condition is unstable. The electron will eventually lose energy (by releasing a photon) and drop into the lower orbital. Thus, electrons fill orbitals in the order specified by the energy sequence given above.

            This behavior is responsible for the structure of the periodic table. The table may be divided into several rows (called 'periods'), numbered starting with 1 at the top. The presently known elements occupy seven periods. If a certain period has number i, it consists of elements whose outermost electrons fall in the ith shell.

            The periodic table may also be divided into several numbered rectangular 'blocks'. The elements belonging to a given block have this common feature: their highest-energy electrons all belong to the same \ell-state (but the n associated with that \ell-state depends upon the period). For instance, the leftmost two columns constitute the 's-block'. The outermost electrons of Li and Be respectively belong to the 2s subshell, and those of Na and Mg to the 3s subshell.

            The number of electrons in a neutral atom increases with the atomic number. The electrons in the outermost shell, or valence electrons, tend to be responsible for an element's chemical behavior. Elements that contain the same number of valence electrons can be grouped together and display similar chemical properties.

            Chapter 3. Basic types of Chemical bonds

            3.1. The basic concept about chemical bond

            A chemical bond is the physical process responsible for the attractive interactions between atoms and molecules, and that which confers stability to diatomic and polyatomic chemical compounds. The explanation of the attractive forces is a complex area that is described by the laws of quantum electrodynamics. In practice, however, chemists usually rely on quantum theory or qualitative descriptions that are less rigorous but more easily explained to describe chemical bonding. In general, strong chemical bonding is associated with the sharing or transfer of electrons between the participating atoms. Molecules, crystals, and diatomic gases—indeed most of the physical environment around us—are held together by chemical bonds, which dictate the structure of matter.

            Bonds vary widely in their strength. Generally covalent and ionic bonds are often described as "strong", whereas hydrogen bonds and van der Waals' bonds are generally considered to be "weak". Care should be taken because the strongest of the "weak" bonds can be stronger than the weakest of the "strong" bonds.

Examples of Lewis dot-style chemical bonds between carbon C, hydrogen H, and oxygen O.  Lewis dot depictures represent an early attempt to describe chemical bonding and are still widely used today.

Fig.7. The simplest models of atoms and molecules

 

            Examples of Lewis dot-style chemical bonds between carbon C, hydrogen H, and oxygen O. Lewis dot depictures represent an early attempt to describe chemical bonding and are still widely used today.

Remembering that opposite charges attract, and that the electrons orbiting the nucleus are negatively charged and protons in the nucleus are positively charged, then imagine two atoms near each other which form a covalent bond.

In the simplest view of a so-called covalent bond, one or more electrons—often a pair as in this example—is drawn into the space between the two atomic nuclei. Here the negatively charged electrons are attracted to the positive charges of both nuclei, instead of just their own. This overcomes the repulsion between the two positively charged nuclei of the two atoms and so this overwhelming attraction holds the two nuclei in a relatively fixed configuration of equilibrium, even though they will still vibrate about that equilibrium position. In summary, covalent bonding involves sharing of electrons in which the positively charged nuclei of two or more atoms simultaneously attract the negatively charged electrons that are being shared.

In a simplified view of an ionic bond, the positive charge of one of the nuclei overwhelms the positive charge of the other nucleus, thus effectively transferring an electron from one atom to another, causing one atom to assume a net positive charge, and the other to assume a net negative charge. The bond then results from electrostatic attraction between atoms, and the atoms become positive or negatively charged ions.

All bonds can be explained by quantum theory, but in practice, simplification rules allow chemists to predict the strength, directionality, and polarity of bonds. The octet rule and VSEPR theory are two examples. More sophisticated theories are valence bond theory which includes orbital hybridization and resonance, and the linear combination of atomic orbitals molecular orbital method which includes ligand field theory. Electrostatics are used to describe bond polarities and the effects they have on chemical substances.

Early speculations into the nature of the chemical bond, from as early as the 12th century, supposed that certain types of chemical species were joined by a type of chemical affinity. In 1704, Issac Newton famously outlined his atomic bonding theory, in "Query 31" of his Opticks, whereby atoms attach to each other by some "force". Specifically, after acknowledging the various popular theories, in vogue at the time, of how atoms were reasoned to attach to each other, i.e. “hooked atoms”, “glued together by rest”, or “stuck together by conspiring motions”, Newton states that he would rather infer from their cohesion, that:

Particles attract one another by some force, which in immediate contact is exceedingly strong, at small distances performs the chemical operations, and reaches not far from the particles with any sensible effect.

            In 1819, on the heels of the invention of the voltaic pile, Jöns Jakob Berzelius developed a theory of chemical combination stressing the electronegative and electropositive character of the combining atoms. By the mid 19th century, Edward Frankland, F.A. Kekule, A.S. Couper, A.M. Butlerov, and Hermann Kolbe, building on the theory of radicals, developed the theory of valency, originally called “combining power”, in which compounds were joined owing to an attraction of positive and negative poles. In 1916, chemist Gilbert N. Lewis developed the concept of the electron-pair bond, in which two atoms may share one to six electrons, thus forming the single electron bond, a single bond, a double bond, or a triple bond:

Fig. 8. Lewis-type Chemical bond

           

            In Lewis' own words:

            An electron may form a part of the shell of two different atoms and cannot be said to belong to either one exclusively.

            That same year, Walther Kossel put forward a theory similar to Lewis' only his model assumed complete transfers of electrons between atoms, and was thus a model of polar bonds. Both Lewis and Kossel structured their bonding models on that of Abegg's rule (1904).

In 1927, the first mathematically complete quantum description of a simple chemical bond, i.e. that produced by one electron in the hydrogen molecular ion, H2+, was derived by the Danish physicist Oyvind Burrau. This work showed that the quantum approach to chemical bonds could be fundamentally and quantitatively correct, but the mathematical methods used could not be extended to molecules containing more than one electron. A more practical, albeit less quantitative, approach was put forward in the same year by Walter Heitler and Fritz London. The Heitler-London method forms the basis of what is now called valence bond theory. In 1929, the linear combination of atomic orbitals molecular orbital method (LCAO) approximation was introduced by Sir John Lennard-Jones, who also suggested methods to derive electronic structures of molecules of F2 (fluorine) and O2 (oxygen) molecules, from basic quantum principles. This molecular orbital theory represented a covalent bond as a orbitals formed by combining the quantum mechanical Schrödinger atomic orbitals which had been hypothesized for electrons in single atoms. The equations for bonding electrons in multi-electron atoms could not be solved to mathematical perfection (i.e., analytically), but approximations for them still gave many good qualitative preditions and results. Most quantitative calculations in modern quantum chemistry use either valence bond or molecular orbital theory as a starting point, although a third approach, Density Functional Theory, has become increasingly popular in recent years.

In 1935, H. H. James and A. S. Coolidge carried out a calculation on the dihydrogen molecule that, unlike all previous calculation which used functions only of the distance of the electron from the atomic nucleus, used functions which also explicitly added the distance between the two electrons. With up to 13 adjustable parameters they obtained a result very close to the experimental result for the dissociation energy. Later extensions have used up to 54 parameters and give excellent agreement with experiment. This calculation convinced the scientific community that quantum theory could give agreement with experiment. However this approach has none of the physical pictures of the valence bond and molecular orbital theories and is difficult to extend to larger molecules.

Chemical bonds in molecules are indicating in special chemical formulas. However the 3-dimensionality of atoms and molecules makes it difficult to use a single technique for indicating orbitals and bonds. In molecular formulas the chemical bonds (binding orbitals) between atoms are indicated by various different methods according to the type of discussion. Sometimes, they are completely neglected. For example, in organic chemistry chemists are sometimes concerned only with the functional groups of the molecule. Thus, the molecular formula of ethanol (a compound in alcoholic beverages) may be written in a paper in conformational, 3-dimensional, full 2-dimensional (indicating every bond with no 3-dimensional directions), compressed 2-dimensional (CH3–CH2–OH), separating the functional group from another part of the molecule (C2H5OH), or by its atomic constituents (C2H6O), according to what is discussed. Sometimes, even the non-bonding valence shell electrons ( with the 2-dimensionalized approximate directions) are marked, i.e. for elemental carbon .'C'. Some chemists may also mark the respective orbitals, i.e. the hypothetical ethene−4 anion (\/C=C/\ −4) indicating the possibility of bond formation.

Depending on energy of chemical binding  chemical bonds are divided on following types:

  • Strong chemical bonds,
  • Inetrmolecular bonds.

            3.2. Strong chemical bonds

 These chemical bonds are intramolecular forces, which hold atoms together in molecules. In the simplistic localized view of bonding, the number of electrons participating in a bond (or located in a bonding orbital) is typically multiples of two, four, or six, respectively. Even numbers are common because electrons enjoy lower energy states, if paired. Substantially more advanced bonding theories have shown that bond strength is not always a whole number, depending on the distribution of electrons to each atom involved in a bond. For example, the carbons in benzene are connected to each other with about 1.5 bonds, and the two atoms in nitric oxide NO, are connected with about 2.5 bonds. Quadruple bonds are also well known. The type of strong bond depends on the difference in electronegativity and the distribution of the electron orbital paths available to the atoms that are bonded. The larger the difference in electronegativity, the more an electron is attracted to a particular atom involved in the bond, and the more "ionic" properties the bond is said to have ("ionic" means the bond electron(s) are unequally shared). The smaller the difference in electronegativity, the more covalent properties (full sharing) the bond has.

These chemical bonds are intramolecular forces, which hold atoms together in molecules. In the simplistic localized view of bonding, the number of electrons participating in a bond (or located in a bonding orbital) is typically multiples of two, four, or six, respectively. Even numbers are common because electrons enjoy lower energy states, if paired. Substantially more advanced bonding theories have shown that bond strength is not always a whole number, depending on the distribution of electrons to each atom involved in a bond. For example, the carbons in benzene are connected to each other with about 1.5 bonds, and the two atoms in nitric oxide NO, are connected with about 2.5 bonds. Quadruple bonds are also well known. The type of strong bond depends on the difference in electronegativity and the distribution of the electron orbital paths available to the atoms that are bonded. The larger the difference in electronegativity, the more an electron is attracted to a particular atom involved in the bond, and the more "ionic" properties the bond is said to have ("ionic" means the bond electron(s) are unequally shared). The smaller the difference in electronegativity, the more covalent properties (full sharing) the bond has.

Typical bond length (in pm) and bond energies in kJ/mol

 

Bonds

length

energy

Bonds

length

energy

Bonds

length

energy

H-H

H-C

H-N

H-O

H-F

H-Cl

H-Br

H-I

C-C

74

109

101

96

92

127

141

161

154

436

413

391

366

568

432

366

298

348

C=C

CºC

C-O

C-N

C-F

C-Cl

C-Br

C-I

134

120

143

147

135

177

194

214

 

614

839

360

308

488

330

170

216

C-S

N-N

NºN

O-O

F-F

Cl-Cl

Br-Br

I-I

182

145

110

148

142

199

228

267

 

272

170

945

145

158

243

193

151

 

 

Now we shall consider the main types of such bonds.

            3.2.1. Covalent bond

            The term "covalence" in regards to bonding was first used in 1919 by Irving Langmuir in a Journal of American Chemical Society article entitled The Arrangement of Electrons in Atoms and Molecules: “… we shall denote by the term covalence the number of pairs of electrons which a given atom shares with its neighbors”.

The idea of covalent bonding can be traced several years prior to 1919 to Gilbert N. Lewis, who in 1916 described the sharing of electron pairs between atoms. He introduced the so called Lewis notation or electron dot notation or The Lewis Dot Structure in which valence electrons (those in the outer shell) are represented as dots around the atomic symbols. Pairs of electrons located between atoms represent covalent bonds. Multiple pairs represent multiple bonds, such as double and triple bonds. Some examples of Electron Dot Notation are shown in the following figure. An alternative form, in which bond-forming electron pairs are represented as solid lines, is shown alongside.

            Covalent bonding is a common type of bonding, in which the electronegativity difference between the bonded atoms is small or non-existent. Bonds within most organic compounds are described as covalent. See sigma bonds and pi bonds for LCAO-description of such bonding.

            Early concepts in covalent bonding arose from this kind of image of the molecule of methane. Covalent bonding is implied in the Lewis structure that indicates sharing of electrons between atoms.

                                                                     

Fig.9. Formation of covalent chemical bonds according Lewis theory

 

            While the idea of shared electron pairs provides an effective qualitative picture of covalent bonding, quantum mechanics is needed to understand the nature of these bonds and predict the structures and properties of simple molecules. Walter Heitler and Fritz London are credited with the first successful quantum mechanical explanation of a chemical bond, specifically that of molecular hydrogen, in 1927. Their work was based on the valence bond model, which assumes that a chemical bond is formed when there is good overlap between the atomic orbitals of participating atoms. These atomic orbitals are known to have specific angular relationships between each other, and thus the valence bond model can successfully predict the bond angles observed in simple molecules.

            Covalent bonding is characterized by new concept named by the bond order. Depending on it chemical bond can be strong or weak. Bond order is a number that indicates the number of pairs of electrons shared between atoms forming a covalent bond. The term is only applicable to diatomic molecules, but is used to describe bonds within polyatomic compounds as well.

            The most common type of covalent bond is the single bond, the sharing of only one pair of electrons between two atoms. It usually consists of one sigma bond. All bonds with more than one shared pair are called multiple bonds.

            Sharing two pairs is called a double bond. An example is in ethylene (between the carbon atoms). It usually consists of one sigma bond and one π-bond.

            Sharing three pairs is called a triple bond. An example is in hydrogen cyanide (between C and N). It usually consists of one sigma bond and two π-bonds.

            Quadruple bonds are found in the transition metals. Molybdenum and rhenium are the elements most commonly observed with this bonding configuration. An example of a quadruple bond is also found in Di-tungsten tetra(hpp).

            Quintuple bonds have been found to exist in certain dichromium compounds.

The only known molecules with true sextuple bonds (order 6) are diatomic Mo2 and W2, in the gaseous phase at very low temperatures. Although diatomic Cr2 and U2 have formal structures with twelve-electron bonds, their effective bond orders (derived from quantum chemistry calculations) are less than 5. There is strong evidence to believe that no two elements in the periodic table can form a bond with greater order than 6.

            In a more advanced context bond order need not be an integer. A good example of this is bonds between carbon in the molecule benzene where the delocalized molecular orbitals contain 6 pi electrons over six carbons essentially yielding half a pi bond. Together with the sigma bond the bond order is 1.5. Furthermore bond orders of 1.1 for example may be spoken about in more complex terms and are essentially referring to bond strength relative to B.O.=1 bonds.

            In Molecular Orbital Theory bond order is also defined as the difference between the number of bonding electrons and the number of antibonding electrons, divided by two. This often but not always yields the same result as most simpler theories.

            Bond order is an index of bond strength and is used extensively in valence bond theory. In short: the bond order is the net amount of bonding between two atoms.

BO = 1/2 ((no. of bonding electrons) - (no. of antibonding electrons))

            Most bonding of course, is not localized, so the following classification, while powerful and pervasive, is of limited validity. Three center bond do not conform readily to the above conventions.

The bond order concept is used e.g. in molecular dynamics bond order potentials.

            3.2.2. Polar covalent bond

            Polar covalent bonding is intermediate in character between a covalent and an ionic bond.

 [Show more]

A commonly-used example of a polar compound is water (H2O). The electrons of water's hydrogen atoms are strongly attracted to the oxygen atom, and are actually closer to oxygen's nucleus than to the hydrogen nuclei; thus, water has a relatively strong negative charge in the middle (red shade), and a positive charge at the ends (blue shade).

Fig.10. Electric dipols in the water molecule

            Polarity refers to the dipole-dipole intermolecular forces between the slightly positively-charged end of one molecule to the negative end of another or the same molecule. Molecular polarity is dependent on the difference in electronegativity between atoms in a compound and the asymmetry of the compound's structure. For example, water is thought to be polar because of the uneven sharing of its electrons. However, methane is considered non-polar because the carbon shares the hydrogen molecules uniformally.

            3.2.3. Theory of polarity of molecule

            Diagram showing the net effect of symmetrical polar bonds (direction of yellow arrows show the migration of electrons) within boron trifluoride cancelling out to give a net polarity of zero. δ- shows an increase in negative Electrons are not always shared equally between two bonding atoms: one atom might exert more of a force on the electron cloud than the other. This "pull" is termed electronegativity and measures the attraction for electrons a particular atom has. The unequal sharing of electrons within a bond leads to the formation of an electric dipole: a separation of positive and negative electric charge.charge and δ+ shows an increase in positive charge.

            Atoms with high electronegativities - such as fluorine, oxygen, and nitrogen - exert a greater pull on electrons than atoms with lower electronegativities. In a bonding situation this can lead to unequal sharing of electrons between atoms as electrons will spend more time closer to the atom with the higher electronegativity.

Fig.11. Occurrence of electric dipoles in a molecule Boron Trifluoride

 

            Bonds can fall between one of two extremes - being completely non-polar or completely polar. A completely non-polar bond occurs when the electronegativities are identical and therefore possess a difference of zero. A completely polar bond is more correctly termed ionic bonding and occurs when the difference between electronegativities is large enough that one atom takes an electron from the other. The terms "polar" and "non-polar" bonds usually refer to covalent bonds. To determine the polarity of a covalent bond using numerical means, the difference between the electronegativity of the atoms is taken. If the result is between 0.5 and 2 then, generally, the bond is polar.

            A compound is composed of one or more chemical bonds between different atoms. The polarity of each bond within the compound may determine the overall polarity of the compound: how polar or non-polar it is.

            A polar molecule may be polar as a result of polar bonds or as a result of an asymmetric arrangement of non-polar bonds and non bonding pairs of electrons.

Example 1. A polar molecule by virtue of polar bonds (bonds which have unequal sharing of electrons between the two atoms involved in bonding), e.g hydrogen fluoride, HF, where the bonding pair of electron is displaced towards the more electronegative fluorine atom.

Example 2. In ammonia, NH3, the three N-H bonds have only a slight polarity (towards the more electronegative nitrogen atom), however the lone pair of electrons (pointing towards the fourth apex of the approximate tetrahedron, is electron rich and results in a powerful dipole across the whole ammonia molecule.

            A non-polar compound may be non polar because there is (almost) no polarity in the bonds or because of the symmetrical arrangement of polar bonds.

Example 3. Methane, CH4 The four C-H bonds, arranged tetrahedral around the carbon atom, has very little polarity in the bonds and so there is no dipole in the molecule.

Example 4. BF3, boron trifluoride has a trigonal planar arrangement of three polar bonds at 120o This results in no overall dipole in the molecule.

            While molecules can be described as "polar," "non-polar," or "semi-polar," it must be noted that this is often a relative term, with one molecule simply being more polar or more non-polar than another. As such, there are no ultimate properties which can be ascribed to polar or non-polar molecules. However, the following properties are typical of such molecules.

            Examples of household polar molecules include ammonia and sugar (glucose). Polar molecules are generally able to dissolve in water (hydrophilic) due to the polar nature of water. Polar molecules have slightly positive and slightly negatively charged ends.

            Examples of household non-polar compounds include fats, oil and petrol. Most non-polar molecules are water insoluble (hydrophobic) at room temperature. However many non-polar organic solvents, such as turpentine, are able to dissolve non-polar substances - like dissolves like. Also a non-polar compound occurs when there is an equal sharing of electrons between two atoms. When comparing a polar and non-polar molecule with similar molar masses, the polar molecule generally has a higher boiling point, because of the dipole-dipole interaction between their molecules.

Examples of  polar and nonpolar molecule

 

Formula

Description

Example

Polar

AB

Linear Molecules

CO

HAx

Molecules with a single H

HCl

AxOH

Molecules with an OH at one end

C2H5OH

OxAy

Molecules with an O at one end

H2O

NxAy

Molecules with an N at one end

NH3

Nonpolar

Ax

All elements

O2

CxAy

Most carbon compounds

CO2

            Polarity of covalent bonds is dependent on a electronegativity difference between two bonded atoms as follows:

Electronegativity difference predictions

Type of  covalent bonds

Different  in negativity

Non-polar covalent bond

0.0-0.4

Slightly polar bond

0.5-0.9

Moderately polar bond

1-1.3

Highly polar bond

1.4-1.7

Slightly ionic bond

1.8-2.2

Ionic Bond

< 2.3

            3.2.4. Ionic bond

            Ionic bonding is a type of electrostatic interaction between atoms which have an electronegativity difference of over 1.6 (this limit is a convention). These form in a solution between two ions after the excess of the solvent is removed. Ionic charges are commonly between −3e to +7e.

Sodium and chlorine bonding ionically to form sodium chloride. Sodium loses its outer electron endothermically to give it a noble gas electron configuration, and this electron enters the chlorine atom exothermically. The oppositely charged ions are then attracted to each other, and their bonding releases energy. The net transfer of energy is that energy leaves the atoms, so the reaction is able to take place.

Fig. 12. The distribution of electrons on atomic orbitals

            Sodium and chlorine bonding ionically to form sodium chloride. Sodium loses its outer electron endothermically to give it a noble gas electron configuration, and this electron enters the chlorine atom exothermically. The oppositely charged ions are then attracted to each other, and their bonding releases energy. The net transfer of energy is that energy leaves the atoms, so the reaction is able to take place.

            An ionic bond (or electrovalent bond) is a type of chemical bond that can often form between metal and non-metal ions (or polyatomic ions such as ammonium) through electrostatic attraction.

            The metal donates one or more electrons, forming a positively charged ion or cation with a stable electron configuration. These electrons then enter the non metal, causing it to form a negatively charged ion or anion which also has a stable electron configuration. The electrostatic attraction between the oppositely charged ions causes them to come together and form a bond.

For example, common table salt is made of sodium chloride. In the ionic bonding, the sodium (Na) loses one electron, forming a cation, and the chlorine (Cl) gains an electron to form an anion. These ions are then attracted to each other in a 1:1 ratio to form sodium chloride (NaCl).

Na+ + Cl- → NaCl

            Electron configurations of lithium and fluorine. Lithium has one electron in its outer shell, held rather loosely because the ionization energy is low. Fluorine carries 7 electrons in its outer shell. When one electron moves from lithium to fluorine, each ion acquires the noble gas configuration. The bonding energy from the electrostatic attraction of the two oppositely-charged ions has a large enough negative value that the overall bonded state energy is lower than the unbonded state

            The removal of electrons from the atoms is endothermic and causes the ions to have a higher energy. There may also be energy changes associated with breaking of existing bonds or the addition of more than one electron to form anions. However, the attraction of the ions to each other lowers their energy.

            Ionic bonding will occur only if the overall energy change for the reaction is favourable – when the bonded atoms have a lower energy than the free ones. The larger the resulting energy change the stronger the bond. The low electronegativity of metals and high electronegativity of non-metals means that the energy change of the reaction is most favorable when metals lose electrons and non-metals gain electrons.

            Pure ionic bonding is not known to exist. All ionic bonds have a degree of covalent bonding or metallic bonding. The larger the difference in electronegativity between two atoms, the more ionic the bond. Ionic compounds conduct electricity when molten or in solution. They generally have a high melting point and tend to be soluble in water.

             There is the polarization effect in ionic crystals which determine many of physical and chemical properities of that crystals. Ions in crystal lattices of purely ionic compounds are spherical; however, if the positive ion is small and/or highly charged, it will distort the electron cloud of the negative ion. This polarization of the negative ion leads to a build-up of extra charge density between the two nuclei, i.e., to partial covalency. Larger negative ions are more easily polarized, but the effect is usually only important when positive ions with charges of 3+ (e.g., Al3+) are involved (e.g., pure AlCl3 is a covalent molecule). However, 2+ ions (Be2+) or even 1+ (Li+) show some polarizing power because their sizes are so small (e.g., LiI is ionic but has some covalent bonding present).

            Ionic compounds in the solid state form a continuous ionic lattice structure in an ionic crystal. The simplest form of ionic crystal is a simple cubic. This is as if all the atoms were placed at the corners of a cube. This unit cell has a weight that is the same as 1 of the atoms involved. When all the ions are approximately the same size, they can form a different structure called a face-centered cubic (where the weight is 4 ´ atomic weight), but, when the ions are different sizes, the structure is often body-centered cubic (2 times the weight). In ionic lattices the coordination number refers to the number of connected ions.

            In an ionic bond, the atoms are bound by attraction of opposite ions, whereas, in a covalent bond, atoms are bound by sharing electrons. In covalent bonding, the molecular geometry around each atom is determined by VSEPR rules, whereas, in ionic materials, the geometry follows maximum packing rules. Below it is shown some examples of ionic structures.

Substances in ionic form

Common Cations

Stock System Name

Formula

Historic Name

Simple Cations

Aluminium

Al3+

 

Barium

Ba2+

 

Beryllium

Be2+

 

Caesium

Cs+

 

Calcium

Ca2+

 

Chromium(II)

Cr2+

Chromous

Chromium(III)

Cr3+

Chromic

Chromium(VI)

Cr6+

Chromyl

Cobalt(II)

Co2+

Cobaltous

Cobalt(III)

Co3+

Cobaltic

Copper(I)

Cu+

Cuprous

Copper(II)

Cu2+

Cupric

Copper(III)

Cu3+

 

Gallium

Ga3+

 

Gold(I)

Au+

 

Gold(III)

Au3+

 

Helium

He2+

(Alpha particle)

Hydrogen

H+

(Proton)

Iron(II)

Fe2+

Ferrous

Iron(III)

Fe3+

Ferric

Lead(II)

Pb2+

Plumbous

Lead(IV)

Pb4+

Plumbic

Lithium

Li+

 

Magnesium

Mg2+

 

Manganese(II)

Mn2+

Manganous

Manganese(III)

Mn3+

Manganic

Manganese(IV)

Mn4+

Manganyl

Manganese(VII)

Mn7+

 

Mercury(II)

Hg2+

Mercuric

Nickel(II)

Ni2+

Nickelous

Nickel(III)

Ni3+

Nickelic

Potassium

K+

 

Silver

Ag+

 

Sodium

Na+

 

Strontium

Sr2+

 

Tin(II)

Sn2+

Stannous

Tin(IV)

Sn4+

Stannic

Zinc

Zn2+

 

Polyatomic Cations

Ammonium

NH4+

 

Hydronium

H3O+

 

Nitronium

NO2+

 

Mercury(I)

Hg22+

Mercurous

Common Anions

Formal Name

Formula

Alt. Name

Simple Anions

Arsenide

As3−

 

Azide

N3−

 

Bromide

Br−

 

Chloride

Cl−

 

Fluoride

F−

 

Hydride

H−

 

Iodide

I−

 

Nitride

N3−

 

Oxide

O2−

 

Phosphide

P3−

 

Sulphide

S2−

 

Peroxide

O22−

 

Oxoanions

Arsenate

AsO43−

 

Arsenite

AsO33−

 

Borate

BO33−

 

Bromate

BrO3−

 

Hypobromite

BrO−

 

Carbonate

CO32−

 

Hydrogen Carbonate

HCO3−

Bicarbonate

Chlorate

ClO3−

 

Perchlorate

ClO4−

 

Chlorite

ClO2−

 

Hypochlorite

ClO−

 

Chromate

CrO42−

 

Dichromate

Cr2O72−

 

Iodate

IO3−

 

Nitrate

NO3−

 

Nitrite

NO2−

 

Phosphate

PO43−

 

Hydrogen Phosphate

HPO42−

 

Dihydrogen Phosphate

H2PO4−

 

Permanganate

MnO4−

 

Phosphite

PO33−

 

Sulphate

SO42−

 

Thiosulphate

S2O32−

 

Hydrogen Sulphate

HSO4−

Bisulphate

Sulphite

SO32−

 

Hydrogen Sulphite

HSO3−

Bisulphite

Anions from Organic Acids

Acetate

C2H3O2−

 

Formate

HCO2−

 

Oxalate

C2O42−

 

Hydrogen Oxalate

HC2O4−

Bioxalate

Other Anions

Hydrogen Sulphide

HS−

Bisulphide

Telluride

Te2−

 

Amide

NH2−

 

Cyanate

OCN−

 

Thiocyanate

SCN−

 

Cyanide

CN−

 

            3.2.5. Coordinate covalent bond

            Coordinate covalent bonding, sometimes referred to as dative bonding, is a kind of covalent bonding, in which the covalent bonding electrons originate solely from one of the atoms, the electron-pair donor or Lewis base but are approximately equally shared in the formation of a covalent bond. This concept is somewhat fading as chemists increasingly embrace molecular orbital theory.

            Examples of coordinate covalent bonding occur in nitrones and ammonia borane. The arrangement is different from an ionic bond in that the electronegativity difference is small, resulting in covalency. They are shown by an arrow unlike other bonds . This arrow shows its head towards the electron acceptor or lewis acid and its tail towards the lewis base. This bond is seen in ammonium.

            A coordinate covalent bond (also known as dative bond) is a description of covalent bonding between two atoms in which both electrons shared in the bond come from the same atom. The distinction from ordinary covalent bonding is artificial, but the terminology is popular in textbooks, especially those describing coordination compounds. Once the bonds have been formed using this, its strength and description is no different from that of other polar covalent bonds.

            Coordinate covalent bonds are invoked when a Lewis base (an electron donor or giver) donates a pair of electrons to a Lewis acid (an electron acceptor) to give a so-called adduct. The process of forming a dative bond is called coordination. The electron donor acquires a positive formal charge, while the electron acceptor acquires a negative formal charge.

An example of coordinate covalent bonding in the ammonium ion

 

Fig.13. Formation of the coordinate covalent bond in ammonium

           

            Classically, any compound that contains a lone pair of electrons is capable of forming a coordinate bond. The bonding in diverse chemical compounds can be described as coordinate covalent bonding.

            Carbon monoxide (CO) can be viewed as containing one coordinate bond and two "normal" covalent bonds between the carbon atom and the oxygen atom. This highly unusual description illustrates the flexibility of this bonding description. Thus in CO, carbon is the electron acceptor and oxygen is the electron donor.

            The ammonium ion (NH4+), can be viewed as consisting of four coordinate covalent bonds between the protons (the H+ ion) and the nitrogen trianion N3-.

The beryllium dichloride (BeCl2) is described as electron deficient in the sense that the triatomic species (which does exist in the gas phase) features Be centers with four valence electrons. When treated with excess chloride, the Be2+ ion binds four chloride ions to form tetrachloroberyllate anion, BeCl42-, wherein all ions achieve the octet configuration of electrons.

            Coordinate bonding is popularly used to describe coordination complexes, especially involving metal ions. In such complexes, several Lewis bases "donate" their "free" pairs of electrons to an otherwise naked metal cation, which acts as a Lewis acid and "accepts" the electrons. Coordinate bonds form and the resulting compound is called a coordination complex, and the electron donors are called ligands. A more useful description of bonding in coordination compounds is provided by Ligand Field Theory, which embraces molecular orbitals as a description of bonding in such polyatomic compounds.

            Many chemical compounds can serve as ligands, often these contain oxygen, sulfur, nitrogen, and halide ions. The most common ligand is water (H2O), which forms coordination complexes with metal ions (like the hexaaquacopper(II) ion, [Cu(H2O)6]2+). Ammonia (NH3) is also a common ligand, as well as anions, especially fluoride (F-), chloride (Cl-), and cyanide (CN-).

            3.2.6. Bent bonds

            Bent bond, also known as banana bond, is a term in chemistry that refers to a type of covalent chemical bond with a geometry somewhat reminiscent of a banana. Bent bonds are in strained or otherwise sterically hindered molecules those binding orbitals are forced into a banana-like form. Bent bonds are often more susceptible to reactions than ordinary bonds. One of the first bent bond theories for cyclopropane was the so-called Coulson-Moffitt model (1947).

            The term itself is a general representation of electron density or configuration resembling a similar "bent" structure within small ring molecules, such as cyclopropane (C3H6), three-center two-electron bonds found in diborane (B2H6), or as a representation of double or triple bonds within a compound that is an alternative to the sigma and pi bond model.

 

             

Fig. 14. Two kinds of banana bonds are distinguished: 3c-2e and 4c-3e bonds

 

            In three-center two-electron bonds three atoms share two electrons in bonding. This type of bonding occurs in electron deficient compounds like diborane. Each such bond (2 per molecule in diborane) contains a pair of electrons which connect the boron atoms to each other in a banana shape (shown as a more sharply angled section in the stick model at right), with a proton (nucleus of a hydrogen atom) in the middle of the bond, sharing electrons with both boron atoms. Three-center four-electron bonds also exist which explain the bonding in hypervalent molecules. In certain cluster compounds so-called four-center two-electron bonds also have been postulated.

            One- and three-electron bonds have own properities. These bonds with one or three electrons can be found in radical species, which have an odd number of electrons. The simplest example of a 1-electron bond is found in the hydrogen molecular cation, H2+. One-electron bonds often have about half the bond energy of a 2-electron bond, and are therefore called "half bonds". However, there are exceptions: in the case of dilithium, the bond is actually stronger for the 1-electron Li2+ than for the 2-electron Li2. This exception can be explained in terms of hybridization and inner-shell effects.

            The simplest example of three-electron bonding can be found in the helium dimer cation, He2+, and can also be considered a "half bond" because, in molecular orbital terms, the third electron is in an anti-bonding orbital which cancels out half of the bond formed by the other two electrons. Another example of a molecule containing a 3-electron bond, in addition to two 2-electron bonds, is nitric oxide, NO. The oxygen molecule, O2 can also be regarded as having two 3-electron bonds and one 2-electron bond, which acounts for its paramagnetism and its formal bond order of 2.

            Molecules with odd-electron bonds are usually highly reactive. These types of bond are only stable between atoms with similar electronegativities.

            3.2.7. Aromatic bond

            The first known use of the word "aromatic" as a chemical term -- namely, to apply to compounds that contain the phenyl radical -- occurs in an article by August Wilhelm Hofmann in 1855. If this is indeed the earliest introduction of the term, it is curious that Hofmann says nothing about why he introduced an adjective indicating olfactory character to apply to a group of chemical substances, only some of which have notable aromas. It is the case, however, that many of the most odoriferous organic substances known are terpenes, which are not aromatic in the chemical sense. But terpenes and benzenoid substances do have a chemical characteristic in common, namely higher unsaturation indexes than many aliphatic compounds, and Hofmann may not have been making a distinction between the two categories.

            The cyclohexatriene structure for benzene was first proposed by August Kekulé in 1865. Over the next few decades, most chemists readily accepted this structure, since it accounted for most of the known isomeric relationships of aromatic chemistry. However, it was always puzzling that this purportedly highly-unsaturated molecule was so unreactive toward addition reactions.

            The discoverer of the electron J. J. Thomson, in 1921 placed three equivalent electrons between each carbon atom in benzene.

            In most cases, the locations of electrons cannot be simplified to simple lines (place for two electrons) or dots (a single electron). In aromatic bonds which occur in rings of atoms where the 4n+2 rule determines whether ring molecules comprised of C=C bonds would show behavior extra stability by allowing extra sharing of electrons below and above the ring plane.

            In benzene, the prototypical aromatic compound, 18 bonding electrons bind 6 carbon atoms together to form a planar ring structure. The bond "order" (average number of bonds) between the different carbon atoms may be said to be (18/6)/2=1.5, but in this case the bonds are all identical from the chemical point of view. They may sometimes be written as single bonds alternating with double bonds, but the view of all ring bonds as being equivalently about 1.5 bonds in strength, is much closer to truth.

            In the case of heterocyclic aromatics and substituted benzenes, the electronegativity differences between different parts of the ring may dominate the chemical behaviour of aromatic ring bonds, which otherwise are equivalent.

Fig.15. Distribution of bonds in the molecule  benzene

           

            Aromaticity is a chemical property in which a conjugated ring of unsaturated bonds, lone pairs, or empty orbitals exhibit a stabilization stronger than would be expected by the stabilization of conjugation alone. It can also be considered a manifestation of cyclic delocalization and of resonance. 

                          

Fig.16. Kekulé resonance interconvertions of benzene

           

            This is usually considered to be because electrons are free to cycle around circular arrangements of atoms, which are alternately single- and double-bonded to one another. These bonds may be seen as a hybrid of a single bond and a double bond, each bond in the ring identical to every other. This commonly-seen model of aromatic rings, namely the idea that benzene was formed from a six-membered carbon ring with alternating single and double bonds (cyclohexatriene), was developed by Kekulé. The model for benzene consists of two resonance forms, which corresponds to the double and single bonds' switching positions. Benzene is a more stable molecule than would be expected without accounting for charge delocalization.

            By convention,  the  double-headed   arrow   indicates   that   the  two  structures  are  simply

Fig.17. Aromatic bond in the molecule benzene

hypothetical, since neither is an accurate representation of the actual compound. The actual molecule is best represented by a hybrid (average) of these structures, which can be seen at right. A C=C bond is shorter than a C−C bond, but benzene is perfectly hexagonal--all six carbon-carbon bonds have the same length, intermediate between that of a single and that of a double bond.A better representation is that of the circular π bond (Armstrong's inner cycle), in which the electron density is evenly distributed through a π bond above and below the ring. This model more correctly represents the location of electron density within the aromatic ring.

            The single bonds are formed with electrons in line between the carbon nuclei--these are called sigma bonds. Double bonds consist of a sigma bond and another bond - a π bond. The π-bonds are formed from overlap of atomic p-orbitals above and below the plane of the ring. The following diagram shows the positions of these p-orbitals:

Fig.18. Orientation of  p-orbitals in acyclic and cyclic carbon compounds

 

            Since they are out of the plane of the atoms, these orbitals can interact with each other freely, and become delocalised. This means that instead of being tied to one atom of carbon, each electron is shared by all six in the ring. Thus, there are not enough electrons to form double bonds on all the carbon atoms, but the "extra" electrons strengthen all of the bonds on the ring equally. The resulting molecular orbital has π symmetry.

 

Fig.19. Formation of π-electron system from p-AO

 

            Aromatic (Aryl) compounds possess following characteristics. An aromatic compound contains a set of covalently-bound atoms with specific characteristics:

            Whereas benzene is aromatic (6 electrons, from 3 double bonds), cyclobutadiene is not, since the number of π delocalized electrons is 4, which of course is a multiple of 4. The cyclobutadienide (2−) ion, however, is aromatic (6 electrons). An atom in an aromatic system can have other electrons that are not part of the system, and are therefore ignored for the 4n + 2 rule. In furan, the oxygen atom is sp² hybridized. One lone pair is in the π system and the other in the plane of the ring (analogous to C-H bond on the other positions). There are 6 π electrons, so furan is aromatic.

            Aromatic molecules typically display enhanced chemical stability, compared to similar non-aromatic molecules. The circulating π electrons in an aromatic molecule produce ring currents that oppose the applied magnetic field in NMR. The NMR signal of protons in the plane of an aromatic ring are shifted substantially further down-field than those on non-aromatic sp² carbons. This is an important way of detecting aromaticity. By the same mechanism, the signals of protons located near the ring axis are shifted up-field. Planar monocyclic molecules containing 4n π electrons are called antiaromatic and are, in general, destabilized. Molecules that could be antiaromatic will tend to alter their electronic or conformational structure to avoid this situation, thereby becoming non-aromatic. For example, cyclooctatetraene (COT) distorts itself out of planarity, breaking π overlap between adjacent double bonds. Aromatic molecules are able to interact with each other in so-called π-π stacking: the π systems form two parallel rings overlap in a "face-to-face" orientation. Aromatic molecules are also able to interact with each other in an "edge-to-face" orientation: the slight positive charge of the substituents on the ring atoms of one molecule are attracted to the slight negative charge of the aromatic system on another molecule.

            Many of the earliest-known examples of aromatic compounds, such as benzene and toluene, have distinctive pleasant smells. This property led to the term "aromatic" for this class of compounds, and hence to "aromaticity" being the eventually-discovered electronic property of them.

            Metal aromaticity is believed to exist in certain metal clusters of aluminium. Möbius aromaticity occurs when a cyclic system of molecular orbitals formed from pπ atomic orbitals and populated in a closed shell by 4n (n is an integer) electrons is given a single half-twist to correspond to a Möbius topology. Because the twist can be left-handed or right-handed, the resulting Möbius aromatics are dissymmetric or chiral. Up to now there is no doubtless proof, that a Möbius aromatic molecule was synthesized. Aromatics with two half-twists corresponding to the paradromic topologies first suggested by Johann Listing have been proposed by Rzepa in 2005. In carbo-benzene the ring bonds are extended with alkyne and allene groups.

            3.2.9. Metallic bond

            In a metallic bond, bonding electrons are delocalized over a lattice of atoms. By contrast, in ionic compounds, the locations of the binding electrons and their charges are static. Metallic bonds are found in metals like copper:

.

Fig. 20. Fragment of  the copper piece

 

            Metallic bonding is the bonding between atoms within metals. It involves the delocalized sharing of free electrons among a lattice of metal atoms. Thus, metallic bonds may be compared to molten salts.

            Metallic bonding is the electrostatic attraction between the metal atoms or ions and the delocalized electrons, also called conduction electrons. This is why atoms or layers are allowed to slide past each other, resulting in the characteristic properties of malleability and ductility.

            Metal atoms typically contain fewer electrons in their valence shell relative to their period or energy level. These electrons can be easily lost by the atoms and therefore become delocalized and form a sea of electrons surrounding a giant lattice of positive ions.

The electrons and the positive ions in the metal have a strong attractive force between them. This means that more energy is required to negate these forces. Therefore metals often have high melting or boiling points. The principle is similar to that of ionic bonds.

            Metallic bonding is non-polar, because for pure elemental metals and even for alloys there is no (or a very small) electronegativity difference among the atoms participating in the bonding interaction, and the electrons involved in that interaction are delocalized across the crystalline structure of the metal.

            The metallic bond accounts for many physical characteristics of metals, such as strength, malleability, ductility, conduction of heat and electricity, and lustre.

Due to the fact that the electrons move independently of the positive ions in a sea of negative charge, the metal gains some electrical conductivity. It allows the energy to pass quickly through the electrons generating a current. Heat conduction works on the same principle - the free electrons can transfer the energy at a faster rate than other substances such as those which are covalently bonded, as these have their electrons fixed into position. There also are few non-metals which conduct electricity: graphite (because, like metals, they have free electrons), and molten and aqueous ionic compounds which have free moving ions.

            Metal atoms have at least one valence electron which they do not share with neighboring atoms, nor do they lose electrons to form ions. Instead the outer energy levels of the

            3.2.10. Intermolecular bonding

            In physics, chemistry, and biology, intermolecular forces are forces that act between stable molecules or between functional groups of macromolecules. These non-covalent forces, which give rise to bonding energies of less than a few kcal/mol, are generally much weaker than the chemical bonding forces. Nevertheless, intermolecular forces are responsible for a wide range of physical, chemical, and biological phenomena. For instance, they play a role in the deviation from the ideal gas law for real gases, the tertiary structure of macromolecules and signal induction in neurotransmitters.

            There are four basic types of bonds that can be formed between two or more (otherwise non-associated) molecules, ions or atoms. We will distinguish four fundamental interactions:

  • exchange
  • electrostatic
  • induction
  • dispersion.

 Intermolecular forces cause molecules to be attracted or repulsed by each other. Often, these define some of the physical characteristics (such as the melting point) of a substance.

            In general one distinguishes short and long range intermolecular forces. The former are due to intermolecular exchange and charge penetration. They fall off exponentially as a function of intermolecular distance R and are repulsive for interacting closed-shell systems. In chemistry they are well known, because they give rise to steric hindrance, also known as Born or Pauli repulsion. Long range forces fall off with inverse powers of the distance, R-n, typically 3 ≤ n ≤ 10, and are mostly attractive.

            The sum of long and short range forces gives rise to a minimum, referred to as Van der Waals minimum. The position and depth of the Van der Waals minimum depends on distance and mutual orientation of the molecules.

            3.2.11. Permanent dipole to permanent dipole

            A large electronegativity difference (see 3.2.3) between two strongly bonded atoms within a molecule causes a dipole to form (a dipole is a pair of permanent partial charges). Dipoles will attract or repel each other. Dipole-dipole interactions, also called Keesom interactions or Keesom forces after Willem Hendrik Keesom, who produced the first mathematical description in 1921, are the forces that occur between two molecules with permanent dipoles. They result from the dipole-dipole interaction between two molecules. An example of this can be seen in hydrochloric acid:

Hd+-Cld-…Hd+-Cld-

            The molecules are depicted here as two point dipoles. A point dipole is an idealization similar to a point charge (a finite charge in an infinitely small volume). A point dipole consists of two equal charges of opposite sign δ+ and δ-, which are a distance d apart. This distance d is so small that at any distance R from the point dipole it can be assumed that d/R >> (d/R)2. In this idealization the electrostatic field outside the charge distribution consists of one (R-3) term only, see this article. The electrostatic interaction between two point dipoles is given by the single term lA = 1 and lB = 1 in the expansion above.

            Obviously, no molecule is an ideal point dipole, and in the case of the HCl dimer, for instance, dipole-quadrupole, quadrupole-quadrupole, etc. interactions are by no means negligible (and neither are induction or dispersion interactions).

            Note that almost always the dipole-dipole interaction between two atoms is zero, because atoms rarely carry a permanent dipole, see atomic dipoles.

To get the mathematical equation for the dipole-dipole interaction we must consider the term with lA = 1 and lB = 1 in the expansion of the electrostatic energy. Because this expansion is termwise rotational invariant, we can choose a convenient system of axes to evaluate the term. We choose a coordinate system centered on A with its z-axis coinciding with the intermolecular vector RAB. Under this circumstance it holds for the irregular solid harmonic that

                                              (46)

Hence, the dipole-dipole term becomes after substitution of two Clebsch-Gordan coefficients

        (47)

where

            (48)

Analogous relations hold for the permanent dipole moments on B. Then

                         (49)

Writing

                    (50)

and similarly for B, we get finally the well-known expression

                  (51)

            As a numerical example we consider the HCl dimer depicted above. We assume that the left molecule is A and the right B, so that the z-axis is along the molecules and points to the right. Our (physical) convention of the dipole moment is such that it points from negative to positive charge. Note parenthetically that in organic chemistry the opposite convention is used. Since organic chemists hardly ever perform vector computations with dipoles, confusion hardly ever arises. In organic chemistry dipoles are mainly used as a measure of charge separation in a molecule. So,

                            (52)

            The value of μHCl is 0.43 (atomic units), so that at a distance of 10 bohr the dipole-dipole attraction is -3.698 10-4 hartree (-0.97 kJ/mol).

            If one of the molecules is neutral and freely rotating, the total electrostatic interaction energy becomes zero. (For the dipole-dipole interaction this is most easily proved by integrating over the spherical polar angles of the dipole vector, while using the volume element sinθ dθdφ). In gases and liquids molecules are not rotating completely freely—the rotation is weighted by the Boltzmann factor exp(-Edip-dip/kT), where k is the Boltzmann constant and T the absolute temperature. It was first shown by Lennard-Jones[4] that the temperature-averaged dipole-dipole interaction is

                                                  (53)

Since the averaged energy has an R-6 dependence, it is evidently much weaker than the unaveraged one, but it is not completely zero. It is attractive, because the Boltzmann weighting favors somewhat the attractive regions of space. In HCl-HCl we find for T = 300 K and RAB = 10 bohr the averaged attraction -62 J/mol, which shows a weakening of the interaction by a factor of about 16 due to thermal rotational motion.

            3.2.12. London dispersion forces. Instantaneous dipole to induced dipole

            Instantaneous dipole to induced dipole, or van der Waals forces, are the weakest, but also the most prolific—occurring between all chemical substances.

            Imagine a helium atom. At any one point in time, the electron cloud around the (otherwise neutral) atom can be thought to be slightly imbalanced, with momentarily more negative charge on one side. This is referred to as an instantaneous dipole. This dipole, with its slight charge imbalance, may attract or repel the electrons within a neighbouring helium atom, setting up another dipole. The two atoms will be attracted for an instant, before the charge rebalances and the atoms move on.

            Also called London forces, instantaneous dipole (or multipole) effects (spatially variable δ+) or Van der Waals forces, these involve the attraction between temporarily induced dipoles in nonpolar molecules (often disappear within an instant). This polarization can be induced either by a polar molecule or by the repulsion of negatively charged electron clouds in nonpolar molecules. An example of the former is chlorine dissolving in water:

    (+)(-)(+)     (-) (+)

[Permanent Dipole] H-O-H-----Cl-Cl [Induced Dipole]

            Sketched is an interaction between the permanent dipole on water and an induced dipole on chlorine. The latter dipole is induced by the electric field offered by the permanent dipole of water (see field from an electric dipole).

            This permanent dipole-induced dipole interaction is referred to as induction (or polarization) interaction and is to be distinguished from the London dispersion interaction. The latter is sometimes described as an interaction between two instantaneous dipoles, see molecular dipole. The Cl2—Cl2 interaction that now follows is an example of a proper London dispersion interaction.

(+) (-)    (+) (-)

[instantaneous dipole] Cl-Cl------Cl-Cl [instantaneous dipole]

            Note added by other author:

            It must be pointed out that the London interaction is not the only interaction between two chlorine molecules in the region where the overlap between the respective charge distributions may be neglected. Each chlorine molecule carries permanent multipole moments of even order, the first one being a permanent quadrupole moment (order 2). The interaction between two permanent multipole moments also contributes to the intermolecular force and the first term (quadrupole-quadrupole) is as important as the London dispersion force. London dispersion forces exist between all atoms. London forces are the only reason for rare-gas atoms to condense at low temperature.

            The first explanation of the attraction between noble gas atoms was given by Fritz London in 1930. He used a quantum mechanical theory based on second-order perturbation theory. The perturbation is the Coulomb interaction V between the electrons and nuclei of the two monomers (atoms or molecules) that constitute the dimer. The second-order perturbation expression of the interaction energy contains a sum over states. The states appearing in this sum are simple products of the excited electronic states of the monomers. Thus, no intermolecular antisymmetrization of the electronic states is included and the Pauli exclusion principle is only partially satisfied.

            London developed the perturbation V in a Taylor series in , where R is the distance between the nuclear centers of mass of the monomers. This Taylor expansion is known as the multipole expansion of V because the terms in this series can be regarded as energies of two interacting multipoles, one on each monomer. Substitution of the multipole-expanded form of V into the second-order energy yields an expression that resembles somewhat an expression describing the interaction between instantaneous multipoles (see the qualitative description above). Additionally an approximation, named after Albrecht Unsöld, must be introduced in order to obtain a description of London dispersion in terms of dipole polarizabilities and ionization potentials.

            In this manner the following approximation is obtained for the dispersion interaction  between two atoms A and B. Here αA and αB are the dipole polarizabilities of the respective atoms. The quantities IA and IB are the first ionization potentials of the atoms and R is the intermolecular distance.

                                                    (54)

Note that this final London equation does not contain instantaneous dipoles. The "explanation" of the dispersion force as the interaction between two such dipoles was invented after London gave the proper quantum mechanical theory. The London theory has much similarity to the quantum mechanical theory of light dispersion, which is why London coined the phrase "dispersion effect" for the interaction that we described in this lemma.

            3.13. Hydrogen bond

            Hydrogen bonding is an intermolecular interaction with a hydrogen atom being present in the intermolecular bond. A hydrogen atom attached to a relatively electronegative atom is a hydrogen bond donor. This electronegative atom is usually fluorine, oxygen, or nitrogen. An electronegative atom such as fluorine, oxygen, or nitrogen is a hydrogen bond acceptor, regardless of whether it is bonded to a hydrogen atom or not. An example of a hydrogen bond donor is ethanol, which has a hydrogen bonded to oxygen; an example of a hydrogen bond acceptor which does not have a hydrogen atom bonded to it is the oxygen atom on diethyl ether.

            Carbon can also participate in hydrogen bonding, especially when the carbon atom is bound to several electronegative atoms, as is the case in chloroform, CHCl3. The electronegative atom attracts the electron cloud from around the hydrogen nucleus and, by decentralizing the cloud, leaves the atom with a positive partial charge. Because of the small size of hydrogen relative to other atoms and molecules, the resulting charge, though only partial, nevertheless represents a large charge density. A hydrogen bond results when this strong positive charge density attracts a lone pair of electrons on another heteroatom, which becomes the hydrogen-bond acceptor.

            The most ubiquitous, and perhaps simplest, example of a hydrogen bond is found between water molecules. In a discrete water molecule, water has two hydrogen atoms and one oxygen atom. This hydrogen is covalently (chemically) bound in one molecule, which acts as the proton donor. The other molecule acts as the proton acceptor. In the following important example of the water dimer, the water molecule on the right is the proton donor, while the one on the left is the proton acceptor:

Fig. 21. Formation of intermolecular H-bond between two water molecules

 

            Two molecules of water can form a hydrogen bond between them; the simplest case, when only two molecules are present, is called the water dimer and is often used as a model system.

            When more molecules are present, as is the case in liquid water, more bonds are possible because the oxygen of one water molecule has two lone pairs of electrons, each of which can form a hydrogen bond with hydrogens on two other water molecules. This can repeat so that every water molecule is H-bonded with up to four other molecules, as shown in the figure (two through its two lone pairs, and two through its two hydrogen atoms.)

            Liquid water's high boiling point is due to the high number of hydrogen bonds each molecule can have relative to its low molecular mass, not to mention the great strength of these hydrogen bonds. Realistically the water molecule has a very high boiling point, melting point and viscosity compared to other similar substances not conjoined by hydrogen bonds. The reasoning for these attributes is the inability to, or the difficulty in, breaking these bonds. Water is unique because its oxygen atom has two lone pairs and two hydrogen atoms, meaning that the total number of bonds of a water molecule is up to four.

            For example, hydrogen fluoride—which has three lone pairs on the F atom but only one H atom—can have a total of only two bonds (ammonia has the opposite problem: three hydrogen atoms but only one lone pair).

H-F...H-F...H-F.

            The exact number of hydrogen bonds in which a molecule in liquid water participates fluctuates with time and depends on the temperature. From TIP4P liquid water simulations at 25 °C, it was estimated that each water molecule participates in an average  of 3.59  hydrogen

           

Fig. 22. System of H-bonds in liquid water

 

bonds. At 100 °C, this number decreases to 3.24 due to the increased molecular motion and decreased density, while at 0 °C, the average number of hydrogen bonds increases to 3.69. A more recent study found a much smaller number of hydrogen bonds: 2.357 at 25 °C. The differences may be due to the use of a different method for defining and counting the hydrogen bonds.

            Were the bond strengths more equivalent, one might instead find the atoms of two interacting water molecules partitioned into two polyatomic ions of opposite charge, specifically hydroxide (OH−) and hydronium (H3O+) (Hydronium ions are also known as 'hydroxonium' ions.)

H-O− H3O+

Indeed, in pure water under conditions of standard temperature and pressure, this latter formulation is applicable only rarely; on average about one in every 5.5 × 108 molecules gives up a proton to another water molecule, in accordance with the value of the dissociation constant for water under such conditions. It is a crucial part of the uniqueness of water.

            The hydrogen atom participating in the hydrogen bond is often covalently bound in the donor to an electronegative atom. Examples of such atoms are nitrogen, oxygen, or fluorine. The electronegative atom is negatively charged (carries a charge δ-) and the hydrogen atom bound to it is positively charged. Consequently the proton donor is a polar molecule with a relatively large dipole moment. Often the positively charged hydrogen atom points towards an electron rich region in the acceptor molecule. The fact that an electron rich region exists in the acceptor molecule, implies already that the acceptor has a relatively large dipole moment as well. The result is a dimer that to a large extent is bound by the dipole-dipole force.

            The hydrogen bond is often described as an electrostatic dipole-dipole interaction. However, it also has some features of covalent bonding: it is directional, strong, produces interatomic distances shorter than sum of van der Waals radii, and usually involves a limited number of interaction partners, which can be interpreted as a kind of valence. These covalent features are more significant when acceptors bind hydrogens from more electronegative donors.

            For quite some time it was believed that hydrogen bonding required an explanation that was different from the other intermolecular interactions. However, reliable computer calculations that became possible since the 1980s, have shown that only the four effects listed above play a role, with the dipole-dipole interaction being particularly important. Since the four effects account completely for the bonding in small dimers like the water dimer, for which highly accurate calculations are feasible, it is now generally believed that no other bonding effects are operative.

            An example of a quadruple hydrogen bond between a self-assembled dimer complex reported by Meijer and coworkers.

An example of a quadruple hydrogen bond between a self-assembled dimer complex reported by Meijer and coworkers.

Fig. 23. Intramolecular H-bonds betwee two molecules of acetylacetone

Intramolecular hydrogen bonding in acetylacetone helps stabilize the enol tautomer

Fig. 24. Tautomerization in acetylacetone

            Carboxylic acids often form dimers in vapor phase

 

Fig. 25. Formation of dimeric molecule by H-bonds    

            A hydrogen bond is a special type of dipole-dipole bond that exists between an electronegative atom and a hydrogen atom bonded to another electronegative atom. This type of bond always involves a hydrogen atom, thus the name. Hydrogen bonds can occur between molecules (intermolecularly), or within different parts of a single molecule (intramolecularly).[2] The typical hydrogen bond is stronger than van der Waals forces, but weaker than covalent, ionic and metallic bonds.

            Intermolecular hydrogen bonding is responsible for the high boiling point of water (100 °C). This is because of the strong hydrogen bond, as opposed to other group 16 hydrides. Intramolecular hydrogen bonding is partly responsible for the secondary, tertiary, and quaternary structures of proteins and nucleic acids.

We distinguish following kinds of H-bonds:

  • Symmetric hydrogen bond
  • Dihydrogen bond
  • Bifurcated and over-coordinated hydrogen bonds

            A symmetric hydrogen bond is a special type of hydrogen bond in which the proton is spaced exactly halfway between two identical atoms. The strength of the bond to each of those atoms is equal. It is an example of a 3-center 4-electron bond. This type of bond is much stronger than "normal" hydrogen bonds, in fact, its strength is comparable to a covalent bond. It is seen in ice at high pressure, and also in the solid phase of many anhydrous acids such as hydrofluoric acid and formic acid at high pressure. It is also seen in the ion [F-H-F]−. Much has been done to explain the symmetric hydrogen bond quantum-mechanically, as it seems to violate the duet rule for the first shell: The proton is effectively surrounded by four electrons. Because of this problem, some consider it to be an ionic bond.

            Symmetric hydrogen bonds have been observed recently spectroscopically in formic acid at high pressure (>GPa). Each hydrogen atom forms a partial covalent bond with two atoms rather than one. Symmetric hydrogen bonds have been postulated in ice at high pressure (ice-X). Low-barrier hydrogen bonds form when the distance between two heteroatoms is very small.

            The partially covalent nature of a hydrogen bond raises the questions: "To which molecule or atom does the hydrogen nucleus belong?" and "Which should be labeled 'donor' and which 'acceptor'?" Usually, this is easy to determine simply based on interatomic distances in the X-H...Y system: X-H distance is typically ~1.1 Å, whereas H...Y distance is ~ 1.6 to 2.0 Å. Liquids that display hydrogen bonding are called associated liquids.

            Hydrogen bonds can vary in strength from very weak (1-2 kJ mol−1) to extremely strong (>155 kJ mol−1), as in the ion HF2-.

            Typical values include:

F—H ... :F (155 kJ/mol or 40 kcal/mol)

O—H ... :N (29 kJ/mol or 6.9 kcal/mol)

O—H ... :O (21 kJ/mol or 5.0 kcal/mol)

N—H ... :N (13 kJ/mol or 3.1 kcal/mol)

N—H ... :O (8 kJ/mol or 1.9 kcal/mol)

      HO—H ... :OH3+ (18 kJ/mol or 4.3 kcal/mol)

{Data obtained using molecular dynamics as detailed in the reference and should be compared to 7.9 kJ/mol, obtained using the same molecular dynamics}.

            The length of hydrogen bonds depends on bond strength, temperature, and pressure. The bond strength itself is dependent on temperature, pressure, bond angle, and environment (usually characterized by local dielectric constant). The typical length of a hydrogen bond in water is 1.97 Å (197 pm).

            Hydrogen bonds are found throughout nature. They give water its unique properties that are so important to life on earth. Hydrogen bonds between hydrogen atoms and nitrogen atoms of adjacent base pairs provide the intermolecular force that help more precisely bind together the two strands in a molecule of DNA. Hydrophobic effects between the double-stranded DNA and the surrounding aqueous environment, however, are more important in maintaining the DNA in its double stranded form.

            In some ways this is an especially strong example of a permanent dipole, as above. However, in the hydrogen bond, the hydrogen proton comes closer to being shared between target and donor atoms, in a three-center two-electron bond like that in diborane. Hydrogen bonds explain the relatively high boiling points of liquids like water, ammonia, and hydrogen fluoride, compared with their heavier counterparts in the same periodic table column.

            Dihydrogen bond take place in cases when the hydrogen bond can be compared with the closely related dihydrogen bond, which is also an intermolecular bonding interaction involving hydrogen atoms. These structures have been known for some time, and well characterized by crystallography; however, an understanding of their relationship to the conventional hydrogen bond, ionic bond, and covalent bond remains unclear. Generally, the hydrogen bond is characterized by a proton acceptor that is a lone pair of electrons in nonmetallic atoms (most notably in the nitrogen, and chalcogen groups). In some cases, these proton acceptors may be pi-bonds or metal complexes. In the dihydrogen bond, however, a metal hydride serves as a proton acceptor; thus forming a hydrogen-hydrogen interaction. Neutron diffraction has shown that the molecular geometry of these complexes are similar to hydrogen bonds, in that the bond length is very adaptable to the metal complex/hydrogen donor system.

            Bifurcated and over-coordinated hydrogen bonds take place in water. It can be sometimes, that single hydrogen participates in two hydrogen bonds, rather than one. This type of bonding is called "bifurcated". It was suggested that a bifurcated hydrogen is an essential step in water reorientation,] however - the case of an oxygen lone pair participating in more then two hydrogens bonds is rarly given attention in the scientific literature.

            Advanced theory of the hydrogen bond has appeared in 1980s. The hydrogen bond was long a fairly mysterious object in the theoretical study of quantum chemistry and physics. However, in the early 1980s it was shown to be primarily of electrostatic origin, i.e., the hydrogen bond attraction is mainly due to the interaction of permanent dipoles and higher permanent multipoles of the molecules participating in the bond. In 1983 Buckingham and Fowler modelled the interaction of a large number of hydrogen bonded van der Waals molecules by atomic hard sphere repulsions and distributed multipole interactions. They found a qualitative agreement between predicted and measured structures, which showed that their model accounted for the major effects of the hydrogen bonding. These results have been borne out by many quantum chemical calculations since then, of the perturbative as well as of the supermolecule type. However, until recently some controversy existed about the nature of the bond. A widely publicized article claimed from interpretations of the anisotropies in the Compton profile of ordinary ice, that the hydrogen bond is partly covalent. Some NMR data on hydrogen bonds in proteins also seemed to indicate covalent bonding. However, these interpretations were refuted by Ernest R. Davidson and coworkers. It is now commonly assumed that for hydrogen bonding the same effects (exchange, electrostatic, polarization, and dispersion) play a role as for "ordinary" intermolecular forces, with electrostatics plus Pauli (hard sphere) repulsion being dominant for hydrogen bonds.

            Hydrogen bonding causes following phenomena:

1.      Dramatically higher boiling points of NH3, H2O, and HF compared to the heavier analogues PH3, H2S, and HCl.

2.      Viscosity of anhydrous phosphoric acid and of glycerol.

3.      Dimer formation in carboxylic acids and hexamer formation in hydrogen fluoride, which occur even in the gas phase, resulting in gross deviations from the ideal gas law.

4.      High water solubility of many compounds such as ammonia is explained by hydrogen bonding with water molecules.

5.      Negative azeotropy of mixtures of HF and water.

6.      Deliquescence of NaOH is caused in part by reaction of OH- with moisture to form hydrogen-bonded H2O3- species. An analogous process happens between NaNH2 and NH3, and between NaF and HF.

            The fact that Ice is less dense than liquid water is due to a crystal structure resulting from hydrogen bonds.

            3.2.14. Cation-π-interaction

            Cation-pi interactions occur between the localized negative charge of π orbital electrons, located above and below the plane of an aromatic ring, and a positive charge. A cation-pi interac­tion is a noncovalent molecular interaction between the electron-rich π orbitals of an aromatic ring with adjacent cation. Such interactions are relatively strong, being roughly equivalent in energy to a hydrogen bond. This interaction provides stability in thermophilic proteins.

            3.2.15. Electrons in chemical bonds

            Many simple compounds involve covalent bonds. These molecules have structures that can be predicted using valence bond theory, and the properties of atoms involved can be understood using concepts such as oxidation number. Other compounds that involve ionic structures can be understood using theories from classical physics.

            In the case of ionic bonding, electrons are mainly localized on the individual atoms, and electrons do not travel between the atoms very much. Each atom is assigned an overall electric charge to help conceptualize the molecular orbital's distribution. The forces between atoms (or ions) are largely characterized by isotropic continuum electrostatic potentials. By contrast, in covalent bonding, the electron density within a bond is not assigned to individual atoms, but is instead delocalized in the MOs between atoms. The widely accepted theory of the linear combination of atomic orbitals (LCAO) helps describe the molecular orbital structures and energies based on the atomic orbitals of the atoms they came from. Unlike pure ionic bonds, covalent bonds may have directed anisotropic properties. These may have their own names, too, such as σ-(Sigma) and π(Pi)-bond. Atoms can also form bonds that are intermediates between ionic and covalent. This is because these definitions are based on the extent of electron delocalization. Electrons can be partially delocalized between atoms, but spend more time around one atom than another. This type of bond is often called polar covalent.

            Thus, the electrons in a molecular orbital (or in a polar covalent, or in a covalent bond) can be said to be either localized on certain atom(s) or delocalized between two or more atoms. The type of bond between two atoms is defined by how much the electron density is localized or delocalized among the atoms of the bonds.

            Chapter 4. Theory of Chemical Bonds 

            4.1. Valence bond theory

            In 1927, valence bond theory was formulated which argued essentially that a chemical bond forms when two valence electrons, in their respective atomic orbitals, work or function to hold two nuclei together, by virtue of system energy lowering effects. In 1931, building on this theory, chemist Linus Pauling published what some consider one of the most important papers in the history of chemistry: “On the Nature of the Chemical Bond”. In this paper, building on the works of Lewis, and the valence bond theory (VB) of Heitler and London, and his own earlier work, he presented six rules for the shared electron bond, the first three of which were already generally known:

1. The electron-pair bond forms through the interaction of an unpaired electron on each of two atoms.

2. The spins of the electrons have to be opposed.

3. Once paired, the two electrons cannot take part in additional bonds.

His last three rules were new:

4. The electron-exchange terms for the bond involves only one wave function from each atom.

5. The available electrons in the lowest energy level form the strongest bonds.

6. Of two orbitals in an atom, the one that can overlap the most with an orbital from another atom will form the strongest bond, and this bond will tend to lie in the direction of the concentrated orbital.

            Building on this article, Pauling’s 1939 textbook: On the Nature of the Chemical Bond would become what some have called the “bible” of modern chemistry. This book helped experimental chemists to understand the impact of quantum theory on chemistry. However, the later edition in 1959 failed to address adequately the problems that appeared to be better understood by molecular orbital theory. The impact of valence theory declined during the 1960's and 1970's as molecular orbital theory grew in popularity and was implemented in many large computer programs. Since the 1980s, the more difficult problems of implementing valence bond theory into computer programs have been largely solved and valence bond theory has seen a resurgence.

            4.2. Molecular orbital theory

            In chemistry, molecular orbital theory (MO theory) is a method for determining molecular structure in which electrons are not assigned to individual bonds between atoms, but are treated as moving under the influence of the nuclei in the whole molecule.

Molecular orbital theory (MO) uses a linear combination of atomic orbitals to form molecular orbitals which cover the whole molecule. In this theory, each molecule has a set of molecular orbitals, in which it is assumed that the molecular orbital wave function ψf may be written as a simple weighted sum of the n constituent atomic orbitals χi, according to the following equation:

                                                          (55)

            The cij coefficients may be determined numerically by substitution of this equation into the Schrödinger equation and application of the variational principle. This method is called the linear combination of atomic orbitals approximation and is used in computational chemistry. An additional unitary transformation can be applied on the system to accelerate the convergence in some computational schemes. Molecular orbital theory was seen as a competitor to valence bond theory in the 1930s, before it was realised that the two methods are closely related and that when extended they become equivalent.

Molecular orbital theory was developed, in the years after valence bond theory (1927) had been established, primarily through the efforts of Friedrich Hund, Robert Mulliken, John C. Slater, and John Lennard-Jones. MO theory was originally called the Hund-Mulliken theory. The word orbital was introduced by Mulliken in 1932. By 1933, the molecular orbital theory had become accepted as a valid and useful theory. According to German physicist and physical chemist Erich Hückel, the first quantitative use of molecular orbital theory was the 1929 paper of Lennard-Jones. The first accurate calculation of a molecular orbital wavefunction was that made by Charles Coulson in 1938 on the hydrogen molecule. By 1950, molecular orbitals were completely defined as eigenfunctions (wave functions) of the self-consistent field Hamiltonian and it was at this point that molecular orbital theory became fully rigorous and consistent. This rigorous approach is known as the Hartree-Fock method for molecules although it had its origins in calculations on atoms. In calculations on molecules, the molecular orbitals are expanded in terms of an atomic orbital basis set, leading to the Roothaan equations. This led to the development of many ab initio quantum chemistry methods. Parallel to this rigorous development, molecular orbital theory was applied in an approximate manner using some empirically derived parameters in methods now known as semi-empirical quantum chemistry methods.

            These are often divided into bonding orbitals, anti-bonding orbitals, and non-bonding orbitals. A molecular orbital is merely a Schrödinger orbital which includes several, but often only two nuclei. If this orbital is of type in which the electron(s) in the orbital have a higher probability of being between nuclei than elsewhere, the orbital will be a bonding orbital, and will tend to hold the nuclei together. If the electrons tend to be present in a molecular orbital in which they spend more time elsewhere than between the nuclei, the orbital will function as an anti-bonding orbital and will actually weaken the bond. Electrons in non-bonding orbitals tend to be in deep orbitals (nearly atomic orbitals) associated almost entirely with one nucleus or the other, and thus they spend equal time between nuclei or not. These electrons neither contribute nor detract from bond strength.

            4.3. Born-Oppenheimer approximation

            In quantum chemistry, the computation of the energy and wavefunction of an average-size molecule is a formidable task that is alleviated by the Born-Oppenheimer (BO) approximation. For instance the benzene molecule consists of 12 nuclei and 42 electrons. The time-independent Schrödinger equation, which must be solved to obtain the energy and molecular wavefunction of this molecule, is a partial differential eigenvalue equation in 162 variables—the spatial coordinates of the electrons and the nuclei. The BO approximation makes it possible to compute the wavefunction in two less formidable, consecutive, steps. This approximation was proposed in the early days of quantum mechanics by Born and Oppenheimer (1927) and is still indispensable in quantum chemistry.

In the first step of the BO approximation the electronic Schrödinger equation is solved, yielding a wavefunction depending on electrons only. For benzene this wavefunction depends on 126 electronic coordinates. During this solution the nuclei are fixed in a certain configuration, very often the equilibrium configuration. If the effects of the quantum mechanical nuclear motion are to be studied, for instance because a vibrational spectrum is required, this electronic computation must be repeated for many different nuclear configurations. The set of electronic energies thus computed becomes a function of the nuclear coordinates. In the second step of the BO approximation this function serves as a potential in a Schrödinger equation containing only the nuclei—for benzene an equation in 36 variables.

The success of the BO approximation is due to the high ratio between nuclear and electronic masses. The approximation is an important tool of quantum chemistry, without it only the lightest molecule, H2, could be handled; all computations of molecular wavefunctions for larger molecules make use of it. Even in the cases where the BO approximation breaks down, it is used as a point of departure for the computations.

The electronic energies, constituting the nuclear potential, consist of kinetic energies, interelectronic repulsions and electron-nuclear attractions. In a handwaving manner it is sometimes suggested that the nuclear potential is some averaged electron-nuclear attraction. This interpretation of the Born-Oppenheimer approximation is wrong and misleading.

The Born-Oppenheimer (BO) approximation is ubiquitous in quantum chemical calculations of molecular wavefunctions. It consists of two steps.

In the first step the nuclear kinetic energy is neglected, that is, the corresponding operator Tn is subtracted from the total molecular Hamiltonian. In the remaining electronic Hamiltonian He the nuclear positions enter as parameters. The electron-nucleus interactions are not removed and the electrons still "feel" the Coulomb potential of the nuclei clamped at certain positions in space. (This first step of the BO approximation is therefore often referred to as the clamped nuclei approximation.)

            The electronic Schrödinger equation

Heχ(r) = Ee(r) χ(r)                                                         (56)

 

is solved (out of necessity approximately) with a fixed nuclear geometry as input. The quantity r stands for all electronic coordinates. Obviously, the electronic energy eigenvalue Ee depends on the chosen positions R of the nuclei. Varying these positions R in small steps and repeatedly solving the electronic Schrödinger equation, one obtains Ee as a function of R. This is the potential energy surface (PES): Ee(R). Because this procedure of recomputing the electronic wave functions as a function of an infinitesimally changing nuclear geometry is reminiscent of the conditions for the adiabatic theorem, this manner of obtaining a PES is often referred to as the adiabatic approximation and the PES itself is called an adiabatic surface.

In the second step of the BO approximation the nuclear kinetic energy Tn (containing partial derivatives with respect to the components of R) is reintroduced and the Schrödinger equation for the nuclear motion

[Tn + Ee(R)]f = Ef(r)                                                (57)

is solved. This second step of the BO approximation involves separation of vibrational, translational, and rotational motions. This can be achieved by application of the Eckart conditions. The eigenvalue E is the total energy of the molecule, including contributions from electrons, nuclear vibrations, and overall rotation and translation of the molecule.

Now we shall consider application of the theory to molecular system. It will be discussed how the BO approximation may be derived and under which conditions it is applicable. At the same time we will show how the BO approximation may be improved by including vibronic coupling. To that end the second step of the BO approximation is generalized to a set of coupled eigenvalue equations depending on nuclear coordinates only. Off-diagonal elements in these equations are shown to be nuclear kinetic energy terms. It will be shown that the BO approximation can be trusted whenever the PESs, obtained from the solution of the electronic Schrödinger equation, are well separated: E0(R)<<E1(R)<< E2(R), . . . for all n.

            We start from the exact non-relativistic, time-independent molecular Hamiltonian:

H=He+Tn                                                                                        (58)

with

          (59)

The position vectors r º {ri} of the electrons and the position vectors Rº{RA=(RAx,RAy,RAz)} of the nuclei are with respect to a Cartesian inertial frame. Distances between particles are written as riA º |ri-RA|  (distance between electron i and nucleus A) and similar definitions hold for and . We assume that the molecule is in a homogeneous (no external force) and isotropic (no external torque) space. The only interactions are the Coulomb interactions between the electrons and nuclei. The Hamiltonian is expressed in atomic units, so that we do not see Planck's constant, the dielectric constant of the vacuum, electronic charge, or electronic mass in this formula. The only constants explicitly entering the formula are ZA and MA—the atomic number and charge of nucleus A.

It is useful to introduce the total nuclear momentum and to rewrite the nuclear kinetic energy operator as follows:

                                  (60)

            Suppose we have K electronic eigenfunctions of , that is, we have solved

                 (61)

            The electronic wave functions ck will be taken to be real, which is possible when there are no magnetic or spin interactions. The parametric dependence of the functions ck on the nuclear coordinates ck is indicated by the symbol after the semicolon. This indicates that, although is a real-valued function of , its functional form depends on . For example, in the molecular-orbital-linear-combination-of-atomic-orbitals (LCAO-MO) approximation, ck is a MO given as a linear expansion of atomic orbitals (AOs). An AO depends visibly on the coordinates of an electron, but the nuclear coordinates are not explicit in the MO. However, upon change of geometry, i.e., change of , the LCAO coefficients obtain different values and we see corresponding changes in the functional form of the MO ck. We will assume that the parametric dependence is continuous and differentiable, so that it is meaningful to consider

                (62)

which in general will not be zero.

The total wave function Ψ is expanded in terms of ck(r; R):

                                               (63)

with

                                        (64)

and where the subscript (r)indicates that the integration, implied by the braket notation, is over electronic coordinates only. By definition, the matrix with general element

    (65)

is diagonal. After multiplication by the real function ck’ (r; R) from the left and integration over the electronic coordinates the total Schrödinger equation

                                             (66)

is turned into a set of K coupled eigenvalue equations depending on nuclear coordinates only

                               (67)

The column vector f(R) has elementsfk(R), k =1,…K. The matrix  is diagonal and the nuclear Hamilton matrix is non-diagonal with the following off-diagonal (vibronic coupling) terms,

                          (68)

            The vibronic coupling in this approach is through nuclear kinetic energy terms. Solution of these coupled equations gives an approximation for energy and wavefunction that goes beyond the Born-Oppenheimer approximation. Unfortunately, the off-diagonal kinetic energy terms are usually difficult to handle. This is why often a diabatic transformation is applied, which retains part of the nuclear kinetic energy terms on the diagonal, removes the kinetic energy terms from the off-diagonal and creates coupling terms between the adiabatic PESs on the off-diagonal.

            If we can neglect the off-diagonal elements the equations will uncouple and simplify drastically. In order to show when this neglect is justified, we suppress the coordinates in the notation and write, by applying the Leibniz rule for differentiation, the matrix elements of Tn as

   (69)

            The diagonal (k' = k) matrix elements áck|(PAαck)ñ(r) of the operator PAα vanish, because this operator is Hermitian and purely imaginary. The off-diagonal matrix elements satisfy

                                     (70)

The matrix element in the numerator is

        (71)

            The matrix element of the one-electron operator appearing on the right hand side is finite. When the two surfaces come close, Ek(R)» Ek’(R), the nuclear momentum coupling term becomes large and is no longer negligible. This is the case where the BO approximation breaks down and a coupled set of nuclear motion equations must be considered, instead of the one equation appearing in the second step of the BO approximation.

            Conversely, if all surfaces are well separated, all off-diagonal terms can be neglected and hence the whole matrix of is effectively zero. The third term on the right hand side of the expression for the matrix element of Tn (the Born-Oppenheimer diagonal correction) can approximately be written as the matrix of squared and, accordingly, is then negligible also. Only the first (diagonal) kinetic energy term in this equation survives in the case of well-separated surfaces and a diagonal, uncoupled, set of nuclear motion equations results,

               (72)

which are the normal second-step of the BO equations discussed above.

            We reiterate that when two or more potential energy surfaces approach each other, or even cross, the Born-Oppenheimer approximation breaks down and one must fall back on the coupled equations. Usually one invokes then the diabatic approximation.

            4.4. Variation principle

The variational principle is the basic method of calculation molecular orbitals. It is a principle in physics which is expressed in terms of the calculus of variations.   The Variational Principle provides the starting point for almost all methods which aim to find an approximate solution to Schrödinger's equation. It may be stated as follows:

The expectation value of a Hamiltonian, Ĥ, calculated using a trial wave-function, ΨT, is never lower in value than the true ground state energy, º0, which is the expectation value of Ĥ calculated using the true ground state wavefunction, Ψ0.

Say you have a system for which you know what the energy depends on, or in other words, you know the Hamiltonian H. If one cannot solve the Schrödinger equation to figure out the ground state wavefunction, you one may try any normalized wavefunction whatsoever, say φ, and the expectation value of the Hamiltonian for your trial wavefunction will be greater than the actual ground state energy. Or in other words:

                                                       (73)

            This holds for any trial φ, and is obvious from the definition of the ground state wavefunction of a system. By definition, the ground state has the lowest energy, and therefore any trial wavefunction will have an energy greater than or equal to the ground state energy.

Your guessed wavefunction, φ, can be expanded as a linear combination of the actual eigenfunctions of the Hamiltonian (which we assume to be normalized and orthogonal):

                                                          (74)

Then, to find the expectation value of the hamiltonian:

Now, the ground state energy is the lowest energy possible, i.e. En≥E0. Therefore, if the guessed wave function φ is normalized:

                                                (75)

For a hamiltonian H that describes the studied system and any normalizable function Ψ with arguments appropriate for the unknown wave function of the system, we define the functional

                                                           (76)

The variational principle states that

  • e ≥ E0, where E0 is the lowest energy eigenstate (ground state) of the hamiltonian
  • e = E0 if and only if Ψ is exactly equal to the wave function of the ground state of the studied system.

Variational calculations rely on making a physically plausible guess at the form of the ground state wavefunction, ΨT, of the Hamiltonian, Ĥ. This guess will be referred to as the trial/guiding wavefunction throughout this thesis. The “trial” part of the name refers to the use of the wavefunction as a guess of the true groundstate wavefunction to be used as the input wavefunction in a Variational quantum Monte Carlo (VMC) calculation. The “guiding” part refers to the use of the same wavefunction as an input wavefunction in the Diffusion quantum Monte Carlo (DMC) algorithm as part of the mechanism to introduce importance sampling. This will described in more detail in section. The trial/guiding wavefunction depends on a number of variable parameters which can be adjusted to minimise the energy expectation value. If the guessed values of these parameters are good and the chosen functional form builds in enough variational freedom to adequately describe the physics of the system being studied, then very accurate estimates of the ground state energy can be obtained. Variational quantum Monte Carlo (VMC) calculations are direct applications of the above variational principle.

            4.5. Slater determination

            In quantum mechanics, a Slater determinant is an expression which describes the wavefunction of a multi-fermionic system that satisfies the Pauli exclusion principle and subsequently anti-symmetry (Φ = 0) requirements by changing sign upon exchange of fermions. It is named for its discoverer, John C. Slater who published Slater determinants as a means of ensuring the antisymmetry of a wave function through the use of matrices. The Slater determinant arises from the consideration of a wave function for a collection of electrons, each with a wave function known as the spin-orbital, , where x denotes the position and spin of the singular electron; two electrons within the same spin orbital resulting in no wave function.

Two-particle system is simplest case of muny-particle system and therefofe Slater determi­nant have also simplest view. For simplicity we shall consider this variant.

This quantum approximate method to approximate the wavefunction of a many-particle system is to take the product of properly chosen one-electron wavefunctions of the individual particles. For the two-particle case, we have

     Y(x1, x2)=χ(x1)χ(x2).                                                    (77)

            This expression is used in the Hartree method as an ansatz for the molecular wavefunction and is known as a Hartree product. However, it is not satisfactory for fermions, such as electrons, because the wavefunction is not antisymmetric. An antisymmetric wavefunction can be mathematically described as follows:

     Y(x1, x2)= - Y(x2, x1)                                                 (78)

            Therefore the Hartree product does not satisfy the Pauli principle. This problem can be overcome by taking a linear combination of both Hartree products

       (79)

where the coefficient is a normalization factor. This wavefunction is antisymmetric and no longer distinguishes between electrons. Moreover, it also goes to zero if any two wavefunctions or two electrons are the same. This is equivalent to satisfying the Pauli exclusion principle.

Now we shall consider the general case of many particles. The expression can be generalised to any number of fermions by writing it as a determinant. For an N-electron system, the Slater determinant is defined as

                          (80)

The linear combination of Hartree products for the two-particle case can clearly be seen as identical with the Slater determinant for N = 2. It can be seen that the use of (Slater) determinants assures an antisymmetrized function on the outset, symmetric functions are automatically rejected. In the same way, the use of Slater determinants assures the obeying of the Pauli principle. Indeed, the Slater determinant vanishes if the set {χi } is linearly dependent. In particular this is the case when two (or more) spinorbitals are the same. In chemistry one expresses this fact by stating that no two electrons can occupy the same spinorbital. In general the Slater determinant is evaluated by the Laplace expansion. Mathematically, a Slater determinant is an antisymmetric tensor, also known as a wedge product.

A single Slater determinant is used as an approximation to the electronic wavefunction in Hartree-Fock theory. In more accurate theories (such as configuration interaction and MCSCF), a linear combination of Slater determinants is needed.

            4.6. Hartree-Fock approximation

            In computational physics and computational chemistry, the Hartree-Fock (HF) method is an approximate method for the determination of the ground-state wavefunction and ground-state energy of a quantum many-body system.

            The origin of the Hartree-Fock method dates back to the end of the 1920s, soon after the derivation of the Schrödinger equation in 1926. In 1927 D.R. Hartree introduced a procedure, which he called the self consistent field method, to calculate approximate wavefunctions and energies for atoms and ions. In 1930 Slater and V.A. Fock independently pointed out that the Hartree method did not respect the principle of antisymmetry of the wavefunction. Hartree's method used the Pauli exclusion principle in its older formulation, forbidding the presence of two electrons in the same quantum state. However, this was shown to be fundamentally incomplete in its neglect of quantum statistics. The original Hartree method can then be viewed as an approximation to the Hartree-Fock method by neglecting exchange. Fock's original method relied heavily on group theory and was too abstract for contemporary physicists to understand and implement. In 1935 Hartree reformulated the method more suitably for the purposes of calculation. The Hartree-Fock method assumes that the exact, N-body wavefunction of the system can be approximated by a single Slater determinant (in the case where the particles are fermions) or by a single permanent (in the case of bosons) of N spin-orbitals.

      Invoking the variational principle one can derive a set of N coupled equations for the N spin-orbitals. Solution of these equations yields the Hartree-Fock wavefunction and energy of the system, which are approximations of the exact ones. The Hartree-Fock method finds its typical application in the solution of the electronic Schrödinger equation of atoms, molecules and solids but it has also found widespread use in nuclear physics. See Hartree-Fock-Bogolyubov for a discussion of its application in nuclear structure theory. The rest of this article will focus on applications in electronic structure theory. The Hartree-Fock method is also called, especially in the older literature, the self-consistent field method (SCF). The solutions to the resulting non-linear equations behave as if each particle is subjected to the mean field created by all other particles (see the Fock operator below). The equations are almost universally solved by means of an iterative, fixed-point type algorithm (see the following section for more details). This solution scheme is not the only one possible and is not an essential feature of the Hartree-Fock method. For molecules, Hartree-Fock is the central method for all ab initio quantum chemistry methods. The discussion here is only for the Restricted Hartree-Fock method, where the atom or molecule is a closed-shell system with all orbitals (atomic or molecular) are doubly occupied. Open shell systems, where some of the electrons are not paired, can be dealt with by one of two Hartree-Fock methods:

            Restricted Open-shell Hartree-Fock (ROHF) is a variant of Hartree-Fock theory for open shell molecules. It uses doubly occupied molecular orbitals as far as possible and then singly occupied orbitals for the unpaired electrons. This is the simple picture for open shell molecules but it is difficult to implement. As with restricted Hartree-Fock theory for closed shell molecules, it leads to Roothaan equations written in the form of a generalized eigenvalue problem

FC=SCº                                                                    (81)

Where F is the so-called Fock matrix, C is a matrix of coefficients, S is the overlap matrix of the basis functions, and º is the (diagonal, by convention) matrix of orbital energies. Unlike restricted Hartree-Fock theory for closed shell molecules, the form of the Fock matrix is not unique. Different so-called canonicalisations can be used leading to different orbitals and different orbital energies, but the same total wavefunction, total energy, and other observables.

            In contrast to Unrestricted Hartree-Fock (UHF), the ROHF wave function is a satisfactory eigenfunction of the total spin operator –S2.Developing post Hartree-Fock methods from the ROHF wave function has proved to be more difficult than from the UHF wave function, but a restricted open shell version of Møller-Plesset perturbation theory to second order (ROMP2) has been developed.

Unrestricted Hartree-Fock (UHF) theory is the most common molecular orbital method for open shell molecules where the number of electrons of each spin are not equal. While restricted Hartree-Fock theory uses a single molecular orbital twice, once multiplied by the α spin function and once multiplied by the β spin function in the Slater determinant, unrestricted Hartree-Fock theory uses different molecular orbitals for the α and β electrons. This has been called a different orbitals for different spins (DODS) method. The result is a pair of coupled Roothaan equations, known as the Pople-Nesbet equations:

FαCα=SαCαº                                                                                                 (82)

FβCβ=SβCβº                                                              (82’)

            Where Fα and Fβ are the Fock matrices for the α and β orbitals, Cα and Cβ are the matrices of coefficients for the α and β orbitals, S is the overlap matrix of the basis functions, and ºα  and ºβ  are the (diagonal, by convention) matrices of orbital energies for the α and β orbitals. The pair of equations are coupled because the Fock matrix elements of one spin contains coefficients of both spin as the orbital has to be optimised in the average field of all other electrons. The final result is a set of molecular orbitals and orbital energies for the α spin electrons and a set of molecular orbitals and orbital energies for the β electrons.

This method has one drawback. A single Slater determinant of different orbitals for different spins is not a satisfactory eigenfunction of the total spin operator – S2. The ground state is contaminated by excited states. If there is one more electron of α spin than β spin, the ground state is a doublet. The average value of S2, written <S2>, should be  but will actually be rather more than this value as the doublet state is contaminated by a quadruplet state. A triplet state with two excess α electrons should have <S2> = 1 (1 + 1) = 2, but it will be larger as the triplet is contaminated by a quintruplet state. When carrying out unrestricted Hartree-Fock calculations, it is always necessary to check this contamination. For example, with a doublet state, if <S2> = 0.8 or less, it is probably satisfactory. If it is 1.0 or so, it is certainly not satisfactory and the calculation should be rejected and a different approach taken. It requires experience to make this judgement.

            Given this drawback, why is the unrestricted Hartree-Fock method uses so frequently and in preference to the restricted open-shell Hartree-Fock (ROHF) method? The answer is that UHF is simpler to code, it is easier to develop post-Hartree-Fock methods, and it is unique, unlike ROHF where different Fock operators can give the same final wave function.

            Unrestricted Hartree-Fock theory was developed by John Pople and it is found in almost all ab initio programs.

            4.7. Hartree-Fock algorithm

            The Hartree-Fock method is typically used to solve the time-independent Schrödinger equation for a multi-electron atom or molecule as described in the Born-Oppenheimer appro­xi­mation. Since there are no known solutions for many-electron systems (hydrogenic atoms and the diatomic hydrogen cation being notable one-electron exceptions), the problem is solved numerically. Due to the nonlinearities introduced by the Hartree-Fock approximation, the equations are solved using a nonlinear method such as iteration, which gives rise to the name "self-consistent field method".

      The Hartree-Fock method makes five major simplifications in order to deal with this task:

·           The Born-Oppenheimer approximation is inherently assumed. The full molecular

wavefunction is actually a function of the coordinates of each of the nuclei, in addition to those of the electrons.

·                    Typically, relativistic effects are completely neglected. The momentum operator is assumed to be completely non-relativistic.

Fig.26. Greatly simplified algorithmic flowchart illustrating the Hartree-Fock Method

            The variational solution is assumed to be a linear combination of a finite number of basis functions, which are usually (but not always) chosen to be orthogonal. The finite basis set is assumed to be approximately complete.

            Each energy eigenfunction is assumed to be describable by a single Slater determinant, an antisymmetrized product of one-electron wavefunctions (i.e. orbitals). The mean field approximation is implied. Effects arising from deviations from this assumption, known as electron correlation, are completely neglected. (Electron correlation should not be confused with electron exchange, which is fully accounted for in the Hartree-Fock method.) 

            Relaxation of the last two approximations give rise to many so-called post-Hartree-Fock methods.

            4.8.  Variational Optimization of Orbitals

            The variational theorem states that for a time-independent Hamiltonian operator, any trial wavefunction will have an energy expectation value that is greater than or equal to the true ground state wavefunction corresponding to the given Hamiltonian. Because of this, the Hartree-Fock energy is an upper bound to the true ground state energy of a given molecule. In the context of the Hartree-Fock method, the best possible solution is at the Hartree-Fock limit, i.e. the limit of the Hartree-Fock energy as the basis set approaches completeness. (The other is the full-CI limit, where the last two approximations of the Hartree-Fock theory as described above are completely undone. It is only when both limits are attained that the exact solution is obtained.) In that case, the Hartree-Fock energy is the minimal energy over the set of Slater determinants.

The starting point for the Hartree-Fock method is a set of approximate one-electron wavefunctions known as orbitals. For an atomic calculation, these are typically the orbitals for a hydrogenic atom (an atom with only one electron, but the appropriate nuclear charge). For a molecular or crystalline calculation, the initial approximate one-electron wavefunctions are typically a linear combination of atomic orbitals (LCAO).

The orbitals above only account for the presence of other electrons in an average manner. In the Hartree-Fock method, the effect of other electrons are accounted for in a mean-field theory context. The orbitals are optimized by requiring them to minimize the energy of the respective Slater determinant. The resultant variational conditions on the orbitals lead to a new one-electron operator, the Fock operator. At the minimum, the occupied orbitals are eigensolutions to the Fock operator via a unitary transformation between themselves. The Fock operator is an effective one-electron Hamiltonian operator being the sum of two terms. The first is a sum of kinetic energy operators for each electron, the internuclear repulsion energy, and a sum of nuclear-electronic coulombic attraction terms. The second are Coulombic repulsion terms between electrons in a mean-field theory description; a net repulsion energy for each electron in the system, which is calculated by treating all of the other electrons within the molecule as a smooth distribution of negative charge. This is the major simplification inherent in the Hartree-Fock method, and is equivalent to the fifth simplification in the above list.

Since the Fock operator depends on the orbitals used to construct the corresponding Fock matrix, the eigenfunctions of the Fock operator are in turn new orbitals which can be used to construct a new Fock operator. In this way, the Hartree-Fock orbitals are optimized iteratively until the change in total electronic energy falls below a predefined threshold. In this way, a set of self-consistent one-electron orbitals are calculated. The Hartree-Fock electronic wavefunction is then the Slater determinant constructed out of these orbitals. Following the basic postulated of quantum mechanics, the Hartree-Fock wavefunction can then be used to compute any desired chemical or physical property within the framework of the Hartree-Fock method and the approximations employed.

            4.9. The Fock operator

Because the electron-electron repulsion term of the electronic molecular Hamiltonian involves the coordinates of two different electrons, it is necessary to reformulate it in an approximate way. Under this approximation, (outlined under Hartree-Fock algorithm), all of the terms of the exact Hamiltonian except the nuclear-nuclear repulsion term are re-expressed as the sum of one-electron operators outlined below. The "(1)" following each operator symbol simply indicates that the operator is 1-electron in nature.

\hat F[\{\phi_j\}](1) = \hat H^{core}(1)+\sum_{j=1}^{n/2}[2\hat J_j(1)-\hat K_j(1)]                                 (83)

where:

\hat F[\{\phi_j\}](1)

is the one-electron Fock operator generated by the orbitals φj ,

\hat H^{core}(1)=-\frac{1}{2}\nabla^2_1 - \sum_{\alpha} \frac{Z_\alpha}{r_{1\alpha}}                                            (84)

is the one-electron core Hamiltonian,  is the Coulomb operator, defining the electron-electron repulsion energy due to the orbital of the j-th electron, is the exchange operator, defining the electron exchange energy. Finding the Hartree-Fock one-electron wavefunctions is now equivalent to solving the eigenfunction equation:

\hat F(1)\phi_i(1)=\epsilon_i \phi_i(1)                                              (85)

where φi(1) are a set of one-electron wavefunctions, called the Hartree-Fock Molecular Orbitals.

Typically, in modern Hartree-Fock calculations, the one-electron wavefunctions are approximated by a Linear combination of atomic orbitals. These atomic orbitals are called Slater-type orbitals. Furthermore, it is very common for the "atomic orbitals" in use to actually be composed of a linear combination of one or more Gaussian-type orbitals, rather than Slater-type orbitals, in the interests of saving large amounts of computation time.

Various basis sets are used in practice, most of which are composed of Gaussian functions. In some applications, an orthogonalization method such as the Gram-Schmidt process is performed in order to produce a set of orthogonal basis functions. This can in principle save computational time when the computer is solving the Roothaan-Hall equations by converting the overlap matrix effectively to an identity matrix. However in most modern computer programs for molecular Hartree-Fock calculations this procedure is not followed due to the high numerical cost of orthogonalization and the advent of more efficient, often sparse, algorithms for solving the generalized eigenvalue problem, of which the Roothan-Hall equations are an example.

The Roothaan equations are a representation of the Hartree-Fock equation in a non orthonormal basis set which can be of Gaussian-type or Slater-type. It applies to closed-shell molecules or atoms where all molecular orbitals or atomic orbitals, respectively, are doubly occupied. This is generally called Restricted Hartree-Fock theory.

The method was developed independently by Clemens C. J. Roothaan and George G. Hall in 1951, and is thus sometimes called the Roothaan-Hall equations. The Roothaan equations can be written in the form of generalized eigenvalue problem

FC=SCΕ                                                                                                  (86)

Where F is the so-called Fock matrix, C is a matrix of coefficients, S is the overlap matrix of the basis functions, and ε is the (diagonal, by convention) matrix of orbital energies. In the case of an orthonormalised basis set the overlap matrix, S, reduces to the identity matrix.

            4.10. Numerical stability

            Numerical stability can be a problem with this procedure and there are various ways of combating this instability. One of the most basic and generally applicable is called F-mixing or damping. With F-mixing, once a single electron wavefunction is calculated it is not used directly. Instead, some combination of that calculated wavefunction and the previous wavefunctions for that electron is used - the most common being a simple linear combination of the calculated and immediately preceding wavefunction. A clever dodge, employed by Hartree, for atomic calculations was to increase the nuclear charge, thus pulling all the electrons closer together. As the system stabilised, this was gradually reduced to the correct charge. In molecular calculations a similar approach is sometimes used by first calculating the wavefunction for a positive ion and then to use these orbitals as the starting point for the neutral molecule. Modern molecular Hartree-Fock computer programs use a variety of methods to ensure convergence of the Roothaan-Hall equations.

Of the four simplifications outlined under Hartree-Fock algorithm, the fourth is typically the most important. Neglecting electron correlation can lead to large deviations from experimental results. A number of approaches to this weakness, collectively called post-Hartree-Fock methods, have been devised to include electron correlation to the multi-electron wave function. One of these approaches, Møller-Plesset perturbation theory, treats correlation as a perturbation of the Fock operator. Others expand the true multi-electron wavefunction in terms of a linear combination of Slater determinants - such as Multi-configurational self-consistent field, Configuration interaction, Quadratic configuration interaction, Complete active space SCF (CASSCF).

An alternative to Hartree-Fock calculations used in some cases is density functional theory, which gives approximate solutions to both exchange and correlation energies, but is not a purely ab initio method in practice. Indeed, it is common to use calculations that are a hybrid of the two methods - the popular B3LYP schema is one such hybrid functional method.

            Multi-configurational self-consistent field (MCSCF) is a method in quantum chemistry used to generate qualitatively correct reference states of molecules in cases where Hartree-Fock and density functional theory are not adequate (e.g., for molecular ground states which are quasi-degenerate with low lying excited states or in bond breaking situations). It uses a linear combination of configuration state functions (CSF) or configuration determinants to approximate the exact electronic wavefunction of an atom or molecule. In an MCSCF calculation, the set of coefficients of both the CSFs or determinants and the basis functions in the molecular orbitals are varied to obtain the total electronic wavefunction with the lowest possible energy. This method can be considered an intermediate between configuration interaction (where the molecular orbitals are not varied but the expansion of the wave function) and Hartree-Fock (where there is only one determinant but the molecular orbitals are varied).

MCSCF wave functions are often used as reference states for Multireference configuration interaction (MRCI) or multi-reference perturbation theories like Complete Active Space Perturbation Theory (CASPT2). These methods can deal with extremely complex chemical situations and, if computing power permits, may be used to reliably calculate molecular ground- and excited states if all other methods fail.

            4.11. Comparison of valence bond and molecular orbital theory

            In some respects valence bond theory is superior to molecular orbital theory. When applied to the simplest two-electron molecule, H2, valence bond theory, even at the simplest Heitler-London approach, gives a much closer approximation to the bond energy, and it provides a much more accurate representation of the behavior of the electrons as chemical bonds are formed and broken. In contrast simple molecular orbital theory predicts that the hydrogen molecule dissociates into a linear superposition of hydrogen atoms and positive and negative hydrogen ions, a completely unphysical result. This explains in part why the curve of total energy against interatomic distance for the valence bond method lies above the curve for the molecular orbital method at all distances and most particularly so for large distances. This situation arises for all homonuclear diatomic molecules and is particularly a problem for F2, where the minimum energy of the curve with molecular orbital theory is still higher in energy than the energy of two F atoms.

The concepts of hybridization are so versatile, and the variability in bonding in most organic compounds is so modest, that valence bond theory remains an integral part of the vocabulary of organic chemistry. However, the work of Friedrich Hund, Robert Mulliken, and Gerhard Herzberg showed that molecular orbital theory provided a more appropriate description of the spectroscopic, ionization and magnetic properties of molecules. The deficiencies of valence bond theory became apparent when hypervalent molecules (e.g. PF5) were explained without the use of d orbitals that were crucial to the bonding hybridisation scheme proposed for such molecules by Pauling. Metal complexes and electron deficient compounds (e.g. diborane) also appeared to be well described by molecular orbital theory, although valence bond descriptions have been made.

            In the 1930s the two methods strongly competed until it was realised that they are both approximations to a better theory. If we take the simple valence bond structure and mix in all possible covalent and ionic structures arising from a particular set of atomic orbitals, we reach what is called the full configuration interaction wave function. If we take the simple molecular orbital description of the ground state and combine that function with the functions describing all possible excited states using unoccupied orbitals arising from the same set of atomic orbitals, we also reach the full configuration interaction wavefunction. It can be then seen that the simple molecular orbital approach gives too much weight to the ionic structures, while the simple valence bond approach gives too little. This can also be described as saying that the molecular orbital approach is too delocalised, while the valence bond approach is too localised.

            The two approaches are now regarded as complementary, each providing its own insights into the problem of chemical bonding. Modern calculations in quantum chemistry usually start from (but ultimately go far beyond) a molecular orbital rather than a valence bond approach, not because of any intrinsic superiority in the former but rather because the MO approach is more readily adapted to numerical computations. However better valence bond programs are now available.

            Chapter 5. Bases of Computational Chemistry

            Computational chemistry is a branch of chemistry that uses computers to assist in solving chemical problems. It uses the results of theoretical chemistry, incorporated into efficient computer programs, to calculate the structures and properties of molecules and solids. While its results normally complement the information obtained by chemical experiments, it can in some cases predict hitherto unobserved chemical phenamena. It is widely used in the design of new drugs and materials.

            Examples of such properties are structure (i.e. the expected positions of the constituent atoms), absolute and relative (interaction) energies, electronic charge distributions, dipoles and higher multipole moments, vibrational frequinces, reactivity or other sprctroscopic quantities, and cross sections for collision with other particles.

            The methods employed cover both static and dynamic situations. In all cases the computer time increases rapidly with the size of the system being studied. That system can be a single molecule, a group of molecules or a solid. The methods are thus based on theories which range from highly accurate, but are suitable only for small systems, to very approximate, but suitable for very large systems. The accurate methods are used called ab initio methods, as they are based entirely on theory from first principles. The less accurate methods are called empirical or semi-empirical because some experimental results, often from atoms or related molecules, are used along with the theory.

            5.1. The basic stages of the development of Computational Chemistry

            Building on the founding discoveries and theories in the history of quantum mechanics, the first theoretical calculations in chemistry were those of W. Heitler and F. London in 1927. The books that were influential in the early development of computational quantum chemistry include: L. Pauling and E. Bright Wilson’s 1935 “Introduction to Quantum Mechanics” – with Applications to Chemistry, Eyring, Walter and Kimball's 1944 “Quantum Chemistry”, Heitler’s 1945 “Elementary Wave Mechanics” – with Applications to Quantum Chemistry, and later Coulson's 1952 textbook “Valence”, each of which served as primary references for chemists in the decades to follow.

            With the development of efficient computer technology in the 1940s the solutions of elaborate wave equations for complex atomic systems began to be a realizable objective. Below the chronology of development of the computer chemistry is put in order:

v                  In the early 1950s, the first semi-empirical atomic orbital calculations were carried out. Theoretical chemists became extensive users of the early digital computers.

v                  At the same time a very detailed account of such use in the United Kingdom is given by Smith and Sutcliffe.

v                  The first polyatomic calculations using Gaussian orbitals were carried out in the late 1950s.

v                  The first configuration interaction calculations were carried out in Cambridge on the EDSAC computer in the 1950s using Gaussian orbitals by Boys and coworkers.

v                  The first ab initio Hartree-Fock calculations on diatomic molecules were carried out in 1956 at Massachusetts Institute of Technology (MIT) using a basis set of Slater orbitals.

v                  For diatomic molecules a systematic study using a minimum basis set and the first calculation with a larger basis set were published by Ransil and Nesbet respectively in 1960.

v                  In 1964, Hückel method calculations, which are a simple LCAO method for the determination of electron energies of molecular orbitals of π electrons in conjugated hydrocarbon systems, ranging from simple systems such as butadiene and benzene to ovalene with 10 fused six-membered rings , were generated on computers at Berkeley and Oxford. These empirical methods were replaced in the 1960s by semi-empirical methods such as CNDO.

v                  In the early 1970s, efficient ab initio computer programs such as ATMOL, GAUSSIAN, IBMOL, and POLYAYTOM, began to be used to speed up ab initio calculations of molecular orbitals. Of these four programs only GAUSSIAN, massively expanded, is still in use, but many other programs are now in use. At the same time, the methods of molecular mechanics, such as MM2, were developed, primarily by Norman Allinger.

v                  By 1971, when a bibliography of ab initio calculations was published, the largest molecules included were naphthalene and azulene. Abstracts of many earlier developments in ab initio theory have been published by Schaefer.

            One of the first mentions of the term “computational chemistry” can be found in the 1970 book Computers and Their Role in the Physical Sciences by Sidney Fernbach and Abraham Haskell Taub, where they state “It seems, therefore, that 'computational chemistry' can finally be more and more of a reality.” During the 1970s, widely different methods began to be seen as part of a new emerging discipline of computational chemistry. The Journal of Computational Chemistry was first published in 1980.

            5.2. The basic concepts of Computional Chemistry

            The term theoretical chemistry may be defined as a mathematical description of chemistry, whereas computational chemistry is usually used when a mathematical method is sufficiently well developed that it can be automated for implementation on a computer. Note that the words exact and perfect do not appear here, as very few aspects of chemistry can be computed exactly. Almost every aspect of chemistry, however, can be described in a qualitative or approximate quantitative computational scheme.

            Molecules consist of nuclei and electrons, so the methods of quantum mechanics apply. Computational chemists often attempt to solve the non-relativistic Schrödinger equation, with relativistic corrections added, although some progress has been made in solving the fully relativistic Schrödinger equation. It is, in principle, possible to solve the Schrödinger equation, in either its time-dependent form or time-independent form as appropriate for the problem in hand, but this in practice is not possible except for very small systems. Therefore, a great number of approximate methods strive to achieve the best trade-off between accuracy and computational cost. Accuracy can always be improved with greater computational cost.

            Present computational chemistry can routinely accurately calculate the properties of molecules that contain up to about 40 electrons. Errors for energies can be less than 1 kcal/mol. For geometries, bond lengths can be predicted within a few picometres and bond angles within 0.5o. The treatment of larger molecules that contain a few dozen electrons is computationally tractable by approximate methods such as density functional theory (DFT). There is some dispute within the field whether the latter methods are sufficient to describe complex chemical reactions, such as those in biochemistry. Large molecules can be studied by semi-empirical approximate methods. Even larger molecules are treated by classical mechanics methods that are called molecular mechanics.

            In theoretical chemistry, chemists, physicists and mathematicians develop algorithms and computer programs to predict atomic and molecular properties and reaction paths for chemical reactions. Computational chemists, in contrast, may simply apply existing computer programs and methodologies to specific chemical questions. There are some aspects of computational chemistry:

·          computational studies can be carried out in order to find a starting point for a laboratory synthesis, or to assist in understanding experimental data, such as the position and source of spectroscopic peaks,

·          computational studies can be used to predict the possibility of so far entirely unknown molecules or to explore reaction mechanisms that are not readily studied by experimental means,

·          the prediction of the molecular structure of molecules by the use of the simulation of forces, or more accurate quantum chemical methods, to find stationary points on the energy hypersurface as the position of the nuclei is varied,

·          storing and searching for data on chemical entities),

·          identifying correlations between chemical structures and properties,

·          computational approaches to help in the efficient synthesis of compounds.

·          computational approaches to design molecules that interact in specific ways with other molecules.

       Quantum Chemistry possesses several methods of calculation molecular orbital. Among them the following are especially important:

·          ab initio methods,

·          density functional methods,

·          semi-empirical methods,

·          molecular mechanics,

·          methods for solids,

·          Chemical dynamics.

            Each of that methods have own features, advantages and lacks. In the further we shall consider each of them in details.

            5.3. Ab initio Quantum Chemistry methods

            The programs used in computational chemistry are based on many different quantum chemical methods that solve the molecular Schrödinger equation associated with the molecular Hamiltonian. Methods that do not include any empirical or semi-empirical parameters in their equations - being derived directly from theoretical principles, with no inclusion of experimental data - are called as ab initio methods. This does not imply that the solution is an exact one; they are all approximate quantum mechanical calculations. It means that a particular approximation is rigorously defined on first principles (quantum theory) and then solved within an error margin that is qualitatively known beforehand. If numerical iterative methods have to be employed, the aim is to iterate until full machine accuracy is obtained (the best that is possible with a finite word length on the computer, and within the mathematical and/or physical approximations made).

            Ab initio methods are one of the basic methods of computational chemistry. It includes in itself following calculation methods:

Ø      Hartree-Fock calculation methods,

Ø      Post-Hartree-Fock methods,

Ø      Multi-reference methods,

            Many types of calculations begin with a Hartree-Fock calculation and subsequently correct for electron-electron repulsion, referred to also as electronic correlation. Møller-Plesset perturbation theory (MPn) and coupled cluster theory (CC) are examples of these post-Hartree-Fock methods. In some cases, particularly for bond breaking processes, the Hartree-Fock method is inadequate and this single-determinant reference function is not a good basis for post-Hartree-Fock methods. It is then necessary to start with a wave function that includes more than one determinant such as Multi-configurational self-consistent field and methods have been developed that use these multi-determinant references for improvements.

            Almost always the basis set (which is usually built from the LCAO combination) used to solve the Schrödinger equation is not complete, and does not span the Hilbert space associated with ionization and scattering processes (see continuous spectrum for more details). In the Hartree-Fock method and the Configuration interaction method, this approximation allows one to treat the Schrödinger equation as a "simple" eigenvalue equation of the electronic molecular Hamiltonian, with a discrete set of solutions.

            The simplest type of ab initio electronic structure calculation is the Hartree-Fock (HF) scheme, in which the instantaneous Coulombic electron-electron repulsion is not specifically taken into account. Only its average effect (mean field) is included in the calculation. This is a variational procedure, therefore the obtained approximate energies, expressed in terms of the system's wave function, are always equal to or greater than the exact energy, and tend to a limiting value called the Hartree-Fock limit as the size of the basis is increased.

            The most popular classes of ab initio electronic structure in the frame Hartree-Fock methods are:

-         Hartree-Fock (HF),

-         Restricted Open-shell Hartree-Fock (ROHF),

-         Unrestricted Hartree-Fock (UHF).

And in the frame Post-Hartree-Fock methods important methods are:

-         Møller-Plesset perturbation theory (MPn)

-         Configuration interaction (CI)

-         Coupled cluster (CC)

-         Quadratic configuration interaction (QCI)

-         Quantum chemistry methods

            Following methods concern to multi-reference methods:

-         Multi-configurational self-consistent field (MCSCF)

-         Multi-reference configuration interaction (MRCI)

-         N-Electron Valence state Perturbation Theory (NEVPT)

-         Complete Active Space Perturbation Theory (CASPTn)

            Many types of calculations, known as post-Hartree-Fock methods, begin with a Hartree-Fock calculation and subsequently correct for electron-electron repulsion, referred to also as electronic correlation. As these methods are pushed to the limit, they approach the exact solution of the non-relativistic Schrödinger equation. Below relative values of molecular energy of calculated by various methods are shown.

 Fig. 27. Diagram illustrating various ab initio electronic structure methods in terms of energy

           

            In order to obtain exact agreement with experiment, it is necessary to include relativistic and spin orbit terms, both of which are only really important for heavy atoms. In all of these approaches, in addition to the choice of method, it is necessary to choose a basis set. This is a set of functions, usually centered on the different atoms in the molecule, which are used to expand the molecular orbitals with the LCAO ansatz. Ab initio methods need to define a level of theory (the method) and a basis set.

            The Hartree-Fock wave function is a single configuration or determinant. In some cases, particularly for bond breaking processes, this is quite inadequate and several configurations need to be used. Here the coefficients of the configurations and the coefficients of the basis functions are optimized together.

            The total molecular energy can be evaluated as a function of the molecular geometry, in other words the potential energy surface. Such a surface can be used for reaction dynamics. The stationary points of the surface lead to predictions of different isomers and the transition structures for conversion between isomers, but these can be determined without a full knowledge of the complete surface.

            Ab initio electronic structure methods have the advantage that they can be made to converge to the exact solution, when all approximations are sufficiently small in magnitude. In particular configuration interaction where all possible configurations are included (called "Full CI") tends to the exact non-relativistic solution of the Schrödinger equation. The convergence, however, is usually not monotonic, and sometimes the smallest calculation gives the best result for some properties. The downside of ab initio methods is their computational cost. They often take enormous amounts of computer time, memory, and disk space. The HF method scales nominally as N4 (N being the number of basis functions) – i.e. a calculation twice as big takes 16 times as long to complete. However in practice it can scale closer to N³ as the program can identify zero and extremely small integrals and neglect them. Correlated calculations scale even less favorably - MP2 as N5; MP4 as N6 and coupled cluster as N7. DFT methods scale in a similar manner to Hartree-Fock but with a larger proportionality term. Thus DFT calculations are always more expensive than an equivalent Hartree-Fock calculation.

            The problem of computational expense can be alleviated through simplification schemes. In the density fitting scheme, the four-index integrals used to describe the interaction between electron pairs are reduced to simpler two- or three-index integrals, by treating the charge densities they contain in a simplified way. This reduces the scaling with respect to basis set size. Methods employing this scheme are denoted by the prefix "df-", for example the density fitting MP2 is df-MP2 (lower-case is advisable to prevent confusion with DFT). In the local approximation, the molecular orbitals are first localized by a unitary rotation in the orbital space (which leaves the reference wave function invariant, i.e., is not an approximation) and subsequently interactions of distant pairs of localized orbtials are neglected in the correlation calculation. This sharply reduces the scaling with molecular size, a major problem in the treatment of biologically-sized molecules. Methods employing this scheme are denoted by the prefix "L", e.g. LMP2. Both schemes can be employed together, as in the recently developed df-LMP2 and df-LCCSD(T0) methods. In fact, df-LMP2 calculations are faster than df-Hartree-Fock calculations and thus are feasible in nearly all situations in which also DFT is.

            5.3.1. Valence bond methods in ab initio

            Valence bond (VB) methods are generally ab initio although some semi-empirical versions have been proposed. Current VB approaches are:

            The generalized valence bond (GVB) method is one of the simplest and oldest valence bond method that uses flexible orbitals in the general way used by modern valence bond theory. The method was developed by the group of William A. Goddard, III around 1970. The generalised Coulson-Fisher for the hydrogen molecule, discussed in Modern valence bond theory is generalised to describe every electron pair in a molecule. The pair of orbitals for each electron pair are expanded in terms of the full basis set and are non-orthogonal. Orbitals from different pairs are forced to be orthogonal - the strong orthogonality condition. This condition simplifies the calculation but can lead to some difficulties.

            GVB code in some programs, particularly GAMES (US), can also be used to do a variety of restricted open-shell Hartree-Fock calculations, such as those with one or three electrons in two pi-electron molecular orbitals while retaining the degeneracy of the orbitals. This wave function is essentially a two-determinant function, rather than the one-determinant function of the restricted Hartree-Fock method.

            GAMESS (US) is a computational chemistry software program that stands for General Atomic and Molecular Electronic Structure System. The original code split in 1981 into GAMESS (US) and GAMESS (UK) variants, which now differ significantly. GAMESS is maintained by the members of the Gordon research group at Iowa State University. GAMESS (US) can perform a number of general computational chemistry calculations, including Hartree-Fock, Density functional theory (DFT), Generalized Valence Bond (GVB), and Multi-configurational self-consistent field (MCSCF). Correlation corrections after these SCF calculations can be estimated by configuration interaction (CI), second order Møller-Plesset perturbation theory, and coupled cluster theory. Solvent effect can be considered using discrete effective fragment potentials or continuum models (such as PCM). Relativistic corrections can be calculated, including third order Douglas-Kroll scalar terms.

            While the program does not perform molecular mechanics, it can be interfaced with the TINKER code for molecular mechanics to do mixed molecular mechanics/quantum mechanics calculations. The Fragment Molecular Orbital method can be used to treat large systems, by dividing them into fragments.

            It can also be interfaced with the valence bond VB2000 and XMVB programs and the Natural Bond Order (NBO) analysis program.

            Modern valence bond theory is the term used to describe applications of valence bond theory with computer programs that are competitive in accuracy and economy with programs for the Hartree-Fock method and other molecular orbital based methods. The latter methods dominated quantum chemistry from the advent of digital computers because they were easier to program. The early popularity of valence bond methods thus declined. It is only recently that the programming of valence bond methods has improved. These developments are due to and described by Gerratt, Cooper, Karadakov and Raimondi (1997); Li and McWeeny (2002); Song, Mo, Zhang and Wu (2005); and Shaik and Hiberty (2004).

            In its simplest form the overlapping atomic orbitals are replaced by orbitals which are expanded as linear combinations of the atom-based basis functions. This expansion is optimized to give the lowest energy. This procedure gives good energies without including ionic structures.

For example, in the hydrogen molecule, classic valence bond theory uses two 1s atomic orbitals (a and b) on the two hydrogen atoms respectively and then constructs a covalent structure:

ΦC = [a(1)b(2) + b(1)a(2)] [α(1)β(2) - β(1)α(2)]                                  (87)

and then an ionic structure:

ΦI = [a(1)a(2) + b(1)b(2)] [α(1)β(2) - β(1)α(2)]                                   (88)

The final wave function is a linear combination of these two functions. Coulson and Fischer pointed out that a completely equivalent function is:

ΦCF = ((a+kb)(1)(b+ka)(2) + (b+ka)(1)(a+kb)(2)) ((α(1)β(2) - β(1)α(2))                 (89)

as expanding this out gives a linear combination of the covalent and ionic structures. Modern valence bond theory replaces the simple linear combination of the two atomic orbitals with a linear combination of all orbitals in a larger basis set. The two resulting valence bond orbitals look like an atomic orbital on one hydrogen atom slightly distorted towards the other hydrogen atom. Modern valence bond theory is thus an extension of the Coulson-Fischer method.

            There are a large number of different valence bond methods. Most use n valence bond orbitals for n electrons. If a single set of these orbitals is combined with all linear independent combinations of the spin functions, we have spin-coupled valence bond theory. The total wave function is optimized using the variation principle by varying the coefficients of the basis functions in the valence bond orbitals and the coefficients of the different spin functions. In other cases only a sub-set of all possible spin functions is used. Many valence bond methods use several sets of the valence bond orbitals. Be warned that different authors use different names for these different valence bond methods.

            Several groups have produced computer programs for modern valence bond calculations that are freely available.

            5.3.2. Quantum Monte Carlo methods in ab initio

            A method that avoids making the variational overestimation of HF in the first place is Quantum Monte Carlo (QMC), in its variational, diffusion, and Green's function forms. These methods work with an explicitly correlated wave function and evaluate integrals numerically using a Monte Carlo integration. Such calculations can be very time-consuming, but they are probably the most accurate methods known today.

            Quantum Monte Carlo is a large class of computer algorithms that simulate quantum systems with the idea of solving the many-body problem. They use, in one way or another, the Monte Carlo method to handle the many dimensional integrals that arise. Quantum Monte Carlo allows a direct representation of many-body effects in the wavefunction, at the cost of statistical uncertainty that can be reduced with more simulation time. For bosons, there exist numerically exact and polynomial-scaling algorithms. For fermions, there exist very good approximations and numerically exact exponentially scaling quantum Monte Carlo algorithms, but none that are both.

            In principle, any physical system can be described by the many-body Schrödinger equation, as long as the constituent particles are not moving 'too' fast; that is, they are not moving near the speed of light. This includes the electrons in almost every material in the world, so if we could solve the Schrödinger equation, we could predict the behavior of any electronic system, which has important applications in fields from computers to biology. This also includes the nuclei in Bose-Einstein condensates and superfluids like liquid helium. The difficulty is that the Schrödinger equation involves a function of three times the number of particles (in 3 dimensions), and is difficult (and impossible in the case of fermions) to solve in a reasonable amount of time. Traditionally, theorists have approximated the many-body wave function as an antisymmetric function of one-body orbitals:

,          (90)

 for an example, see Hartree-Fock theory. This kind of formulation either limits the possible wave functions, as in the case of Hartree-Fock, or converges very slowly, as in configuration interaction. One of the reasons for the difficulty with a Hartree-Fock ansatz is that it is very difficult to model the electronic and nuclear cusps in the wavefunction. As two particles approach each other, the wavefunction has exactly known derivatives.

            Quantum Monte Carlo is a way around these problems because it allows us to model a many-body wave function of our choice directly. Specifically, we can use a Hartree-Fock wavefunction as our starting point, but then multiply it by any symmetric function, of which Jastrow functions are typical, designed to enforce the cusp conditions. Most methods aim at computing the ground state wave function of the system, with the exception of Path integral Monte Carlo and finite-temperature auxiliary field Monte Carlo, which calculate the density matrix.

            There are several quantum Monte Carlo flavors, each of which uses Monte Carlo in different ways to solve the many-body problem:

  • Variational Monte Carlo  have a good place to start; it is commonly used in many sorts of quantum problems.
  • Diffusion Monte Carlo is the most common high-accuracy method for electrons (that is, chemical problems), since it comes quite close to the exact ground state energy fairly efficiently. Also used for simulating the quantum behavior of atoms, etc.
  • Path integral Monte Carlo is method finite temperature technique mostly applied to bosons where temperature is very important, especially superfluid helium.
  • Auxiliary field Monte Carlo is the method which usually applied to lattice problems, although there has been recent work on applying it to electrons in chemical systems.
  • Reputation Monte Carlo is recent zero-temperature method related to Path integral Monte Carlo, with applications similar to Diffusion Monte Carlo, but with some different tradeoffs.
  • Gaussian quantum Monte Carlo is a quantum Monte Carlo method that shows a potential solution to the fermion sign problem without the deficiencies of alternative approaches. Instead of the Hilbert space, this method works in the space of density matrices that can be spanned by an over-complete basis of gaussian operators using only positive coefficients. Containing only quadratic forms of the fermionic operators, no anti-commuting variables occur and any quantum state can be expressed as a real probability distribution.

            The following computer programs is working in framework of ab initio by a method of Monte-Carlo:

Ø      CASINO,

Ø      CHAMP,

Ø      QMcBeaver,

Ø      QMCMOL,

Ø      QMC@HOME,

Ø      QWalk,

Ø      ZORI,

Ø      TurboPair.

            5.3.3. Gaussian theory

            The Gaussian methods (G1, G2, and G3) are also unique types of computations. These methods arose from the observation that certain ab initio methods tended to show a systematic error for predicting the energies of the ground states of organic molecules. This observation resulted in a correction equation that uses the energies from several different ab initio calculations in order to extrapolate to a very-high-accuracy result. All the calculations that go into this extrapolation are ab initio methods. However, the extrapolation equation itself is an empirically defined equation parameterized to reproduce results from a test set of molecules as accurately as possible. The extrapolation to complete correlation is based on the number of electrons times an empirically determined constant. For this reason, these methods show the same strengths and weaknesses as other semiempirical methods. The accuracy can be extremely good for systems similar to those for which they were parameterized, the ground state of organic molecules. However, for other systems, such as transition structures or clusters, these methods often are less accurate than some less computationally intensive ab initio methods. J. A. Pople once referred to this as a ``slightly empirical theory.''

            The G1 method is seldom used since G2 yields an improved accuracy of results. G2 has proven to be a very accurate way to model small organic molecules, but gives poor accuracy when applied to chlorofluorocarbons.

            5.4. Density Functional Theory

            Density functional theory (DFT) is a quantum mechanical theory used in physics and chemistry to investigate the ground state of many-body systems, in particular atoms, molecules and the condensed phases. DFT is among the most popular and versatile methods available in condensed matter physics, computational physics, and computational chemistry.

            Density functional theory (DFT) methods are often considered to be ab initio methods for determining the molecular electronic structure, even though many of the most common functionals use parameters derived from empirical data, or from more complex calculations. This means that they could also be called semi-empirical methods. It is best to treat them as a class on their own. In DFT, the total energy is expressed in terms of the total electron density rather than the wave function. In this type of calculation, there is an approximate Hamiltonian and an approximate expression for the total electron density. DFT methods can be very accurate for little computational cost. The drawback is, that unlike ab initio methods, there is no systematic way to improve the methods by improving the form of the functional. Some methods combine the density functional exchange functional with the Hartree-Fock exchange term and are known as hybrid functional methods.

            5.4.1. Thomas-Fermi model

            The predecessor to density functional theory was the Thomas-Fermi model, developed by Thomas and Fermi in 1927. They used a statistical model to approximate the distribution of electrons in an atom. The mathematical basis used was to postulate that electrons are distributed uniformly in phase space with two electrons in every h3 of volume. For each element of coordinate space volume d3r we can fill out a sphere of momentum space up to the fermi momentum pf

equating the number of electrons in coordinate space to that in phase space gives:

solving for pf and substituting in the classical kinetic energy formula then leads directly to a kinetic energy represented as a functional of the electron density:

                                                            (91)

As such they were able to calculate the energy of an atom using this kinetic energy functional combined with the classical expressions for the nuclear-electron and electron-electron interactions (which can both also be represented in terms of the electron density)

                                             (92)

            Although this was an important first step, the Thomas-Fermi equation's accuracy is limited because the resulting kinetic energy functional is only approximate, and also the method does not attempt to represent the exchange energy of an atom as a conclusion of the Pauli principle. An exchange energy functional was added by Dirac in 1928.

            However, the Thomas-Fermi-Dirac theory remained rather inaccurate for most applications. The largest source of error was in the representation of the kinetic energy, followed by the errors in the exchange energy, and due to the complete neglect of electron correlation.

Teller (1962) showed that Thomas-Fermi theory cannot describe molecular bonding. This can however be overcome by improving the kinetic energy functional.

The kinetic energy functional can be improved by adding the Weizsäcker (1935) correction:                                                       (93)

            Although density functional theory has its conceptual roots in the Thomas - Fermi model, DFT was not put on a firm theoretical footing until the Hohenberg-Kohn theorems (after Pierre Hohenberg and Walter Kohn) (HK). The first of these demonstrates the existence of a one-to-one mapping between the ground state electron density and the ground state wavefunction of a many-particle system. Further, the second HK theorem proves that the ground state density minimizes the total electronic energy of the system. The original HK theorems held only for the ground state in the absence of magnetic field, although they have since been generalized. The first Hohenberg-Kohn theorem is only an existence theorem, stating that the mapping exists, but does not provide any such exact mapping. It is in these mappings that approximations are made. (The theorems can be extended to the time-dependent domain (TDDFT), which can be also used to determine excited states.)

            5.4.2. Description of the DFM

            Traditional methods in electronic structure theory, in particular Hartree-Fock theory and its descendants, are based on the complicated many-electron wavefunction. The main objective of density functional theory is to replace the many-body electronic wavefunction with the electronic density as the basic quantity. Whereas the many-body wavefunction is dependent on 3N variables, three spatial variables for each of the N electrons, the density is only a function of three variables and is a simpler quantity to deal with both conceptually and practically.

            The most common implementation of density functional theory is through the Kohn-Sham method. Within the framework of Kohn-Sham DFT, the intractable many-body problem of interacting electrons in a static external potential is reduced to a tractable problem of non-interacting electrons moving in an effective potential. The effective potential includes the external potential and the effects of the Coulomb interactions between the electrons, e.g. the exchange and correlation interactions. Modeling the latter two interactions becomes the difficulty within KS DFT. The simplest approximation is the local-density approximation (LDA), which is based upon exact exchange energy for a uniform electron gas, which can be obtained from the Thomas-Fermi model, and from fits to the correlation energy for a uniform electron gas.

            DFT has been very popular for calculations in solid state physics since the 1970s. In many cases DFT with the local-density approximation gives quite satisfactory results, for solid-state calculations, in comparison to experimental data at relatively low computational costs when compared to other ways of solving the quantum mechanical many-body problem. However, it was not considered accurate enough for calculations in quantum chemistry until the 1990s, when the approximations used in the theory were greatly refined to better model the exchange and correlation interactions. DFT is now a leading method for electronic structure calculations in both fields. Despite the improvements in DFT, there are still difficulties in using density functional theory to properly describe intermolecular interactions, especially van der Waals forces (dispersion), or in calculations of the band gap in semiconductors. Its poor treatment of dispersion renders DFT unsuitable (at least when used alone) for the treatment of systems which are dominated by dispersion (e.g. interacting noble gas atoms) or where dispersion competes significantly with other effects (e.g. in biomolecules). The development of new DFT methods designed to overcome this problem, by alterations to the functional or by the inclusion of additive terms, is a current research topic.

            5.4.3. Derivation and formalism

As usual in many-body electronic structure calculations, the nuclei of the treated molecules or clusters are seen as fixed (the Born-Oppenheimer approximation), generating a static external potential V in which the electrons are moving. A stationary electronic state is then described by a wave function   fulfilling the many-electron Schrödinger equation

                     (94)

where H is the electronic molecular Hamiltonian, N is the number of electrons and U is the electron-electron interaction. The operators T and U are so-called universal operators as they are the same for any system, while V is system dependent or non-universal. As one can see, the actual difference between a single-particle problem and the much more complicated many-particle problem just arises from the interaction term U. There are many sophisticated methods for solving the many-body Schrödinger equation based on the expansion of the wave function in Slater determinants. While the simplest one is the Hartree-Fock method, more sophisticated approaches are usually categorized as post-Hartree-Fock methods. However, the problem with these methods is the huge computational effort, which makes it virtually impossible to apply them efficiently to larger, more complex systems.

            Here DFT provides an appealing alternative, being much more versatile as it provides a way to systematically map the many-body problem, with U, onto a single-body problem without U. In DFT the key variable is the particle density  which is given by

                   (95)

Hohenberg and Kohn proved in 1964 that the relation expressed above can be reversed, i.e.to a given ground state density  it is in principle possible to calculate the corresponding ground state wavefunction . In other words, Ψ0 is a unique functional of n0, i.e.

Ψ0 = Ψ0[n]

and consequently all other ground state observables O are also functionals of n0

From this follows, in particular, that also the ground state energy is a functional of  n0

,                                (96)

where the contribution of the external potential  can be written explicitly in terms of the density

                                                  (97)

The functionals T[n] and U[n] are called universal functionals while V[n] is obviously non-universal, as it depends on the system under study. Having specified a system, i.e. V is known, one then has to minimise the functional

                                  (98)

with respect to , assuming one has got reliable expressions for T[n] and U[n]. A successful minimisation of the energy functional will yield the ground state density n0 and thus all other ground state observables.

            The variational problem of minimising the energy functional E[n] can be solved by applying the Lagrangian method of undetermined multipliers, which was done by Kohn and Sham in 1965. Hereby, one uses the fact that the functional in the equation above can be written as a fictitious density functional of a non-interacting system

                                              (99)

where TS denotes the non-interacting kinetic energy and VS is an external effective potential in which the particles are moving. Obviously,   if VS is chosen to be

                                                  (100)

Thus, one can solve the so-called Kohn-Sham equations of this auxiliary non-interacting system

                                      (101)

which yields the orbitals fi that reproduce the density of the original many-body system

                                          (102)

            The effective single-particle potential VS can be written in more detail as

                                 (103)

where the second term denotes the so-called Hartree term describing the electron-electron Coulomb repulsion, while the last term VXC is called the exchange correlation potential. Here, VXC includes all the many-particle interactions. Since the Hartree term VXC and depend on , which depends on the fi, which in turn depend on VS, the problem of solving the Kohn-Sham equation has to be done in a self-consistent (i.e. iterative) way. Usually one starts with an initial guess for , then calculates the corresponding and solves the Kohn-Sham equations for the fi. From these one calculates a new density and starts again. This procedure is then repeated until convergence is reached.

            5.4.4. Approximations in the framework of DFT

            The major problem with DFT is that the exact functionals for exchange and correlation are not known except for the free electron gas. However, approximations exist which permit the calculation of certain physical quantities quite accurately. In physics the most widely used approximation is the local-density approximation (LDA), where the functional depends only on the density at the coordinate where the functional is evaluated:

                                              (104)

The local spin-density approximation (LSDA) is a straightforward generalization of the LDA to include electron spin:

                                    (105)

Highly accurate formulae for the exchange-correlation energy density  have been constructed from quantum Monte Carlo simulations of a free-electron gas.

            Generalized gradient approximations (GGA) are still local but also take into account the gradient of the density at the same coordinate:

                      (106)

Using the latter (GGA) very good results for molecular geometries and ground state energies have been achieved. Many further incremental improvements have been made to DFT by developing better representations of the functionals.

            The DFT formalism above breaks down in the presence of a vector potential, i.e. a magnetic field. In such a case, the one-to-one mapping between electron density and external potential breaks down. Generalizations to include the effects of magnetic fields have led to two different theories: current density functional theory and magnetic field functional theory. In both these theories, the functional used for the exchange and correlation must be generalized to include more than just the electron density. In current density functional theory, developed by Vignale and Rasolt, the functionals become dependent on both the electron density and the current density. In magnetic field density functional theory, developed by Salsbury, Grayce and Harris, the functionals depend on the electron density and the magnetic field, and the functional form can depend on the form of the magnetic field. In both of these theories it has been difficult to develop functionals beyond their equivalent to LDA, which are also readily implementable computationally.

            5.4.5. Applications of the DFT

            In practice, Kohn-Sham theory can be applied in several distinct ways depending on what is being investigated. In solid state calculations, the local density approximations are still commonly used along with plane wave basis sets, as an electron gas approach is more appropriate for an infinite solid. In molecular calculations, however, more sophisticated functionals are needed, and a huge variety of exchange-correlation functionals have been developed for chemical applications.

            Some of these are inconsistent with the uniform electron gas approximation, however, they must reduce to LDA in the electron gas limit. Among physicists, probably the most widely used functional is the revised Perdew-Burke-Ernzerhof exchange model (a direct generalized-gradient parametrization of the free electron gas with no free parameters); however, this is not sufficiently calorimetrically accurate for gas-phase molecular calculations. In the chemistry community, one popular functional is known as BLYP (from the name Becke for the exchange part and Lee, Yang and Parr for the correlation part). Even more widely used is B3LYP  which is a hybrid functional in which the exchange energy, in this case from Becke's exchange functional, is combined with the exact energy from Hartree-Fock theory. Along with the component exchange and correlation funñtionals, three parameters define the hybrid functional, specifying how much of the exact exchange is mixed in. The adjustable parameters in hybrid functionals are generally fitted to a 'training set' of molecules. Unfortunately, although the results obtained with these functionals are usually sufficiently accurate for most applications, there is no systematic way of improving them (in contrast to some of the traditional wavefunction-based methods like configuration interaction or coupled cluster theory). Hence in the current DFT approach it is not possible to estimate the error of the calculations without comparing them to other methods or experiments.

            Programs shown below are working by method DFT.

 

 

5.5. Semi-empirical and empirical methods

            Semi-empirical quantum chemistry methods are based on the Hartree-Fock formalism, but make many approximations and obtain some parameters from empirical data. They are very important in computational chemistry for treating large molecules where the full Hartree-Fock method without the approximations is too expensive. The use of empirical parameters appears to allow some inclusion of correlation effects into the methods.

            Semi-empirical methods follow what are often called empirical methods where the two-electron part of the Hamiltonian is not explicitly included. For π-electron systems, this was the Hückel method proposed by Erich Hückel, and for all valence electron systems, the Extended Hückel method proposed by Roald Hoffmann.

            Within the framework of Hartree-Fock calculations, some pieces of information (such as two-elecron integrals) are sometimes approximated or completely omitted. In order to correct for this loss, semi-empirical methods are parametrized, that is their results are fitted by a set of parameters, normally in such a way as to produce results that best agree with experimental data, but sometimes to agree with ab initio results.

            Semi-empirical calculations are much faster than their ab initio counterparts. Their results, however, can be very wrong if the molecule being computed is not similar enough to the molecules in the database used to parametrize the method.

            Semi-empirical calculations have been most successful in the description of organic chemistry, where only a few elements are used extensively and molecules are of moderate size.

            As with empirical methods, we can distinguish methods that are:

-         restricted to π-electrons. These method exist for the calculation of electronically excited states of polyenes, both cyclic and linear. These methods, such as the Pariser-Parr-Pople method (PPP), can provide good estimates of the pi-electronic excited states, when parameterized well. Indeed, for many years, the PPP method outperformed ab initio excited state calculations,

or those:

-         restricted to all valence electrons. These methods can be grouped into several groups:

-         methods such as CNDO/2, INDO and NDDO that were introduced by John Pople. The implementations aimed to fit, not experiment, but ab initio minimum basis set results. These methods are now rarely used but the methodology is often the basis of later methods,

-         methods that are in the Mopac and/or AMPAC computer programs originally from the group of Michel Dewar. These are MINDO, MNDO, AM1, PM3, RM1 and SAM1. Here the objective is to use parameters to fit experimental heats of formation, dipole moments, ionization potentials, and geometries,

-         methods whose primary aim is to predict the geometries of coordination compounds, such as Sparkle/AM1, available for lanthanide complexes,

-         methods whose primary aim is to calculate excited states and hence predict electronic spectra. These include ZINDO and SINDO. the latter being by far the largest group of methods.

            The table below shows some software packages that carry out semi-empirical methods, indicating the other methods that they include where applicable.

Package

Molecular Mechanics

Hartree-Fock

Post-Hartree-Fock methods

Density Functional Theory

AMPAC

N

N

N

N

GAMESS (UK)

N

Y

Y

Y

GAMESS (US)

N

Y

Y

Y

GAUSSIAN

Y

Y

Y

Y

MOLCAS

Y

Y

Y

Y

MOPAC

N

N

N

N

PQS

Y

Y

Y

Y

VASP

N

Y

N

Y

 

       Semiempirical molecular orbital calculations have gone through many stages of refinement and elaboration since Pople’s 1965 papers on CNDO. Programs like PM3, which is widely used in contemporary research, are the cumulative achievement of numerous authors including Michael Dewar (1977), Walter Thiel (1998), James Stewart (1990), and their coworkers.

       5.5.1. Semiempirical calculations

       Semiempirical calculations are set up with the same general structure as a HF calculation in that they have a Hamiltonian and a wave function. Within this framework, certain pieces of information are approximated or completely omitted. Usually, the core electrons are not included in the calculation and only a minimal basis set is used. Also, some of the two-electron integrals are omitted.

       In order to correct for the errors introduced by omitting part of the calculation, the method is parameterized. Parameters to estimate the omitted values are obtained by ®tting the results to experimental data or ab initio calculations.

       Often, these parameters replace some of the integrals that are excluded. The advantage of semiempirical calculations is that they are much faster than ab initio calculations. The disadvantage of semiempirical calculations is that the results can be erratic and fewer properties can be predicted reliably. If the molecule being computed is similar to molecules in the database used to parameterize the method, then the results may be very good. If the molecule being computed is signi®cantly di¨erent from anything in the parameterization set, the answers may be very poor. For example, the carbon atoms in cyclopropane and cubane have considerably di¨erent bond angles from those in most other compounds; thus, these molecules may not be predicted well unless they were included in the parameterization. However, semiempirical methods are not as sensitive to the parameterization set as are molecular mechanics calculations.

            Semiempirical methods are parameterized to reproduce various results. Most often, geometry and energy (usually the heat of formation) are used. Some researchers have extended this by including dipole moments, heats of reaction, and ionization potentials in the parameterization set. A few methods have been parameterized to reproduce a speci®c property, such as electronic spectra or NMR chemical shifts. Semiempirical calculations can be used to compute properties other than those in the parameterization set. Many semiempirical methods compute energies as heats of formation. The researcher should not add zero-point corrections to these energies because the thermodynamic corrections are implicit in the parameterization. CIS calculations from the semiempirical wave function can be used for computing electronic excited states. Some software packages allow CI calculations other than CIS to be performed from the semiempirical reference space.

            This is a good technique for modeling compounds that are not described properly by a single-determinant wave function. Semiempirical CI calculations do not generally improve the accuracy of results since they include correlation twice: once by CI and once by parameterization.

            Semiempirical calculations have been very successful in the description of organic chemistry, where there are only a few elements used extensively and the molecules are of moderate size. Some semiempirical methods have been devised speci®cally for the description of inorganic chemistry as well. The following are some of the most commonly used semiempirical methods.

            5.5.2. Hückel methods

            The Hückel method and is one of the earliest and simplest semiempirical methods. A Hückel  calculation models only the π-valence electrons in a planar conjugated hydrocarbon. A parameter is used to describe the interaction between bonded atoms. There are no second atom affects. Hückel  calculations do re¯ect orbital symmetry and qualitatively predict orbital coefficients. Hückel calculations can give crude quantitative information or qualitative insight into conjugated compounds, but are seldom used today. The primary use of Hückel calculations now is as a class exercise because it is a calculation that can be done by hand.

            5.5.3.  Extended Hückel method

            An extended Hückel calculation is a simple means for modeling the valence orbitals based on the orbital overlaps and experimental electron affinities and ionization potentials. In some of the physics literature, this is referred to as a tight binding calculation. Orbital overlaps

can be obtained from a simplified single STO representation based on the atomic radius. The advantage of extended Hückel calculations over Hückel calculations is that they model all the valence orbitals.

       The primary reason for interest in extended Hückel today is because the method is general enough to use for all the elements in the periodic table. This is not an extremely accurate or sophisticated method; however, it is still used for inorganic modeling due to the scarcity of full periodic table methods with reasonable CPU time requirements. Another current use is for computing band structures, which are extremely computation-intensive calculations. Because of this, extended Hückel is often the method of choice for band structure calculations. It is also a very convenient way to view orbital symmetry. It is known to be fairly poor at predicting molecular geometries.

       The neglect of electron-electron interactions in the Extended Hückel model has several consequences. For example, the atomic orbital binding energies are fixed and do not depend on charge density. With the more accurate NDO semi-empirical treatments, these energies are appropriately sensitive to the surrounding molecular environment.

       5.5.4. Pariser-Parr-Pople ( PPP) method

The Pariser-Parr-Pople ( PPP) method is an extension of the Hückel method that allows heteroatoms other than hydrogen. It is still occasionally used when very minimal amounts of electronic e¨ects are required. For example, PPP-based terms have been incorporated in molecular mechanics calculations to describe aromaticity. This method is also popular for developing simple parameterized analytic expressions for molecular properties.

       5.5.5. CNDO

       The complete neglect of differential overlap (CNDO) method is the simplest of the neglect of differential overlap (NDO) methods. This method models valence orbitals only using a minimal basis set of Slater type orbitals. The CNDO method has proven useful for some hydrocarbon results but little else. CNDO is still sometimes used to generate the initial guess for ab initio calculations on hydrocarbons.

       The electron repulsion between electrons in different orbitals depends only on the nature of the atoms involved, and not on the particular orbital. This creates a very simple picture. One of its disadvantages is that because it neglects almost all exchange integrals, it cannot calculate differences between states of multiplicity arising from the same electronic configuration. CNDO treats singlet-triplet energy gaps poorly.

            Practically all CNDO calculations are actually performed using the CNDO/2 method, which is an improved parameterization over the original CNDO/1 method. There is a CNDO/S method that is parameterized to reproduce electronic spectra. The CNDO/S method does yield improved prediction of excitation energies, but at the expense of the poorer prediction of molecular geometry. There have also been extensions of the CNDO/2 method to include elements with occupied d- orbitals. These techniques have not seen widespread use due to the limited accuracy of results.

            5.5.6. MINDO

There are three modi®ed intermediate neglect of di¨erential overlap (MINDO) methods: MINDO/1, MINDO/2, and MINDO/3. The MINDO/3 method is by far the most reliable of these. This method has yielded qualitative results for organic molecules. However its use today has been superseded by that of more accurate methods such as Austin model 1 (AM1) and parameterization method (PM3). MINDO/3 is still sometimes used to obtain an initial guess for ab initio calculations.

            5.5.7. MNDO

            The modified neglect of diatomic overlap (MNDO) method has been found to give reasonable qualitative results for many organic systems. It has been incorporated into several popular semiempirical programs as well as the MNDO program. Today, it is still used, but the more accurate AM1 and PM3 methods have surpassed it in popularity.

            There are a some known cases where MNDO gives qualitatively or quantitatively incorrect results, for example:

·        computed electronic excitation energies are underestimated,

·        activation barriers tend to be too high,

·        the correct conformer is not always computed to be lowest in energy,

·        barriers to bond rotation are often computed to be too small,

·        hypervalent compounds and sterically crowded molecules are computed to be too unstable,

·        four-membered rings are predicted to be too stable,

·        oxygenated functional groups on aromatic rings are predicted to be out-of-plane,

·        the peroxide bond is too short by about 0.17 Å,

·        the ether C-O-C bond angle is too large by about 9o,

·        bond lengths between electronegative elements are too short,

·        hydrogen bonds are too weak and long.

       A variation on MNDO is MNDO/d. This is an equivalent formulation including d orbitals. This improves predicted geometry of hypervalent molecules. This method is sometimes used for modeling transition metal systems, but its accuracy is highly dependent on the individual system being studied. There is also a MNDOC method that includes electron correlation.

       5.5.8. AM1

       The Austin Model 1 (AM1) method is still popular for modeling organic compounds. AM1 generally predicts the heats of formation (ΔHf) more accurately than MNDO, although a few exceptions involving Br atoms have been documented. Depending on the nature of the system and information desired, either AM1 or PM3 will often give the most accurate results obtainable for organic molecules with semiempirical methods.

       There are some known strengths and limitations in the results obtained from these methods. Activation energies are improved over MNDO. AM1 tends to predict results for aluminum better than PM3. It tends to poorly predict nitrogen paramidalization. AM1 tends to give O-Si-O bonds that are not bent enough. There are some known limitations to AM1 energies, such as predicting rotational barriers to be one-third the actual barrier and predicting five-membered rings to be too stable. The predicted heat of formation tends to be inaccurate for molecules with a large amount of charge localization. Geometries involving phosphorus are predicted poorly. There are systematic errors in alkyl group energies predicting them to be too stable. Nitro groups are too positive in energy. The peroxide bond is too short by about 0.17 Å. Hydrogen bonds are predicted to have the correct strength, but often the wrong orientation. On average, AM1 predicts energies and geometries better than MNDO, but not as well as PM3. Computed bond enthalpies are consistently low.

       5.5.9. PM3

       Parameterization method 3 ( PM3) uses nearly the same equations as the AM1 method along with an improved set of parameters. The PM3 method is also currently extremely

 popular for organic systems. It is more accurate than AM1 for hydrogen bond angles, but AM1 is more accurate for hydrogen bond energies. The PM3 and AM1 methods are also more popular than other semi­empi­rical methods due to the availability of algorithms for including salvation effects in these calculations.

       There are also some known strengths and limitations of PM3. Overall heats of formation are more accurate than with MNDO or AM1. Hypervalent molecules are also predicted more accurately. PM3 tends to predict that the barrier to rotation around the CaN bond in peptides is too low. Bonds between Si and the halide  atoms  are   too  short.  PM3   also   tends  to   predict

incorrect electronic states for germanium compounds. It tends to predict sp 3 nitrogen as always being pyramidal. Some spurious minima are predicted. Proton affnities are not accurate. Some polycyclic rings are not ¯at. The predicted charge on nitrogen is incorrect. Nonbonded distances are too short. Hydrogen bonds are too short by about 0.1 Å, but the orientation is usually correct. On average, PM3 predicts energies and bond lengths more accurately than AM1 or MNDO.

            5.5.10. INDO

            The intermediate neglect of differential overlap (INDO) method was at one time used for organic systems. Today, it has been superseded by more accurate methods. INDO is still sometimes used as an initial guess for ab initio calculations.

            The INDO method corrects some of the worst problems with CNDO. For example, INDO exchange integrals between electrons on the same atom need not be equal, but can depend on the orbitals involved. Though this introduces more parameters, additional computation time is negligible. INDO and MINDO/3 (Modified INDO, version 3) methods are different implementations of the same approximation.

            5.5.11. ZINDO

            The Zerner's INDO method (ZINDO) is also called spectroscopic INDO (INDO/S). This is a reparameterization of the INDO method specifically for the purpose of reproducing electronic spectra results. ZINDO/1 is expected to give geometries of molecules, and ZINDO/S is parametrized to give UV spectra.This method has been found to be useful for predicting electronic spectra. ZINDO is also used for modeling transition metal systems since it is one of the few methods parameterized for metals. It predicts UV transitions well, with the exception of metals with unpaired electrons. However, its use is generally limited to the type of results for which it was parameterized. ZINDO often gives poor results when used for geometry optimization.

            5.5.12. SINDO1

       The symmetrically orthogonalized intermediate neglect of di¨erential overlap method (SINDO1) is both a semiempirical method and a computer program incorporating that method. It is another variation on INDO. SINDO1 is designed for the prediction of the binding energies and geometries of the 1st and 2nd row elements as well as the 3rd row transition metals. Some of the parameters were taken directly from experimental or ab initio results, whereas the rest were parameterized to reproduce geometry and heats of formation. The method was originally designed for modeling ground states of organic molecules. More recently, it has been extended to predict photochemistry and transition metal

results.

       5.6. Molecular mechanics

       In many cases, large molecular systems can be modeled successfully while avoiding quantum mechanical calculations entirely. Molecular mechanics simulations, for example, use a single classical expression for the energy of a compound, for instance the harmonic oscillator. All constants appearing in the equations must be obtained beforehand from experimental data or ab initio calculations.

       The database of compounds used for parameterization - (the resulting set of parameters and functions is called the force field) - is crucial to the success of molecular mechanics calculations. A force field parameterized against a specific class of molecules, for instance proteins, would be expected to only have any relevance when describing other molecules of the same class.

       These methods can be applied to proteins and other large biological molecules, and allow studies of the approach and interaction (docking) of potential drug molecules.

       The term molecular mechanics refers to the use of Newton mechanics to model molecular systems. The potential energy of all systems in molecular mechanics is calculated using force fields. Molecular mechanics can be used to study small molecules as well as large biological systems or material assemblies with many thousands to millions of atoms.

All-atomistic molecular mechanics methods have the following properties:

*        Each atom is simulated as a single particle

*        Each particle is assigned a radius (typically the van der Waals radius), polarizability, and a constant net charge (generally derived from quantum calculations and/or experiment)

*        Bonded interactions are treated as "springs" with an equilibrium distance equal to the experimental or calculated bond length.

       Variations on this theme are possible; for example, many simulations have historically used a "united-atom" representation in which methyl and methylene groups were represented as a single particle, and large protein systems are commonly simulated using a "bead" model that assigns two to four particles per amino acid.

       5.6.1. Functional Form

       The following functional abstraction, known as a potential function or force field in Chemistry, calculates the molecular system's potential energy (E) in a given conformation as a sum of individual energy terms.

            A force field is used to minimize the bond stretching energy of this ethane molecule.

                                                (107)

where the components of the covalent and noncolvalent contributions are given by the following

Fig.28. A force field is used to minimize the bond stretching energy of this ethane molecule

 

summations:

                                           (108)

                              (109)

            The exact functional form of the potential function, or force field, depends on the particular simulation program being used. Generally the bond and angle terms are modeled as harmonic potentials centered around equilibrium bond-length values derived from experiment or theoretical calculations of electronic structure performed with software which does ab-initio type calculations such as Gaussian. For accurate reproduction of vibrational spectra, the Morse potential can be used instead, at computational cost. The dihedral or torsional terms typically have multiple minima and thus cannot be modeled as harmonic oscillators, though their specific functional form varies with the implementation. This class of terms may include "improper" dihedral terms, which function as correction factors for out-of-plane deviations (for example, they can be used to keep benzene benzene rings planar).

            The non-bonded terms are much more computationally costly to calculate in full, since a typical atom is bonded to only a few of its neighbors, but interacts with every other atom in the molecule. Fortunately the van der Waals term falls off rapidly - it is typically modeled using a "6-12 Lennard-Jones potential", which means that attractive forces fall off with distance as r-6 and repulsive forces as r-12, where r represents the distance between two atoms. Generally a cutoff radius is used to speed up the calculation so that atom pairs whose distances are greater than the cutoff have a van der Waals interaction energy of zero.

       The electrostatic terms are notoriously difficult to calculate well because they do not fall off rapidly with distance, and long-range electrostatic interactions are often important features of the system under study (especially for proteins). The basic functional form is the Coulomb potential, which only falls off as r-1. A variety of methods are used to address this problem, the simplest being a cutoff radius similar to that used for the van der Waals terms. However, this introduces a sharp discontinuity between atoms inside and atoms outside the radius. Switching or scaling functions that modulate the apparent electrostatic energy are somewhat more accurate methods that multiply the calculated energy by a smoothly varying scaling factor from 0 to 1 at the outer and inner cutoff radii. Other more sophisticated but computationally intensive methods are known as particle mesh Ewald (PME) and the multipole algorithm.

Common force field terms

       In addition to the functional form of each energy term, a useful energy function must be assigned parameters for force constants, van der Waals multipliers, and other constant terms. These terms, together with the equilibrium bond, angle, and dihedral values, partial charge values, atomic masses and radii, and energy function definitions, are collectively known as a force field. Parameterization is typically done through agreement with experimental values and theoretical calculations results. Each force field is parameterized to be internally consistent, but the parameters are generally not transferable from one force field to another.

       5.6.2. Areas of molecular mechanics application

       The prototypical Molecular Mechanics application is energy minimization. That is, the force field is used as an optimization criterion and the (local) minimum searched by an appropriate algorithm (e.g. steepest descent). Global energy optimization can be

 accomplished using simulated annealing, the Metropolis algorithm and other Monte Carlo methods, or using

 different deterministic methods of discrete or continuous optimization. The main aim of optimization methods is finding the lowest energy conformation of a molecule or identifying a set of low-energy conformers that are in equilibrium with each other. The force field represents only the enthalpic component of free energy, and only this component is included during energy minimization. However, the analysis of equilibrium between different states requires also conformational entropy be included, which is possible but rarely done.

            Other applications of MM include potential energy mapping and ligand docking simulations.

Molecular Mechanics and Molecular Dynamics (MD) are related but different. Main purpose of MD is modeling of molecular motions, although it is also applied for optimization, for example using simulated annealing. MM implements more "static" energy minimization methods to study the potential energy surfaces of different molecular systems. However, MM can also provide important dynamic parameters, such as energy barriers between different conformers or steepness of a potential energy surface around a local minimum. MD and MM are usually based on the same classical force fields. But MD may also be based on quantum chemical methods like DFT. MM is also loosely used to define a set of techniques in molecular modeling.

            5.6.3. Environment and Solvation

            There are several ways of defining the environment surrounding the molecule or molecules of interest in molecular mechanics. A system can be simulated in vacuum (known as a gas-phase simulation) with no surrounding environment at all, but this is usually not desirable because it introduces artifacts in the molecular geometry, especially in charged molecules. Surface charges that would ordinarily interact with solvent molecules instead interact with each other, producing molecular conformations that are unlikely to be present in any other environment.

            The "best" way to solvate a system is to place explicit water molecules in the simulation box with the molecules of interest and treat the water molecules as interacting particles like those in the molecule. A variety of water models exist with increasing levels of complexity, representing water as a simple hard sphere (a united-atom approach), as three separate particles with fixed bond angles, or even as four or five separate interaction centers to account for unpaired electrons on the oxygen atom. Unsurprisingly, the more complex the water model, the more computationally intensive the simulation. A compromise approach has been found in implicit solvation, which replaces the explicitly represented water molecules with a

 mathematical expression that reproduces the average behavior of water molecules (or other solvents such as lipids). This method is useful for preventing artifacts that arise from vacuum simulations and reproduces bulk solvent properties well, but cannot reproduce situations in which individual water molecules have interesting interactions with the molecules under study.

       Software Packages of the Molecular Mechanics involve next limited list of programs: AMBER, CHARMM , Ghemical , GROMOS , GROMACS , NAMD , STR3DI32.

       5.6.4. Methods for solids

       Computational chemical methods can be applied to solid state physics problems. The electronic structure of a crystal is in general described by a band structure, which defines the energies of electron orbitals for each point in the Brillouin zone. Ab initio and semi-empirical calculations yield orbital energies, therefore they can be applied to band structure calculations. Since it is time-consuming to calculate the energy for a molecule, it is even more time-consuming to calculate them for the entire list of points in the Brillouin zone.

       Computational chemical methods in solid state physics follow the same approach as they do for molecules, but with two differences. First, the translational symmetry of the solid has to be utilised, and second, it is possible to use completely delocalised basis functions such as plane waves as an alternative to the molecular atom-centered basis functions. The electronic structure of a crystal is in general described by a band structure, which defines the energies of electron orbitals for each point in the Brillouin zone. Ab initio and semi-empirical calculations yield orbital energies, therefore they can be applied to band structure calculations. Since it is time-consuming to calculate the energy for a molecule, it is even more time-consuming to calculate them for the entire list of points in the Brillouin zone.

       Calculations can use the Hartree-Fock method, some post-Hartree-Fock methods, particularly Møller-Plesset perturbation theory to second order (MP2) and Density functional theory (DFT).  The following computer packages can do these calculations:

Package

Semi-Empirical

Hartree-Fock

Post-Hartree-Fock methods

Density Functional Theory

ADF

N

N

N

Y

Atomistix Toolkit

N

N

N

Y

CASTEP

N

N

N

Y

CRYSTAL

N

Y

N

Y

GAUSSIAN

Y

Y

Y

Y

LMTO

N

N

N

Y

MOPAC

Y

N

N

N

NWChem

N

Y

Y

Y

PLATO

N

N

N

Y

VASP

N

N

N

Y

WIEN (FLAPW)

N

N

N

Y

           

            5.7. Chemical dynamics

            Once the electronic and nuclear variables are separated (within the Born-Oppenheimer representation), in the time-dependent approach, the wave packet corresponding to the nuclear degrees of freedom is propagated via the time evolution operator (physics) associated to the time-dependent Schrödinger equation (for the full molecular Hamiltonian). In the complementary energy-dependent approach, the time-independent Schrödinger equation is solved using the scattering theory formalism. The potential representing the interatomic interaction is given by the potential energy surfaces. In general, the potential energy surfaces are coupled via the vibronic coupling terms.

            The most popular methods for propagating the wave packet associated to the molecular geometry are:

-                     the split operator technique,

-                     the Multi-Configuration Time-Dependent Hartree method (MCTDH),

-                     the semiclassical method.

            Molecular dynamics (MD) examines (using Newton's laws of motion) the time-dependent behavior of systems, including vibrations or Brownian motion, using a classical mechanical description. MD combined with density functional theory leads to the Car-Parrinello method.

            5.7.1.  Adiabatic chemical dynamics

            In adiabatic dynamics, interatomic interactions are represented by single scalar potentials called potential energy surfaces. This is the Born-Oppenheimer approximation introduced by Born and Oppenheimer in 1927. Pioneering applications of this in chemistry were performed by Rice and Ramsperger in 1927 and Kassel in 1928, and generalized into the RRKM theory in 1952 by Marcus who took the transition state theory developed by Eyring in 1935 into account. These methods enable simple estimates of unimolecular reaction rates from a few characteristics of the potential surface.

            5.7.2. Non-adiabatic chemical dynamics

       Non-adiabatic dynamics consists of taking the interaction between several coupled potential energy surface (corresponding to different electronic quantum states of the molecule). The coupling terms are called vibronic couplings. The pioneering work in this field was done by Stueckelberg, Landau, and Zener in the 1930s, in their work on what is now known as the Landau-Zener transition. Their formula allows the transition probability between two diabatic potential curves in the neighborhood of an avoided crossing to be calculated.

       In theoretical chemistry, the vibronic coupling terms (which are neglected within the Born-Oppenheimer approximation) are proportional to the interaction between electronic and nuclear motions of molecules. The term "vibronic" originates from the concatenation of the terms "vibrational" and "electronic". The word coupling denotes the idea that in a molecule, vibrational and electronic interactions are interrelated and influence each other.

       Vibronic coupling is large in the case two adiabatic potential energy surfaces come close to each other, i.e. when the energy gap between them is of the order of magnitude of one oscillation quantum. This usually happens in the neighbourhood of an avoided crossing of potential energy surfaces corresponding to distinct electronic states of the same spatial and spin symmetry. However, vibronic coupling also exists at real crossings. In this case the adiabatic or Born-Oppenheimer approximation fails and non-adiabatic terms (the so-called vibronic coupling terms) have to be taken into account. The vibronic coupling terms are usually difficult to evaluate. This is due to the fact that they are proportional to the first and second derivatives of the electronic wave function with respect to the molecular coordinates. A simpler way to solve this problem is to switch from the adiabatic to the diabatic representation of the potential energy surfaces. The vibronic terms are responsible for example for surface hopping or the Berry phase. The Berry phase has been discovered by Longuet-Higgins in this context. The vibronic coupling becomes infinite in the neighbourhood of a conical intersection. This singularity in the potential energy landscape is the origin of the Berry phase.

       Perhaps the earliest examples of the importance of vibronic coupling were found during the 1930s. In 1934 Renner wrote about the vibronic coupling in an electronically excited Π-state in CO2. Calculations of the lower excited levels of benzene by Sklar in 1937 (with the valence bond method) and later in 1938 by Goeppert-Mayer and Sklar (with the molecular orbital method) demonstrated a correspondence between the theoretical predictions and experimental results of the benzene spectrum. The benzene spectrum was the first qualitative computation of the efficiencies of various vibrations at inducing intensity absorption.

       5.8. Quantum chemistry and quantum field theory

            The application of quantum field theory (QFT) to chemical systems and theories has become increasingly common in the modern physical sciences. One of the first and most fundamentally explicit appearances of this is seen in the theory of the photomagneton. In this system, plasmas, which are ubiquitous in both physics and chemistry, are studied in order to determine the basic quantization of the underlying bosonic field. However, quantum field theory is of interest in many fields of chemistry, including: nuclear chemistry, astrochemistry, sonochemistry, and quantum hydrodynamics. Field theoretic methods have also been critical in developing the ab initio Effective Hamiltonian theory of semi-empirical pi-electron methods.

            Chapter 6. Basis set

            A basis set in chemistry is a set of functions used to create the molecular orbitals, which are expanded as a linear combination of such functions with the weights or coefficients to be determined. Usually these functions are atomic orbitals, in that they are centered on atoms, but functions centered in bonds or lone pairs have been used as have pairs of functions centered in the two lobes of a p orbital. Additionally, basis sets composed of sets of plane waves up to a cutoff wavelength are often used, especially in calculations involving systems with periodic boundary conditions.

            In modern computational chemistry, quantum chemical calculations are typically performed within a finite set of basis functions. In these cases, the wavefunctions under consideration are all represented as vectors, the components of which correspond to coefficients in a linear combination of the basis functions in the basis set used. The operators are then represented as matrices, (rank two tensors), in this finite basis. In this article, basis function and atomic orbital are sometimes used interchangeably, although it should be noted that these basis functions are usually not actually the exact atomic orbitals, even for the corresponding hydrogen-like atoms, due to approximations and simplifications of their analytic formulas. If the finite basis is expanded towards an infinite complete set of functions, calculations using such a basis set are said to approach the basis set limit.

            When molecular calculations are performed, it is common to use a basis composed of a finite number of atomic orbitals, centered at each atomic nucleus within the molecule (linear combination of atomic orbitals ansatz). Initially, these atomic orbitals were typically Slater orbitals, which corresponded to a set of functions which decayed exponentially with distance from the nuclei. Later, it was realized by Frank Boys that these Slater-type orbitals could in turn be approximated as linear combinations of Gaussian orbitals instead. Because it is easier to

calculate overlap and other integrals with Gaussian basis functions, this led to huge computational savings.

       Today, there are hundreds of basis sets composed of Gaussian-type orbitals (GTOs). The smallest of these are called minimal basis sets, and they are typically composed of the minimum number of basis functions required to represent all of the electrons on each atom. The largest of these can contain literally dozens to hundreds of basis functions on each atom.

A minimum basis set is one in which, on each atom in the molecule, a single basis function is used for each orbital in a Hartree-Fock calculation on the free atom. However, for atoms such as lithium, basis functions of p type are added to the basis functions corresponding to the 1s and 2s orbitals of the free atom. For example, each atom in the first row of the periodic system (Li - Ne) would have a basis set of five functions (two s functions and three p functions).

The most common addition to minimal basis sets is probably the addition of polarization functions, denoted by an asterisk, *. Two asterisks, **, indicate that polarization functions are also added to light atoms (hydrogen and helium). These are auxiliary functions with one additional node. For example, the only basis function located on a hydrogen atom in a minimal basis set would be a function approximating the 1s atomic orbital. When polarization is added to this basis set, a p-function is also added to the basis set. This adds some additional needed flexibility within the basis set, effectively allowing molecular orbitals involving the hydrogen atoms to be more asymmetric about the hydrogen nucleus. This is an important result when considering accurate representations of bonding between atoms, because the very presence of the bonded atom makes the energetic environment of the electrons spherically asymmetric. Similarly, d-type functions can be added to a basis set with valence p orbitals, and f-functions to a basis set with d-type orbitals, and so on. Another, more precise notation indicates exactly which and how many functions are added to the basis set, such as (p, d).

Another common addition to basis sets is the addition of diffuse functions, denoted by a plus sign, +. Two plus signs indicate that diffuse functions are also added to light atoms (hydrogen and helium). These are very shallow Gaussian basis functions, which more accurately represent the "tail" portion of the atomic orbitals, which are distant from the atomic nuclei. These additional basis functions can be important when considering anions and other large, "soft" molecular systems.

       6.1. Minimal basis sets

       The most common minimal basis set is STO-nG, where n is an integer. This n value represents the number of Gaussian primitive functions comprising a single basis function. In these basis sets, the same number of Gaussian primitives comprise core and valence orbitals.

 Minimal basis sets typically give rough results that are insufficient for research-quality publication, but are much cheaper than their larger counterparts. Commonly used minimal basis sets of this type are:

·                     STO-3G

·                     STO-4G

·                     STO-6G

·                     STO-3G* - Polarized version of STO-3G

There are several other minimum basis sets that have been used such as the MidiX basis sets.

            6.2. Split-valence basis sets

            During most molecular bonding, it is the valence electrons which principally take part in the bonding. In recognition of this fact, it is common to represent valence orbitals by more than one basis function, (each of which can in turn be composed of a fixed linear combination of primitive Gaussian functions). Basis sets in which there are multiple basis functions corresponding to each valence atomic orbital, are called valence double, triple, or quadruple-zeta basis sets. Since the different orbitals of the split have different spatial extents, the combination allows the electron density in adjust its spatial extend appropriate to the particular molecular environment. Minimum basis sets are fixed and are unable to adjust to different molecular environments. Basis sets in which there are multiple basis functions corresponding to each atomic orbital, including both valence orbitals and core orbitals are called double, triple, or quadruple-zeta basis sets.

            6.3. Pople basis sets

            The notation for the split-valence basis sets arising from the group of John Pople is typically X-YZg. In this case, X represents the number of primitive Gaussians comprising each core atomic orbital basis function. The Y and Z indicate that the valence orbitals are composed of two basis functions each, the first one composed of a linear combination of Y primitive Gaussian functions, the other composed of a linear combination of Z primitive Gaussian functions. In this case, the presence of two numbers after the hyphens implies that this basis set is a split-valence double-zeta basis set. Split-valence triple- and quadruple-zeta basis sets are also used, denoted as X-YZWg, X-YZWVg, etc. Here is a list of commonly used split-valence basis sets of this type:

·         3-21g

·         3-21g* - Polarized

·         3-21+g - Diffuse functions

·         3-21+g* - With polarization and diffuse functions

·         6-31g

·       6-31g*

·       6-31+g*

·       6-31g(3df, 3pd)

·       6-311g

·       6-311g*

·       6-311+g*

       6.4. Correlation consistent basis sets

       Some of the most widely used basis sets are those developed by Dunning and coworkers, since they are designed to converge systematically to the complete basis set (CBS) limit using extrapolation techniques. For first- and second-row atoms, the basis sets are cc-pVNZ where N=D,T,Q,5,6,... (D=double, T=triples, etc.). The 'cc-p', stands for 'correlation consistent polarized' and the 'V' indicates they are valence only basis sets. They include successively larger shells of polarization (correlating) functions (d, f, g, etc.). More recently these 'correlation consistent polarized' basis sets have become widely used and are the current state of the art for correlated or post Hartree-Fock calculations. Examples of these are:

·                     cc-pVDZ - Double-zeta

·                     cc-pVTZ - Triple-zeta

·                     cc-pVQZ - Quadruple-zeta

·                     cc-pV5Z - Quintuple-zeta, etc.

·                     aug-cc-pVDZ, etc. - Augmented versions of the preceding basis sets with added diffuse functions

       For third-row atoms, additional functions are necessary; these are the cc-pV(N+d)Z basis sets. Even large atoms require the cc-pVNZ-PP and cc-pVNZ-DK families of basis sets, where PP and DK stand for pseudopotential and Douglas-Kroll, respectively.

       These basis sets can be augmented with core functions for geometric and nuclear property calculations, and with diffuse functions for electronic excited-state calculations, electric field property calculations, and long-range interactions, such as van der Waals forces. A recipe for constructing additional augmented functions exists; as many as five augmented functions have been used in second hyperpolarizability calculations in the literature. Because of the rigorous construction of these basis sets, extrapolation can be done for almost any property.

       6.5. Other split valence basis sets

Other split valence basis sets often have rather generic names such as:

  • SV(P)
  • SVP
  • DZV
  • TZV
  • TZVPP - Valence triple-zeta plus polarisation
  • QZVPP - Valence quadruple-zeta plus polarisation

            6.6. Plane wave basis sets

            In addition to localized basis sets, plane wave basis sets can also be used in quantum chemical simulations. Typically, a finite number of plane wave functions are used, below a specific cutoff energy which is chosen for a certain calculation. These basis sets are popular in calculations involving periodic boundary conditions. Certain integrals and operations are much easier to code and carry out with plane wave basis functions, than with their localized counterparts. In practice, plane wave basis sets are often used in combination with an 'effective core potential' or pseudopotential, so that the plane waves are only used to describe the valence charge density. This is because core electrons tend to be concentrated very close to the atomic nuclei, resulting in large wavefunction and density gradients near the nuclei which are not easily described by a plane wave basis set unless a very high energy cutoff, (and therefore small plane wavelength), is used. This combined method of a plane wave basis set with a core pseudopotential is often abbreviated as a PSPW calculation. Furthermore, as all functions in the basis are mutually orthogonal, plane wave basis sets do not exhibit basis set superposition error. However, they are less well suited to gas-phase calculations. Using Fast Fourier Transforms, one can work with plane wave basis sets in reciprocal space in which not only the aforementioned integrals, such as the kinetic energy, but also derivatives are computationally less demanding to be carried out. Another important advantage of a plane wave basis is that it is guaranteed to converge to the target wave function while there is no such guarantee for Gaussian type basis sets.

            Chapter 7.  Practical calculation of molecular systems. HyperChem program

            There are many Quantum calculation programs for studying the electron and geometrical structure of molecules. The information on some programs it has been given above. Program   HyperChem is a versatile molecular modeler and editor and a powerful computational package. It offers many types of molecular and quantum mechanics calculations.

            7.1. Building and Displaying Molecules

            Submitting a structure to a calculation can be expensive in terms of human time and effort. HyperChem lets you build and display molecules easily. Since HyperChem contains a graphical interface, you can monitor the construction of molecules. Using the Drawing tool, you can draw a two-dimensional (2D) representation of a molecule, and then use the Model Builder to generate a three-dimensional (3D) structure. The Model Builder adds implicit hydrogens to the molecule at your request. You can also manipulate individual bonds, bond geometries, angles, torsions, and atomic charges during model building (using the Constrain commands on the Build menu) or after model building (using the Set commands on the Edit menu).

            HyperChem contains a database of amino and nucleic acid residues so you can quickly build polymers containing these subunits. You can also read in structures in files from standard databases,

such as the Brookhaven Protein Data Bank (see the HyperChem Reference Manual).

            7.2. Optimizing the Structures of Molecules

            To calculate the properties of a molecule, you need to generate a well-defined structure. A calculation often requires a structure that represents a minimum on a potential energy surface. HyperChem contains several geometry optimizers to do this. You can then calculate single point properties of a molecule or use the optimized structure as a starting point for subsequent calculations, such as molecular dynamics simulations.

            7.3. Investigating the Reactivity of Molecules

            You can use HyperChem to investigate the reactivity of molecules and their functional groups. One method is to use Frontier Molecular Orbital Theory. You can use molecular orbital energies, coefficients, and nodal properties from single point quantum mechanics calculations to investigate issues such as the relative reactivity of different molecular substituents, regioselectivity of reactions, and site-selectivity of nucleophiles and electrophiles.

            7.4. Generating and Viewing Orbitals and Electronic Plots

            HyperChem can plot orbital wave functions resulting from semiempirical and ab initio quantum mechanical calculations. It is interesting to view both the nodal properties and the relative sizes of the wave functions. Orbital wave functions can provide chemical insights. You can also plot the electrostatic potential, the total charge density, or the total spin density determined during a semi-empirical or ab initio calculation. This information is useful in determining reactivity and correlating calculational results with experimental data. These examples illustrate uses of these plots:

• If you calculate the electrostatic potential for cyclopropane, three minima occur in regions that bisect the carbon-carbon bonds. This result is consistent with protonation of cyclopropane occurring along the bond bisector.

• Spin densities help to predict the observed coupling constants in electron spin resonance (ESR) spectroscopy. From spin density plots you can predict a direct relationship between the spin density on a carbon atom and the coupling constant associated with an adjacent hydrogen.

            7.5. Evaluating Chemical Pathways and Mechanisms

            Calculating single point properties and energies provides information about chemical pathways and mechanisms.

Example: Researchers have used MNDO and AM1 semi-empirical methods to calculate possible reaction pathways for the interaction of glycine and cocaine. In choosing possible interaction sites, they first calculated atomic charges and HOMO-LUMO (highest occupied and lowest unoccupied molecular orbitals) electron densities for the two molecules. Using this information, they proposed two possible reaction pathways and computed the heats of formation for geometry-optimized structures.

            7.6. Studying the Dynamic Behavior of Molecules

            A molecular system at room temperature is accurately characterized by its motion. Molecular dynamics simulations calculate the future positions and velocities of atoms based upon their current values. You can obtain qualitative and quantitative data from HyperChem molecular dynamics simulations.

            The greatest value of molecular dynamic simulations is that they complement and help to explain existing data for designing new experiments. The simulations are increasingly useful for structural refinement of models generated from NMR, distance geometry, and X-ray data. The results are also important for investigating macromolecules, where concerted motions may occur during ligand binding, enzymatic activity, and molecular recognition. Using constant temperature molecular dynamics, you can investigate an enzyme that has a loop of residues flanking the active site. You can store snapshots of a molecular dynamic trajectory (the set of positions and velocities of atoms over time) and use them to detect possible conformational changes of the loop region that could change the interaction with the active site. You can further investigate conformational change during the course of the simulation by averaging and plotting the dihedral angle values associated with the loop.

            7.7. HyperChem Calculations of Potential Energy Surfaces

            HyperChem calculations examine potential energy surfaces with single point calculations, optimizations, and molecular dynamic simulations. Energies and derivatives of energy, such as the forces on atoms, are necessary to “construct” a potential energy surface. This concept gives the molecular and quantum mechanics methods a unifying purpose: they generate the energies and energy derivatives necessary to produce and examine potential energy surfaces.

            Both molecular and quantum mechanics methods rely on the Born-Oppenheimer approximation. In quantum mechanics, the Schrödinger equation (1) gives the wave functions and energies of a molecule.

HΨ=EΨ                                                                  (110)

where H is the molecular Hamiltonian, Y is the wave function, and E is the energy. The molecular Hamiltonian is composed of these operators: the kinetic energy of the nuclei (N) and electrons (E), nuclear-nuclear (NN) and electron-electron repulsion (EE), and the attraction between nuclei and electrons (NE) (equation 110).

H = (kinetic energy)N + (kinetic energy)E + (repulsion)NN + (repulsion)EE + (attrction)NE   (111)

Nuclei have many times more mass than electrons. During a very small period of time when the movement of heavy nuclei is negligible, electrons are moving so fast that their distribution is smooth. This leads to the approximation that the electron distribution is dependent only on the fixed positions of nuclei and not on their velocities. This approximation allows two simplifications of the molecular Hamiltonian. The nuclear kinetic energy term drops out (equation 112).

H = (kinetic energy)E + (repulsion)NN +  (repulsion)EE + (attraction)NE                    (112)

Since the nuclear-nuclear repulsion is constant for a fixed configuration of atoms, this term also drops out. The Hamiltonian is now purely electronic.

Helectronic = (kinetic)E + (repulsion)EE + (attraction)NE                             (113)

After solving the electronic Schrödinger equation (equation 113), to calculate a potential energy surface, you must add back nuclear-nuclear repulsions (equation 114).

HelectronicΨelectronic = EelectronicΨelectronic

VPES = Eelectronic + (repulsion)NN                                                 (114)

            Generating the potential energy surface (PES) using this equation requires solutions for many configurations of nuclei. In molecular mechanics, the electronic energy is not evaluated explicitly. Instead, these methods solve the potential energy surface by using a force field equation (see “Molecular Mechanics”). The force field equation represents electronic energy implicitly through parameterization.

            The dimensionality of a potential energy surface depends on the number of degrees of freedom in a molecule. If VPES is a function of two variables, then a plot of the potential energy surface represents a 3D space.

Example: In this plot of a potential energy surface with two variables, points A and C are both minima.

Fig. 29.The path of minimum potential energy that connects reactants and products

 

            This means that the forces on the atoms are zero for the structures found at A and C. The structure at C is more stable than the structure at A. In fact, C is the global minimum for this example. Geometry optimizations seek minima such as A and C.

            Point B is a maximum along the path from A to C (saddle point). The forces on the atoms are also zero for this structure. Point B represents a transition state for the transformation of A to C.

            Convergence of a geometry optimization calculation does not necessarily mean that the calculation found a minimum on the potential energy surface. To categorize a point definitively on the potential energy surface, you must calculate the second derivatives of the potential energy with respect to atomic coordinates. You can then calculate vibrational frequencies from the second derivative matrix. A minimum configuration has six vibrational frequencies with a value of zero. For linear molecules the minimum is five with zero values. A transition state has one vibrational frequency, and this frequency is an imaginary number. The path of minimum potential energy that connects reactants and products is known as the reaction coordinate.

Example: The reaction coordinate for rotation about the central carbon-carbon bond in n-butane has several stationary points. A,C, E, and G are minima and B, D, and F are maxima. Only the structures at the minima represent stable species and of these, the anti conformation is more stable than the gauche.

Fig.30. Conformations of  propane and curve of potential energy

            7.8. Complexity of Potential Energy Surfaces

            Most potential energy surfaces are extremely complex. Elber and Karplus analyzed a 300 psec molecular dynamics trajectory of the protein myoglobin5. They estimate that 2000 thermally accessible minima exist near the native protein structure. The total number of conformations is even larger. Dill derived a formula to calculate the upper bound of thermally accessible conformations in a protein.6 Using this formula, a protein of 150 residues (the approximate size of myoglobin) has an upper limit of 4 x 1034 conformations for the backbone atoms. Here are two of the consequences of

this complexity:

• A single stable structure may not adequately represent the properties of a molecule. You should investigate a Boltzmann distribution of thermally accessible potential structures.

• For all but the smallest of molecules, a search for the global minimum may be impossible. United atom force fields are sometimes used for biomolecules to decrease the number of nonbonded interactions and the computation time. Another reason for using a simplified potential is to reduce the dimensionality of the potential energy surface. This, in turn, allows for more samples of the surface.

Example: Crippen and Snow reported their success in developing a simplified potential for protein folding. In their model, single points represent amino acids. For the avian pancreatic polypeptide, the native structure is not at a potential minimum. However, a global search found that the most stable potential minimum had only a 1.8 Ångstrom root-mean-square deviation from the native structure.

Example: Solvation can have a profound effect on the potential energy profile for a reaction. Jorgensen’s research group provided important insights into the role of solvation 8. Consider the nucleophilic addition of the hydroxide anion to formaldehyde:

*      Ab initio, gas phase calculations found two minima and one maximum for the reaction coordinate, leading to an exothermic formation of the tetrahedral complex. One minimum is an ion-dipole complex at a H2(O)C–OH– distance of 2.74 Ångstroms. The second is the tetrahedral complex at a H2(O)C–OH– distance of 1.47 Ångstroms. A potential-of-mean-force calculation was performed using a Monte Carlo simulation of the reactants in TIP4P water. The reaction coordinate showed no ion-dipole complex. In contrast to the gas phase study, the tetrahedral intermediate was 10–14 kcal/mol less stable than the reactants and there was a 24–28 kcal/mol activation barrier to the formation of the tetrahedral intermediate.

            7.9. Types of Calculations

            There are three types of calculations in HyperChem: single point, geometry optimization or minimization, and molecular dynamics.

            7.9.1. Single Point

            A single point calculation gives the static properties of a molecule. The properties include potential energy, derivatives of the potential energy, electrostatic potential, molecular orbital energies, and the coefficients of molecular orbitals for ground or excited states. The input molecular structure for a single point calculation usually reflects the coordinates of a stationary point on the potential energy surface, typically a minimum or transition state. Use a geometry optimization to prepare a molecule for a single point calculation.

            One type of single point calculation that of calculating vibrational properties is distinguished as a vibrations calculation in HyperChem. A vibrations calculation predicts fundamental vibrational frequencies, infrared absorption intensities, and normal modes for a geometry optimized molecular structure.

            You can use a single point calculation that determines energies for ground and excited states, using configuration interaction, to predict frequencies and intensities of an electronic ultraviolet-visible spectrum.

            7.9.2. Geometry Optimization

            To carry out a geometry optimization (minimization), HyperChem starts with a set of Cartesian coordinates for a molecule and tries to find a new set of coordinates with a minimum potential energy. You should appreciate that the potential energy surface is very complex, even for a molecule containing only a few dihedral angles.

            Since minimization calculations cannot cross or penetrate potential energy barriers, the molecular structure found during an optimization may be a local and not a global minimum. The minimum represents the potential energy closest to the starting structure of a molecule. Researchers frequently use minimizations to generate a structure at a stationary point for a subsequent single point calculation or to remove excessive strain in a molecule, preparing it for a molecular dynamics simulation.

            7.9.3. Transition State Search

            As mentioned earlier, a potential energy surface may contain saddle points; that is, stationary points where there are one or more directions in which the energy is at a maximum. A saddle point with one negative eigenvalue corresponds to a transition structure for a chemical reaction of changing isomeric form. Transition structures also exist for reactions involving separated species, for example, in a bimolecular reaction

A + B → C + D.                                                                (115)

            Activation energy, i.e., the energy of the transition structure relative to reactants, can be observed experimentally. However, the only way that the geometries of transition structures can be evaluated is from theory. Theory also can give energetics and geometry parameters of short-lived reaction intermediates.

            Transition state search algorithms rather climb up the potential energy surface, unlike geometry optimization routines where an energy minimum is searched for. The characterization of even a simple reaction potential surface may result in location of more than one transition structure, and is likely to require many more individual calculations than are necessary to obtain equilibrium geometries for either reactant or product.

            Calculated transition structures may be very sensitive to the level of theory employed. Semi-empirical methods, since they are parametrized for energy minimum structures, may be less appropriate for transition state searching than ab initio methods are.

            Transition structures are normally characterized by weak “partial” bonds, that is, being broken or formed. In these cases UHF calculations are necessary, and sometimes even the inclusion of electron correlation effects.

            7.9.4. Molecular Dynamics

            A molecular dynamics simulation samples phase space (the region defined by atomic positions and velocities) by integrating numerically Newton’s equations of motion. Unlike single point and geometry optimization calculations, molecular dynamics calculation account for thermal motion. Molecules may contain enough thermal energy to cross potential barriers. Molecular dynamics calculations provide information about possible conformations, thermodynamic properties, and dynamic behavior of molecules.

            7.10. Langevin Dynamics

            Molecules in solution undergo collisions with other molecules and experience frictional forces as they move through the solvent. Using Langevin dynamics, you can model these effects and study the dynamical behavior of a molecular system in a liquid environment. These simulations can be much faster than molecular dynamics. Instead of explicitly including solvent molecules, the Langevin method models their effect through random forces applied to the molecule of interest to simulate collisions, and frictional forces are added to model the energy losses associated with these collisions.

            Langevin dynamics simulations can be used to study the same kinds of problems as molecular dynamics: time dependent properties of solvated systems at non-zero temperatures. Because of the implicit treatment of the solvent, this method is particularly wellsuited for studying large molecules in solution. It is possible to decouple the time scales of molecular motion and focus on just the slow modes associated with conformational changes, for example, or to follow the rapid bond stretching motions leading to chemical reaction, while incorporating the influence of molecular collisions.

            Langevin dynamics has been use to study solvent effects on the structural dynamics of the active site of the enzyme lysozyme, conformational equilibria and the dynamics of conformational

transitions in liquid alkanes, and temperature effects on a system of interacting quantum harmonic oscillators.

            7.11. Monte Carlo Simulations

            The Monte Carlo method samples phase space by generating random configurations from a Boltzmann distribution at a given temperature. Averages computed from a properly equilibrated Monte Carlo simulation correspond to thermodynamic ensemble averages. Thus, the Monte Carlo method can be used to find average energies and equilibrium structural properties of complex interacting systems.

            A sequence of successive configurations from a Monte Carlo simulation constitutes a trajectory in phase space; with HyperChem, this trajectory may be saved and played back in the same way as a dynamics trajectory. With appropriate choices of setup parameters, the Monte Carlo method may achieve equilibration more rapidly than molecular dynamics. For some systems, then, Monte-Carlo provides a more direct route to equilibrium structural and thermodynamic properties. However, these calculations can be quite long, depending upon the system studied.

            Monte Carlo simulations provide an alternate approach to the generation of stable conformations. As with HyperChem’s other simulation methods, the effects of temperature changes and salvation are easily incorporated into the calculations.

            7.12. Calculation Methods

            HyperChem uses two types of methods in calculations: molecular mechanics and quantum mechanics. The quantum mechanics methods implemented in HyperChem include semi-empirical quantum mechanics method and ab initio quantum mechanics method. The molecular mechanics and semi-empirical quantum mechanics methods have several advantages over ab initio methods. Most importantly, these methods are fast. While this may not be important for small molecules, it is certainly important for biomolecules. Another advantage is that for specific and well-parameterized molecular systems, these methods can calculate values that are closer to experiment than lower level ab initio techniques.

            The accuracy of a molecular mechanics or semi-empirical quantum mechanics method depends on the database used to parameterize the method. This is true for the type of molecules and the physical and chemical data in the database. Frequently, these methods give the best results for a limited class of molecules or phenomena. A disadvantage of these methods is that you must have parameters available before running a calculation. Developing parameters is time-consuming.

            The ab initio method may overcome this problem. However it is slower than any molecular mechanics and semi-empirical methods.

            7.13. Molecular Mechanics

            Molecular mechanical force fields, as it described above, use the equations of classical mechanics to describe the potential energy surfaces and physical properties of molecules. A molecule is described as a collection of atoms that interact with each other by simple analytical functions. This description is called a force field. One component of a force field is the energy arising from compression and stretching a bond.

            This component is often approximated as a harmonic oscillator and can be calculated using Hooke’s law.

                                                                         (116)

            The bonding between two atoms is analogous to a spring connecting two masses. Using this analogy, equation 7 gives the potential energy of the system of masses, Vspring, and the force constant of the spring, Kr. The equilibrium and displaced distances of the atoms in a bond are r0 and r. Both Kr and r0 are constants for a specific pair of atoms connected by a certain spring. Kr and r0 are force field parameters.

            The potential energy of a molecular system in a force field is the sum of individual components of the potential, such as bond, angle, and van der Waals potentials (equation 117). The energies of the individual bonding components (bonds, angles, and dihedrals) are functions of the deviation of a molecule from a hypothetical compound that has bonded interactions at minimum values.

ETotal = term1 + term2 + …. + termn                                                              (117)

            The absolute energy of a molecule in molecular mechanics has no intrinsic physical meaning; ETotal values are useful only for comparisons between molecules. Energies from single point calculations are related to the enthalpies of the molecules. However, they are not enthalpies because thermal motion and temperature dependent contributions are absent from the energy terms (equation 8).

            Unlike quantum mechanics, molecular mechanics does not treat electrons explicitly. Molecular mechanics calculations cannot describe bond formation, bond breaking, or systems in which electronic delocalization or molecular orbital interactions play a major role in determining geometry or properties. This discussion focuses on the individual components of a typical molecular mechanics force field. It illustrates the mathematical functions used, why those functions are chosen, and the circumstances under which the functions become poor approximations.

            Manual of the HyperChem. Computational Chemistry includes details on the implementation of the MM+, AMBER, BIO+, and OPLS force fields in HyperChem.  

            Chapter 8. Calculation by HyperChem Program

            This section provides an overview and review of quantum mechanics calculations. The information can help you use HyperChem to solve practical problems. For quantitative details of quantum mechanics calculations and how HyperChem implements them, see the second part of HyperChem Manual, Theory and Methods. Ab initio quantum mechanics methods have evolved for many decades. The speed and accuracy of ab initio calculations have been greatly improved by developing new algorithms and introducing better basis functions. Semi-empirical quantum mechanics methods have evolved over the last three decades. Using today’s microcomputers, they can produce meaningful, often quantitative, results for large molecular systems. The roots of the methods lie in the theory of π-electrons, now largely superseded by all-valence electron theories.

            Molecules consist of electrons and nuclei. Most applications of quantum chemistry separate the motion of the nuclei from the motion of electrons (the Born-Oppenheimer approximation). The approximation results in a model of nuclei moving on a potential energy surface, with electrons adjusting instantly to changes in nuclear positions. Nuclear motion is constrained by the interaction of nuclei and electrons. At any fixed positions of the nuclei, potential energy is the sum of repulsions among the positively charged nuclei and attractions arising from the electrons. Electrons are the “glue” holding nuclei together.

            8.1. Molecular Geometry

            The most important chemical structures are stable, equilibrium molecular geometries and transition states. Equilibrium geometry of a molecule (bond lengths and angles) describes the coordinates of a deep minimum on a potential energy surface. A set of atoms may have a number of potential energy minima, each corresponding to a different isomer of the molecular system. Other, less-deep minima may correspond to intermediates in chemical reactions. Saddle points serve as models for transition states of chemical reactions. The potential energy for saddle points is a maximum along one direction, called the reaction coordinate, and a minimum along all others. You can think of saddle points as mountain passes: they are the lowest barriers separating reactant and product potential energy valleys. For a discussion of the potential energy surface, see “Exploring Potential Energy Surfaces” on 6.2.1 of Manual.

            Using the coordinates of special geometries, minima, and saddle points, together with the nearby values of potential energy, you can calculate spectroscopic properties and macroscopic thermodynamic and kinetic parameters, such as enthalpies, entropies, and thermal rate constants. HyperChem can provide the geometries and energy values for many of these calculations.

            8.2. Calculating Electronic Potential Energy

            How is electronic potential energy computed? Electrons, which are more than three orders of magnitude lighter than nuclei, are too small for classical mechanics calculations. Electronic energy must be computed by solving the quantum mechanical Schrödinger equation. The lowest energy solution to this equation is called the ground-state wave function. Molecular orbital methods provide

approximate solutions of the Schrödinger equation, the groundstate wave function, and certain low energy excited states.

            Quantum mechanics methods have several attractive features:

• They require no information about the location or geometry of bonds in a molecular system.

• Parameters for elements (basis functions in ab initio methods usually derived from experimental data and empirical parameters in semi-empirical methods usually obtained from empirical data or ab initio calculations) are independent of the chemical environment. In contrast, parameters used in molecular mechanics methods often depend on the chemical environment.

• With simple precautions, quantum mechanics methods can describe bond breaking.

• HyperChem can perform quantum mechanics MO calculations on molecules containing 100 or more atoms. There is no restriction on the number of atoms, but larger structures may require excessive computing times and computer main memory.

• You can interpret results, including dipole moments and atomic charges, using the simple concepts and familiar vocabulary of the Linear Combination of Atomic Orbitals (LCAO)–molecular orbital (MO) theory.

            Once HyperChem calculates potential energy, it can obtain all of the forces on the nuclei at negligible additional expense. This allows for rapid optimization of equilibrium and transition-state geometries and the possibility of computing force constants, vibrational modes, and molecular dynamics trajectories.

            In addition to computing molecular geometries, energies, and forces, HyperChem computes and displays charge and spin density functions, atomic charges, dipole moments, and electrostatic potential. Individual molecular orbital contributions, important in visualizing and interpreting chemical results, are calculated and displayed graphically.

            8.3. Range of Quantum Mechanics Methods

            The quantum mechanics methods in HyperChem differ in how they approximate the Schrödinger equation and how they compute potential energy. The ab initio method expands molecular orbitals into a linear combination of atomic orbitals (LCAO) and does not introduce any further approximation.

            The Extended Hückel method, for example, does not explicitly consider the effects of electron-electron repulsions but incorporates repulsions into a single-electron potential. This simplifies the solution of the Schrödinger equation and allows HyperChem to compute the potential energy as the sum of the energies for each electron.

            Neglect of Differential Overlap (NDO) methods, such as CNDO and INDO, include electron repulsions, but the resulting equations are nonlinear. The energy calculation includes terms for pairs of electrons. These methods include nuclear repulsions, but only by reducing the charge on each nucleus by the number of core electrons shielding it. The other NDO methods, MINDO/3, MNDO, AM1, and PM3 replace nuclear repulsion terms in the potential energy by parameterized, core-repulsion terms. The terms compensate for considering only valence electrons in the electronic Schrödinger equation, and they incorporate effects of electron correlation.

            8.4. Exclusion Principle

            Because of the quantum mechanical Uncertainty Principle, quantum mechanics methods treat electrons as indistinguishable particles. This leads to the Pauli Exclusion Principle, which states that

the many-electron wave function—which depends on the coordinates of all the electrons—must change sign whenever two electrons interchange positions. That is, the wave function must be antisymmetric with respect to pair-wise permutations of the electron coordinates.

Example: If two functions are sin and cos and two variables are x and y, then sin(x)cos(y) has the same sign if you interchange x and y to give sin(y) cos(x). The antisymmetrized version, sin(x)cos(y)–cos(x)sin(y), does change sign. This is the determinant:

                                                     (118)

Example: If a(1) is a function of the coordinates of electron 1 and b(2) is a function of the coordinates of electron 2, then the product function, a(1)b(2), is not antisymmetric. The function a(1)b(2)–b(1)a(2), an expansion of a 2x2 determinant, is antisymmetric.

            The Exclusion Principle is fundamentally important in the theory of electronic structure; it leads to the picture of electrons occupying distinct molecular orbitals. Molecular orbitals have well-defined energies and their shapes determine the bonding pattern of molecules. Without the Exclusion Principle, all electrons could occupy the same orbital.

            The Exclusion Principle also gives rise to the “exchange” phenomenon. This phenomenon is important in understanding the energy differences between electronic states of different spin symmetries; for example, singlet-triplet splittings and related properties.

            Exchange is also important in understanding the nature of the covalent bond. In the simplest description of bonds for the hydrogen molecule, almost all of the binding energy is due to exchange. (Binding energy is the energy of the two hydrogen atoms separated by infinity minus the energy of the stable hydrogen molecule at its equilibrium bond length.)

            The Exclusion Principle is quantum mechanical in nature, and outside the realm of everyday, “classical” experience. Think of it as the inherent tendency of electrons to stay away from one another; to be mutually excluded. Exclusion is due to the antisymmetry of the wave function and not to electrostatic coulomb repulsion between two electrons. Exclusion exists even in the absence of electrostatic repulsions.

            Consider what happens to the many-electron wave function when two electrons have identical coordinates. Since the electrons have the same coordinates, they are indistinguishable; the wave function should be the same if they trade positions. Yet the Exclusion Principle requires that the wave function change sign. Only a zero value for the wave function can satisfy these two conditions, identity of coordinates and an antisymmetric wave function. For the hydrogen molecule, the antisymmetric wave function is a(1)b(1)–x sin x cosy siny cosb(1)a(1), which is zero. Since the wave function is a probability function, the probability of two electrons having the same coordinates is also zero. In this sense they are mutually excluded.

            8.5. Simplified Wave Functions

            The simplest many-electron wave function that satisfies the Exclusion Principle is a product of N different one-electron functions that have been antisymmetrized, or written as a determinant. Here, N is the number of electrons (or valence electrons) in the molecule. HyperChem uses this form of the wave function for most semi-empirical and ab initio calculations. Exceptions involve using the Configuration Interaction option (see page 119 of Manual). HyperChem computes one-electron functions, termed molecular spin-orbitals, by relatively simple integration and summation calculations.

            The many-electron wave function, which has N! terms (the number of terms in the determinant), never needs to be evaluated. N molecular spin orbitals must be different from one another in a way that satisfies the Exclusion Principle. Because the wave function is written as a determinant, interchanging two rows of the determinant corresponds to interchanging the coordinates of the two electrons. The determinant changes sign according to the antisymmetry requirement. It also changes sign when two columns are interchanged; this corresponds to interchanging two spin orbitals.

            To avoid having the wave function zero everywhere (an unacceptable solution), the spin orbitals must be fundamentally different from one another. For example, they cannot be related by a constant factor. You can write each spin orbital as a product of a space function which depends only on the x, y, and z coordinates of the electron—and a spin function. The space function is usually called the molecular orbital. While an infinite number of space functions are possible, only two spin functions are possible: alpha and beta.

            To use HyperChem for calculations, you specify the total molecular charge and spin multiplicity (see “Charge, Spin, and Excited State” on page 119 of Manual). The calculation selects the appropriate manyelectron wave function with the correct number of alpha or beta electrons. You don’t need to specify the spin function of each orbital.

            8.6. Using of Hartree-Fock Wave Functions

            Quantum mechanics calculations use either of two forms of the wave function: Restricted Hartree-Fock (RHF) or Unrestricted Hartree-Fock (UHF). Use the RHF wave function for singlet electronic states, such as the ground states of stable organic molecules. Spin orbitals are grouped in pairs for an RHF calculation. Each member of the pair differs in its spin function (one alpha and one

beta), but both must share the same space function. For N electrons, N/2 different molecular orbitals (space functions) are doubly occupied, with one alpha (spin up) and one beta (spin down) electron forming a pair.

            The UHF wave function is most often used for multiplicities greater than singlets. In this case, the space orbitals for alpha electrons need not be the same as for beta electrons. For open-shell systems, some of the spins are unpaired — alpha and beta electrons occupy different orbitals. This is because there are more alpha electrons than beta electrons, and they tend to repel each other more than they do beta electrons. This effect is known as exchange repulsion. It is a quantum mechanical effect with no classical analog and would persist even without electrostatic coulomb forces. Essentially, it is due to the requirement that the wave function must be antisymmetric (see “Exclusion Principle”). Exchange repulsion affects the choice of the semi-empirical method (see “Choosing a Semi-Empirical Method”). The UHF wave function can also apply to singlet molecules. Usually, the results are the same as for the faster RHF method. That is, electrons prefer to pair, with an alpha electron sharing a molecular space orbital with a beta electron. Use the UHF method for singlet states only to avoid potential energy discontinuities when a covalent bond is broken and electrons can unpair.

            8.7. Extending the Wave Function Calculation in method of Configuration Interaction

            You can extend the calculation of the Hartree-Fock semi-empirical wave function by choosing Configuration Interaction (CI) in the Semi-empirical Options dialog box and you can also extend the calculation of the Hartree-Fock ab initio wave function by choosing CI in the ab initio  Options dialog box. Use CI for these electron configurations:

• closed-shell singlet ground states (for both semi-empirical and ab initio methods)

• half-electron, excited singlet states (for semi-empirical methods only)

• half-electron, doublet and triplet open-shell ground states (for semi-empirical methods only).

            Configuration Interaction (or electron correlation) adds to the single determinant of the Hartree-Fock wave function a linear combination of determinants that play the role of atomic orbitals. This is similar to constructing a molecular orbital as a linear combination of atomic orbitals. Like the LCAO approximation, CI calculations determine the weighting of each determinant to produce the lowest energy ground state (see “SCF Technique” on page 43). CI calculations can be used to improve the quality of the wavefunction and state energies. Self-consistent field (SCF) level calculations are based on the one-electron model, wherein each electron moves in the average field created by the other n-1 electrons in the molecule. Actually, electrons interact instantaneously and therefore have a natural tendency to avoid each other beyond the requirements of the Exclusion Principle. This correlation results in a lower average interelectronic repulsion and thus a lower state energy. The difference between electronic energies calculated at the SCF level versus the exact nonrelativistic energies is the correlation energy.

            There are two types of electron correlations: static and dynamic. Static correlation refers to a near degeneracy of a given state; a dynamic correlation refers to the instantaneous avoidance of electrons with each other.

            CI calculations are perhaps the most widely used method of going computationally beyond an SCF description. An SCF computation yields a configuration describing the orbital occupancy of the electrons. Other configurations may be generated from this reference configuration by exciting electrons from the set of occupied orbitals to the set of virtual (unoccupied) orbitals. A CI calculation yields a set of improved states, each of which is represented by a linear combination of these configurations.

            There are two types of CI calculations implemented in HyperChem — singly excited CI and microstate CI. The singly excited CI which is available for both ab initio and semi-empirical calculations may be used to generate UV spectra and the microstate CI available only for the semi-empirical methods in HyperChem is used to improve the wave function and energies including the electronic correlation. Only single point calculations can be performed in HyperChem using CI.

            You can use CI calculations to do the following:

• Calculate UV spectra.

• Calculate the energy of excited states.

• Study the making or breaking of bonds, and change of spin couplings (e.g. dissociation of H2).

• Capture the effects of London dispersion forces.

• Describe a nearly degenerate state.

• Study singlet-triplet splittings more accurately.

            You can use CI to predict electronic spectra. Since the CI wave function provides ground state and excited state energies, you can obtain electronic absorption frequencies from the differences between the energy of the ground state and the excited states.

            You can also use a RHF wave function with CI for calculations involving bond breaking, instead of using a UHF wave function.

            The Microstate CI Method lowers the energy of the uncorrelated ground state as well as excited states. The Singly Excited CI Method is particularly appropriate for calculating UV visible spectra, and does not affect the energy of the ground state.

            Be careful when you use the Orbital Criterion for symmetrical systems. To get correct results, you must include all or none of any set of degenerate orbitals in the CI, not just some of them. Carrying out an RHF calculation first and studying the Orbitals dialog box will help you to spot degenerate orbitals and avoid this pitfall.

            In large systems there can be many orbitals in a small energy range, and the size of the CI matrix can be very sensitive to the value of the maximum excitation if you use Energy Criterion. Since calculation time depends heavily on the size of the CI matrix, you can end up with very long calculations, especially if you use the ab initio methods or the MNDO, AM1, or PM3 semi-empirical methods. This could exhaust the memory of your system. Again, inspecting the results of an RHF (no CI) calculation will help you avoid these pitfalls.

            Using CI may not necessarily improve the calculation of ground state energies. Parameters for the MINDO/3, MNDO, AM1, and PM3 methods already include the effects of CI. CI calculations require more computing time.

Note: When you use CI, HyperChem does not calculate forces (gradient).

            8.8. Møller-Plesset Perturbation Theory

            Configuration interaction is a systematic procedure for going beyond the Hartree-Fock approximation. A different systematic approach for finding the correlation energy is perturbation theory (Rayleigh-Schrödinger many-body perturbation theory — RSPT). In this approach, the total Hamiltonian of the system is divided or partitioned into two parts: a zeroth-order part, H0 (which has known eigenfunctions and eigenvalues), and a perturbation, V. The exact energy is then expressed as an infinite sum of contributions of increasing complexity. The expressions for these contributions contain the eigenvalues of H0 and matrix elements of the perturbation between the eigenfunctions of H0. By introducing the ordering parameter λ, the total Hamiltonian can be written as

H= H0 + λV                                                                  (119)

            The exact eigenfunctions and eigenvalues can now be expanded in a Taylor series in λ.

Ψλ = Ψ(0) + λΨ(1) + λ2Ψ(2) + …                                         (120)

Eλ = E(0) + λE(1) +λ2E(2) + …                                          (121)

            If H0 is chosen wisely, then the perturbation is small and the perturbation expansion (i.e., the sum of the 1st, 2nd, ..., nth-order energies) converges quickly. To obtain a perturbation expansion for the correlation energy, the best way is to choose the Hartree-Fock Hamiltonian as the zeroth-order Hamiltonian. The application to N-electron molecular systems is sometimes called Møller-Plesset (MP) perturbation theory. These methods, which can be terminated at second (MP2), third (MP3), or fourth order (MP4), with these three being the most frequently used in different ab initio programs, calculate the correlation energy and rely on a good description of the virtual orbitals in the original SCF function. The calculated total correlation energy is therefore quite dependent on the quality of the basis set.

            HyperChem supports MP2 (second order Møller-Plesset) correlation energy calculations using ab initio methods with any available basis set. In order to save main memory and disk space, the HyperChem MP2 electron correlation calculation normally uses a so called “frozen-core” approximation, i.e. the inner shell (core) orbitals are omitted. A setting in CHEM.INI allows excitations from the core orbitals to be included if necessary (melted core). Only the single point calculation is available for this option.

Chapter 9. Calculations of Molecular Orbitals

            The molecular orbitals computed by HyperChem have several important properties. Each molecular orbital has a uniquely defined orbital energy. Interpretations of orbital energies give useful predictions. For example, the highest energy-occupied molecular orbital (HOMO) provides a remarkably accurate estimate of the first ionization potential (Koopmans’ theorem10). Orbital energies are also important in interpreting molecular geometry (Walsh’s rules) and chemical reactivity (Woodward-Hoffmann rules).

            9.1. Orbital Occupancy

            Because of the LCAO-MO approximation, ab initio and semiempirical calculations produce occupied and unoccupied (virtual) orbitals. The Aufbau or “building up” principle determines the occupancy of orbitals for the ground-state wave function. This first arranges the orbitals in order of increasing energy. Then it assigns electrons to orbitals beginning with the lowest energy. For example, in an RHF calculation for a molecule with ten electrons and twelve molecular orbitals, the five lowest energy orbitals receive two electrons each. The remaining seven virtual orbitals remain unoccupied and do not appear in the many-electron wave function. Some of these virtual orbitals are important in discussions of chemical reactivity and excited electronic states (see “Chemical

Reactivity”).

            In an Extended Hückel calculation, the Aufbau population of molecular orbitals is unambiguous. The calculation method is non-iterative and the total energy is proportional to the sum of the energies of occupied orbitals. The Aufbau guarantees the lowest energy wave function. For NDO calculations (see “NDO Methods” on page 126 of Manual), the Aufbau also produces the lowest energy wave function. However, in some cases, a calculation may converge to a result that is not the ground state (lowest energy), even though the Aufbau is satisfied (for example, RHF calculations on singlet biradicals or molecules with highly distorted bonds).

            In extreme cases, a calculation may not converge on a lowest energy solution because of the iterative nature of the SCF procedure. The highest energy occupied (HOMO) and lowest energy unoccupied (virtual) orbitals (LUMO) may switch occupancies from one SCF cycle to the next when the two orbitals are close in energy (nearly degenerate). It is even possible, though unlikely, that the ground state would not satisfy the Aufbau—there is a virtual orbital with an orbital energy lower than an occupied orbital.

            HOMO and LUMO, also known as Frontier orbitals, are important in interpreting results of a calculation (see “Frontier Molecular Orbitals” on page 141 of Manual). You can use these molecular orbitals to compute the lowest excited electronic singlet state of molecules and the ground states of radicals.

            9.2. Atomic orbitals and their Interactions

            HyperChem uses the Linear Combination of Atomic Orbitals-Molecular Orbital (LCAO-MO) approximation for all of its ab initio and semi-empirical methods. If Ψi represents a molecular orbital and fμ an atomic orbital, then the LCAO representation of the molecular orbital is

                                                             (122)

where Cμi is the coefficient or “weight” of the μ th atomic orbital in the ith molecular orbital.

            This equation is important in interpreting the results of calculations. In ab initio and semi-empirical calculations, atomic orbitals are functions of the x, y, and z coordinates of the electron that closely resemble the valence orbitals of the isolated atoms.

            These atomic orbitals, called Slater Type Orbitals (STOs), are a simplification of exact solutions of the Schrödinger equation for the hydrogen atom (or any one-electron atom, such as Li+2). HyperChem uses Slater atomic orbitals to construct semi-empirical molecular orbitals. The complete set of Slater atomic orbitals is called the basis set. Core orbitals are assumed to be chemically inactive and are not treated explicitly. Core orbitals and the atomic nucleus form the atomic core.

            Because the calculation of multi-center integrals that are inevitable for ab initio method is very difficult and time-consuming, HyperChem uses Gaussian Type Orbital (GTO) for ab initio methods. In truly reflecting a atomic orbital, STO may be better than GTO, so HyperChem uses several GTOs to construct a STO. The number of GTOs depends on the basis sets. For example, in the minimum STO-3G basis set HyperChem uses three GTOs to construct a STO.

            9.3. SCF Technique

            All molecular orbitals are combinations of the same set of atomic orbitals; they differ only by their LCAO expansion coefficients. HyperChem computes these coefficients, Ci, and the molecular orbital energies by requiring that the ground-state electronic energy be at a minimum. That is, any change in the computed coefficients can only increase the energy. The technique for this calculation involves two steps. The first step computes the Hamiltonian or energy matrix. The elements of this matrix are integrals involving the atomic orbitals and terms obtained from the Schrödinger equation. The most important contributions to this matrix involve the overlap integrals, which are mainly responsible for the bonding between atoms.

            The second step determines the LCAO coefficients by standard methods for matrix diagonalization. In an Extended Hückel calculation, this results in molecular orbital coefficients and orbital energies. Ab initio and NDO calculations repeat these two steps iteratively because, in addition to the integrals over atomic orbitals, the elements of the energy matrix depend upon the coefficients of the occupied orbitals. HyperChem ends the iterations when the coefficients or the computed energy no longer change; the solution is then “self-consistent.” The method is known as Self-Consistent Field (SCF) calculation.

Note: You should be aware that the length of semi-empirical calculations increases as the cube of the number of atomic orbitals, because this is the number of operations required for matrix diagonalization; and the length of ab initio calculations increases as the power of four of the number of atomic orbitals (basis functions), because of the calculation of two-electron repulsion integrals. A molecule with 60 atomic orbitals takes about eight times as long to compute as a molecule containing 30 atomic orbitals by using a semi-empirical method and sixteen times as long to compute by using the ab initio method.

            9.4. Virtual Orbitals

            Even with the minimal basis set of atomic orbitals used in most semi-empirical calculations, the number of molecular orbitals resulting from an SCF calculation exceeds the number of occupied molecular orbitals by a factor of about two. The number of virtual orbitals in an ab initio calculation depends on the basis set used in this calculation.

            9.5. Using of RHF and UHF methods

            HyperChem quantum mechanics calculations must start with the number of electrons (N) and how many of them have alpha spins (the remaining electrons have beta spins). HyperChem obtains this information from the charge and spin multiplicity that you specify in the Semi-empirical Options dialog box or ab initio Options dialog box. N is then computed by counting the electrons (valence electrons in semi-empirical methods and all electrons in ab initio method) associated with each (assumed neutral) atom and subtracting the charge. Multiplicity, which is related to the allowed values of the total spin angular momentum, is one for a singlet state, two for a doublet, three for a triplet, and four for a quartet.

            9.6. Multiplicity Considerations

            Most stable organic molecules have an even number of electrons and a singlet ground-state wave function. Half the electrons are alpha. For a radical with an odd number of electrons, the ground state is almost always a doublet with one more alpha than beta electron. For some highly symmetrical molecules, such as oxygen, the ground state is a triplet with two more alpha than beta electrons. Biradicals (metastable species formed, for example, by breaking a bond in a ring molecule) also frequently have triplet ground states.

Fig. 31. Distribution of electrons on orbitals in RHF (a) and UHF (b)methods

            If you specify a multiplicity of one (singlet), then you would most often choose the RHF method, unless the reactions result in bond breaking (see page 46). If the selected multiplicity is greater than one, then the system is open-shell and the usual choice is the UHF method, which uses different orbitals for electrons with different spins.

            9.7. Bond Breaking

            Applications of semi-empirical and ab initio methods frequently follow the course of chemical reactions that break and form covalent bonds. You must take care in choosing a method for these applications.

            If a covalent bond is broken, as in the simple case of dissociation of the hydrogen molecule into atoms, then the RHF wave function without the Configuration Interaction option (see “Extending the Wave Function Calculation” on page 37 of Manual) is inappropriate. This is because the doubly occupied RHF molecular orbital includes spurious terms that place both electrons on the same hydrogen atom, even when they are separated by an infinite distance.

            An RHF wave function is also inappropriate for a reaction where reactants and products share different occupied orbitals, as in a Woodward-Hoffmann orbital symmetry forbidden thermal reaction. For example, in the absence of symmetry, the disrotatory ring opening of cyclobutene to butadiene must involve a change in the orbital occupancy between reactants and products. The lowest virtual orbital in reactants becomes the highest occupied in the product. This gives rise to an artificial and undesirable cusp in the potential energy surface along the reaction path, plus an unrealistically high activation energy. You can prevent this problem and the breaking of a covalent bond by using a UHF wave function for the singlet state. A minor disadvantage of the UHF method is that the space and spin symmetry of the molecular orbitals are broken.

            For example, the UHF wave function is a mixture of singlet and triplet states. To preserve spin symmetry, use RHF plus the Configuration Interaction option instead of UHF. For Woodward-Hoffman allowed thermal reactions (such as the conrotatory ring opening of cyclobutane), orbital symmetry is conserved and there is no change in orbital occupancy. Even though bonds are made and broken, you can use the RHF wave function.

            9.8. RHF Half-Electron Technique

            With all semi-empirical methods, HyperChem can also perform pseudo-RHF calculations for open-shell systems. For a doublet state, all electrons except one are paired. The electron is formally divided into two “half electrons” with paired spins. Each half electron carries half the charge of a whole electron. The calculation produces a set of molecular orbitals appropriate for this pseudowave function. HyperChem then assigns the unpaired electron its proper spin (alpha), substitutes this electron in the orbital formerly occupied by the half electrons, and calculates energy and other properties.

            You can think of the half-electron technique as a device to get reasonable orbitals to populate later with whole electrons.

            The half-electron method can also apply to triplet states. For this calculation, HyperChem populates selected molecular orbitals with pairs of half electrons. The final energy is computed by assigning the proper spins.

            You can study the lowest excited singlet state of a system with a closed-shell ground state. The difference in energy between this state and the singlet ground state is related to the lowest electronic absorption frequency. Here, the calculation is carried out for the half-electron triplet state. The final energy is computed using the appropriate singlet wave function (which involves two determinants) for this singly excited state. An advantage of the half-electron technique is its simplicity.

            HyperChem can carry it out with only minor modifications of the usual calculation. A disadvantage is that forces may not be accurate because of the half electron approximation. Although UHF is often a better theoretical treatment of open-shell systems than the RHF (half-electron) methods, it takes longer to compute. Separate matrices for electrons of each spin roughly double the length of the calculation.

            9.9. SCF Convergence

            Not all iterative semi-empirical or ab initio calculations converge for all cases. For SCF calculations of electronic structure, systems with a small energy gap between the highest occupied orbital and the lowest unoccupied orbital may not converge or may converge slowly. (They are generally poorly described by the Hartree-Fock method.)

            Some systems converge poorly, particularly those with multiple bonds or weak interactions between open-shell systems.

            HyperChem includes two convergence accelerators. One is the default convergence accelerator, effective in speeding up normally convergent calculations. The other, called the Direct Inversion in the Iterative Subspace (DIIS) method, is available for all the SCF methods in HyperChem. You can turn on this accelerator in the Semi-empirical Options dialog box or in the Ab Initio Options dialog box. It uses more memory and computing time.

            If you observe convergence problems, you might try a different semi-empirical method or first try optimizing the geometry using another method. Convergence problems occur more often away from optimized geometries than close to them. Often, a fast geometry optimization with a molecular mechanics calculation may solve convergence difficulties with a semi-empirical optimization.

            9.10. Calculation Results

            The results of calculations are often interpreted in terms of familiar orbital interaction diagrams. The simplest example is two one electron atoms combining to form a molecule. This diagram shows a two-orbital interaction forming a bonding (lower energy) and antibonding (higher energy) combination.

 

Fig.32. Bonding and antibonding MO

Note: This simple orbital interaction picture is useful for interpreting results, but neglects many aspects of a calculation, such as electron-electron interactions. These diagrams are closely related to the results from Extended Hückel calculations.

            A larger interaction energy (the off-diagonal element of an energy matrix, proportional to the overlap integral between two molecular orbitals) causes a larger energy separation between two atomic orbitals, and the lower orbital is doubly occupied according to the Aufbau principle (see “Orbital Occupancy” on page 41 of Manual). The magnitude of this additional splitting is proportional to the magnitude of the overlap integral between the two atomic orbitals. The greater the splitting, the greater the strength of the covalent bond. The occupied molecular orbital results from the in-phase or positive combination of the two atomic orbitals. This is shown for an s-type orbital on the left atom and a sigma p-type on the right. The virtual orbital is the negative or out-of-phase combination. The occupied or “bonding” molecular orbital shows a buildup of electron density between the two atoms. The virtual orbital, on the other hand, has a node (a value of zero) between the atoms.

            A simple example is the formation of the hydrogen molecule from two hydrogen atoms. Here the original atomic energy levels are degenerate (they have equal energy), but as the two atoms approach each other, they interact to form two nondegenerate molecular orbitals, the lowest of which is doubly occupied.

            In the general heteronuclear case, individual atomic orbital and overlap energies determine the magnitude of energy difference between two molecular orbitals. Atomic orbital energies are the diagonal elements of the Hamiltonian matrix, and the interaction energy between atomic orbitals are the off diagonal elements.

            HyperChem can display molecular orbitals and the electron density of each molecular orbital as contour plots, showing the nodal structure and electron distribution in the molecular orbitals.

            9.11. Quantitative Results

            HyperChem provides quantitative results for all calculations. The status line shows the binding energy and gradient (only binding energy if you use the Configuration Interaction option, see page 119 of Manual). The log file for a calculation contains this information, plus more, including heats of formation (for all the SCF semiempirical methods in HyperChem), orbital coefficients, dipole moments, and atomic charges.

            When you perform a single point semi-empirical or ab initio calculation, you obtain the energy and the first derivatives of the energy with respect to Cartesian displacement of the atoms. Since the wave function for the molecule is computed in the process, there are a number of other molecular properties that could be available to you. Molecular properties are basically an average over the wave function of certain operators describing the property. For example, the electronic dipole operator is basically just the operator for the position of an electron and the electronic contribution to the dipole moment is

                                                 (123)

The dipole moment for a calculation is reported and is contained in the log file if logging is turned on. It is also reported on the status line if you Display Dipole Moment on the Display menu. Other expectation values besides the dipole moment (for example, quadrupole moment) could be reported with a wave function but the set reported with this release of HyperChem is limited to only a few. Below we discuss the properties or other characterizations of the calculated wave function that can be interactively visualized.

            9.12. Contour Plots

            HyperChem allows the visualization of two-dimensional contour plots for a certain number of variables. These contour plots show the values of a spatial variable (a property f(x,y,z) in normal three dimensional Cartesian space) on a plane that is parallel to the screen. To obtain these contour plots the user needs to specify:

• The plane to be contoured.

• The contours to be shown.

• The number and positioning of the grid of points on the plane to be used in computing values of the variable.

            The plane is specified as a plane parallel to the screen and offset from the center of mass of the current selection towards the user by a specified number of Ångstroms. For example, if you make no selection and specify an offset of 0.0, the plane is through the center of mass of the whole molecular system. If you select a single atom, again with an offset of 0.0, then the plane is that plane parallel to the screen that goes through the selected atom. If you specify an offset of 1.0 instead of 0.0, then the plane is 1.0 Å closer to the viewer than before. You can also specify a negative offset, which pushes the plane away from the user and into the screen.

            You can specify the number and values of visible contour lines.You specify the total number of contour lines to be shown by simple stating the number, n>0. You normally specify the values of the contour lines as default values. For this case, HyperChem computes the maximum and minimum values on the grid and then draws contours at these values plus n-2 contour lines evenly spaced in between these maximum and minimum values. If you need non-default values, you can specify the starting value and then an increment to define the other n-1 evenly spaced contour lines. If default values were computed previously, HyperChem suggests the starting value and increment of the previous default computation for the new non-default option.

            The grid of computed values for the variable used in defining the contour is a grid of exactly the size of the current working area subdivided evenly such that the total number of grid points is as the user specifies.

            When a molecule rotates, HyperChem abandons the current contour map because it is no longer valid and you must explicitly request a new contour map via the Graph check box.

            9.13. Total Electron Density

            The density matrices are a representation, in the basis set fμ of the electron density at a point in space. The electron density ρ(r), described above by equation (62) on page 198 of Manual, represents a probability distribution function (the probability of an electron being at a particular point in space (rigorously, in a small volume dr at the point r in ordinary three-dimensional space). The total density can be subdivided into the probability of finding a spin-up (alpha) electron plus the probability of finding a spin-down (beta) electron:

ρT(r) =  ρα(r) + ρβ(r)                                                              (124)

where the individual densities are defined to be identical in a

closed-shell RHF calculation but can be different in a UHF calculation,

according to:

                                                          (125)

                                                        (126)

            The neglect of diatomic differential overlap approximation used in the semi-empirical methods reduces these formulae to much simpler formulae that can be computed quite rapidly,

 where the double sum is over each atom and then the orbitals on that atom. Without this last approximation the electron density calculation can be a long one, even longer than solving the approximate Schrödinger equation in the first place. The reason is the large number of grid points and the double sum over all pairs of atomic orbitals at each grid point. For ab initio and extended Hückel calculations where no neglect of overlap (differential or otherwise) is implied in the approximations, the complete calculation is performed and requires considerable patience!

            The CNDO/INDO, MINDO/3, ZINDO/1, and ZINDO/S methods might be expected to imply an even simpler equation for the electron density than the above. For example, a rigorous complete neglect of CNDO approximation, suggests that equations (127) and (128) should be replaced by expressions with a sum only over diagonal elements of the density matrix. This would represent a molecular charge density that is the exact sum of atomic densities. Alternatively, expectation values like the dipole moment, computed from the this density matrix, would have only monopole charge on each atom contributions. The original authors of the CNDO method recognized that a hybridization term (coming from onecenter integrals combining s and p type orbitals) was important to the dipole moment and included it as part of the method. In order to make the electron density consistent with expectation values and take account of these hybridization contributions to dipole moments, electrostatic fields, etc., all the semi-empirical SCF methods in HyperChem use equations (87) and (88) for the electron density and associated expectation values.

            9.14. Spin Density

            For open-shell situations or for some closed-shell UHF situations (A UHF calculation for a normal closed-shell situation will degenerate into the closed-shell RHF solution in most cases. However, a true UHF solution is always obtained when a bond is stretched to such an extent that it is clearly broken and only an unrestricted open shell descriptions of the two fragments is appropriate) non-zero spin densities exist and can be contoured. The definition of a closed-shell situation has subtleties, but a simplistic definition of it might be that the probability of finding a spin-up electron is identical to the probability of finding a spin-down electron, everywhere in space. Thus, in contrast to the total electron density defined as the sum of alpha and beta densities above, it is convenient to define the spin density as the difference:

                                                            (129)

The spin density defines the excess probability of finding spin-up over spin-down electrons at a point in space and is zero everywhere for closed-shell RHF situations. The spin density at the position of a nucleus is a prime determinant of electron spin resonance (ESR) spectra.

            9.15. Orbital Plots

            For many reasons, including the Woodward-Hoffman rules that describe the likelihood of reaction based on arguments about the shapes of orbitals, it is desirable to be able to visualize molecular orbitals.

            You can use the semi-empirical and ab initio orbitals dialog box in HyperChem to request a contour plot of any molecular orbital. When requested, the orbital is contoured for a plane that is parallel to the screen and which is specified by a subset selection and a plane offset, as described above. The index of the orbital and its orbital energy (in electron volts, eV) appears in the status line.

            If a molecule is rotated by changing the position of the viewer (left mouse button rotation) then the molecule's position in the molecular coordinate system has not changed and another contour plot can be requested without recomputing the wave function. That is, many orbitals can be plotted after a single point ab initio or semiempirical calculation. Any contour map is available without recomputation of the wave function.

            When you request an orbital, you also request a plot of either the orbital itself or of its square. The orbital ψi is a signed quantity with nodes where the sign changes and the value of the orbital goes to zero. The nodal structure of orbitals has chemical significance. For planar molecules you can distinguish s-orbitals which have nonzero value in the plane of the molecule from p-orbitals which have a node in the plane of the molecule. As energy increases, an orbital generally has more and more nodes. The orbital squared represents the probability distribution function for an electron in the orbital.

            When you request an orbital, you can use the cardinal number of the orbital (ordered by energy and starting with number=1) or an offset from either the highest occupied molecular orbital (HOMO) or the lowest unoccupied molecular orbital (LUMO). Offset from the HOMO are negative and from the LUMO are positive. Often these frontier orbitals are the ones of most chemical interest.

            9.16. Electrostatic Potential

            The results of electrostatic potential calculations can be used to predict initial attack positions of protons (or other ions) during a reaction. You can use the Contour Plot dialog box to request a plot of the contour map of the electrostatic potential of a molecular system after you done a semi-empirical or ab initio calculation. By definition, the electrostatic potential is calculated using the following expression:

                                             (130)

            The point r is the position of a positive probe charge. ZA is the nuclear charge on atom A located at position RA. The function ρ(r') is the electronic density. In the above equation, the first term represents the contribution of the nuclei to the electrostatic potential and the second term is the electronic contribution. Substituting the electron density expression:

                                         (131)

into the electrostatic potential expression above, the electrostatic potential can be rewritten as

                             (132)

            The second summation is over all the orbitals of the system. This equation is used in HyperChem ab initio calculations to generate contour plots of electrostatic potential. If we choose the approximation whereby we neglect the effects of the diatomic differential overlap (NDDO), then the electrostatic potential can be rewritten as

                              (133)

of the above is over the orbitals of atom A.

            In the given book we have given only short characteristics of program HyperChem. Many theoretical questions have been already considered in theoretical parts of the book. Only detailed studying of theoretical bases of quantum chemistry and acquaintance with a manual of the program ãèïåðõåì  allow to chemists to estimate all charms of quantum chemistry.

 

Acknowledgments

            In the end we would like to note, that the present book is written by wide using of materials free web-site Wikipedia. We are very grateful to authors of these materials and we hope that popularization of quantum chemistry will bring the big advantage for many readers of this site.           

            References  

1.      http://en.Wikipedia.org/wiki/Quantum Chemistry/

2.      Computational Chemistry, David Young, Wiley-Interscience, 2001. Chapter 41, pg 318.

3.      Smith, S. J.; Sutcliffe B. T., (1997). "The development of Computational Chemistry in the United Kingdom". Reviews in Computational Chemistry 70: 271 - 316.

4.      Schaefer, Henry F. III (1972). The electronic structure of atoms and molecules. Reading, Massachusetss: Addison-Wesley Publishing Co., 146.

5.      Boys, S. F.; Cook G. B., Reeves C. M., Shavitt, I. (1956). "Automatic fundamental calculations of molecular structure". Nature 178 (2): 1207.

6.      Richards, W. G.; Walker T. E. H and Hinkley R. K. (1971). A bibliography of ab initio molecular wave functions. Oxford: Clarendon Press.

7.      Preuss, H. (1968). International Journal of Quantum Chemistry 2: 651.

8.      Buenker, R. J.; Peyerimhoff S. D. (1969). Chemical Physics Letters 3: 37.

9.      Schaefer, Henry F. III (1984). Quantum Chemistry. Oxford: Clarendon Press.

10.  Streitwieser, A.; Brauman J. I. and Coulson C. A. (1965). Supplementary Tables of Molecular Orbital Calculations. Oxford: Pergamon Press.

11.  Pople, John A.; David L. Beveridge (1970). Approximate Molecular Orbital Theory. New York: McGraw Hill.

12.  Allinger, Norman (1977). "Conformational analysis. 130. MM2. A hydrocarbon force field utilizing V1 and V2 torsional terms". Journal of the American Chemical Society 99: 8127-8134.

13.  Fernbach, Sidney; Taub, Abraham Haskell (1970). Computers and Their Role in the Physical Sciences. Routledge. ISBN 0677140304.

14.   Reviews in Computational Chemistry vol 1, preface

15.   Dearing, A., Computer-aided Molecular Modelling: Research Study or Research Tool J. Computer-Aided Molecular Design 2:179–189, 1988.

16.  Gund, P.; Barry, D.C.; Blaney, J.M.; Cohen, N.C. Guidelines for Publications in Molecular Modeling Related to Medicinal Chemistry J. Med. Chem. 31:2230–2234, 1988.

17.  Fleming, I. Frontier Orbitals and Organic Chemical Reactions John Wiley & Sons, New York, 1976.

18.  Wu, C.S.; Neely, W.C.; Worley, S.D. A Semiempirical Theoretical Study of the Molecular Interaction of Cocaine with the Biological Substrate Glycine. J. Comput. Chem. 12:862-867, 1991.

19.  Elber, R.; Karplus, M. Multiple conformational states of proteins: a molecular dynamics analysis of myoglobin. Science 235:318-321, 1987.

20.  Dill, K.A. Theory for the folding and stability of globular proteins. Biochemistry 24:1501-1509, 1985.

21.  Crippen, G.M.; Snow, M.E. A 1.8Å resolution potential function for protein folding. Biopolymers 29:1479-1489, 1990.

22.  Madura, J.D.; Jorgensen, W.L. Ab initio and Monte Carlo calculations for a nucleophilic addition reaction in the gas phase and in aqueous solution. J. Am. Chem. Soc. 108:2517-2527, 1986.

23.  W. Kołos and L. Wolniewicz, Nonadiabatic Theory for Diatomic Molecules and Its Application to the Hydrogen Molecule, Rev. Mod. Phys. vol. 35, pp. 473-483 (1963).

24.  R. G. Woolley and B. T. Sutcliffe, P.-O. Löwdin and the Quantum Mechanics of Molecules in Fundamental World of Quantum Chemistry, E. J. Brändas and E. S. Kryachko (Eds.), Vol. 1, 21-65, Kluwer Academic Publshers, 2003.

25.  Christopher J. Cramer Essentials of Computational Chemistry, John Wiley & Sons (2002)

26.  T. Clark A Handbook of Computational Chemistry, Wiley, New York (1985)

27.  R. Dronskowski Computational Chemistry of Solid State Materials, Wiley-VCH (2005)

28.  F. Jensen Introduction to Computational Chemistry, John Wiley & Sons (1999)

29.  D. Rogers Computational Chemistry Using the PC, 3rd Edition, John Wiley & Sons (2003)

30.  Szabo, N.S. Ostlund, Modern Quantum Chemistry, McGraw-Hill (1982)

31.  D. Young Computational Chemistry: A Practical Guide for Applying Techniques to Real World Problems, John Wiley & Sons (2001)

32.  David Young's Introduction to Computational Chemistry

33.  Born, Max; Oppenheimer,Robert (25 August 1927). "Zur Quantentheorie der Molekeln". Annalen der Physik 84: 457-484. 

34.  Moss, R. E. (1973). Advanced Molecular Quantum Mechanics. Chapman and Hall. 

35.  Tinkham, Michael (2003). Group Theory and Quantum Mechanics. Dover Publications.

36.  A readable and thorough discussion on the spin terms in the molecular Hamiltonian is in: R. McWeeny, Methods of Molecular Quantum Mechanics, 2nd edition, Academic, London (1989).

37.  Levine, Ira N. (1991). Quantum Chemistry. Englewood Cliffs, New jersey: Prentice Hall, 455 - 544.

38.  Cramer, Christopher J. (2002). Essentials of Computational Chemistry. Chichester: John Wiley & Sons, Ltd., 153 - 189.

39.  Jensen, Frank (2007). Introduction to Computational Chemistry. Chichester, England: John Wiley and Sons, 98 - 149.

40.  Colegrove, B. T.; Schaefer, Henry F. III (1990). "Disilyne (Si2H2) revisited". Journal of Physical Chemistry 94: 5593.

41.  Grev, R. S.; Schaefer, Henry F. III (1992). "The remarkable monobridged structure of Si2H2". Journal of Chemical Physics 97: 7990.

42.   Palágyi, Zoltán; Schaefer, Henry F. III, Kapuy, Ede (1993). "Ge2H2: A Molecule with a low-lying monobridged equilibrium geometry". Journal of the American Chemical Society 115: 6901 - 6903.

43.  Stephens, J. C.; Bolton, E. E.,Schaefer, H. F. III, and Andrews, L. (1997). "Quantum mechanical frequencies and matrix assignments to Al2H2". Journal of Chemical Physics 107: 119 - 223.

44.  Palágyi, Zoltán; Schaefer, Henry F. III, Kapuy, Ede (1993). "Ga2H2: planar dibridged, vinylidene-like, monobridged and trans equilibrium geometries". Chemical Physics Letters 203: 195 - 200.

45.  DeLeeuw, B. J.; Grev, R. S. and Schaefer, Henry F. III (1992). "A comparison and contrast of selected saturated and unsaturated hydrides of group 14 elements". Journal of Chemical Education 69: 441.

46.  Jensen, Frank (2007). Introduction to Computational Chemistry. Chichester, England: John Wiley and Sons, 80 - 81.

47.  http://en.wikipedia.org/wiki/Ab_initio_quantum_chemistry_methods

48.  Szabo, A.; Ostlund, N. Modern Quantum Chemistry. Macmillan, New York, 1985.

49.  Koopmans, T. On the Relation of Wave Functions and Eigenvalues to Individual Electrons of Atoms Physica (Utrecht) 1:104, 1933.

50.  Woodward, R.B. and Hoffmann, R. Conservation of Orbital Symmetry Verlag Chemie, Weinheim, FRG, 1970.

51.  Laidler, K. J. (1993) The World of Physical Chemistry, Oxford University Press, p. 347.

52.  Weinhold, F.; Landis, C. Valency and bonding, Cambridge, 2005; pp. 96-100.

53.  Pauling, L. The Nature of the Chemical Bond. Cornell University Press, 1960.