UNIT 5 NON-COVALENT INTERACTIONS

Structure
5.1 ' Introduction
Objectives

5.2

The Structure of Ionic and Molecular Crystals Ionic Bonding and Lattice Energies Intermolecular Forces Hydrogen Bonding Classical Molecular Dynamics Summary Glossary Answers to SAQs
.

5.3 5.4 5.5 5.6

5.7
.

5.6

5.9

5.1

INTRODUCTION

By now we have learnt that covalent interactions are a consequence of a sigr'f' l~ lcant sharing of electrons between the atoms that take part in the bonding process to form molecules. In molecules such as Hz or N,, both the participating atoms share their electrons evenly. In bonding between unlike atoms, we get a wide variety of interactions that range from covalent interactions (as in CH,), ionic interactions (as in the NaCl lattice) and phenomena such as hydrogen bonding. In addition, the non-bonding or the weak interactions between molecules characterised by van der Waals forces are always present and are responsible for the formation of liquids and liquid crystals and molecular solids. In this unit, we begin with the study of ionic solids followed by the study of intermolecular forces, hydrogen bonding and liquid structure and dynamics.

Objectives
After studying this unit, you should be able to

characterize different crystalline lattices and estimate lattice energies, calculate the unit cell parameters from diffraction data, calculate the packing fractions in cubic lattices, compute intermolecular forces from simple model potentials and relate them to molecular motions. analyse the consequences of hydrogen bonding, characterize the liquid state, and relate molecular dynamics to macroscopic properties of liquids.

* *
* * * *
5.2

THE STRUCTURE OF IONIC AND MOLECULAR CRYSTALS

A study of solids usually begins with a description of crystalline solids. We obtain beautiful patterns in crystals essentially due to the presence of basic repeating units. As the crystal grows in size due to the attachment of the repeating units oz the unit cells, the growth can be arrested and the crystal can be cut to a required size and shape. The smallest arrangement of atoms or molecules, which, on repeating in one, two or three dimensions results in the crystalline lattice is referred to as the unit cell. Or, repetitions of the unit cells in one or more dimensions in space result in an arrangement which is the crystalline lattice. Periodicity seems to be a common theme

in nature whether in space or time. Depending on the shape and the symmetry of the f the unit cell unit cell, we can classify and characterize lattices. The specifications o must include the length of the three sides or edges of the unit cell, the angles between the three sides and the number and the locations of the various atoms in the unit cell. We are mostly concerned with three dimensional crystals in the present unit. The fourteen common types of lattices found in the seven crystal systems are: cubic (simple cubic, body centered cubic and face centered cubic), orthorhombic (simple, body centered, face centered and end centered), tetragonal (simple and body centered), monoclinic (simple and end centered), rhombohedral, triclinic and hexagonal. Some of the unit cells corresponding to the above lattices are shown in Fig. 5.1.

Simpk cubic

f~antaed cubic (fee)

hexrgd

-'4sd

athomanbic

Figure 5.1 : Different types of unit cells.

SAQ 1 :
Draw the unit cells for the following lattices. (i) body centered cubic (ii) face centered ortho'rhombic and (iii) body centered tetragonal. What is the difference between a unit cell and a lattice?

Due to the long range ordering and symmerty of crystalline lattices, these structures are easier to characterize. With a reduction in symmetry and the long range order, a wide range of other structures such as glasses and liquid crystals emerge and these are a little harder to characterize. The characterization and synthesis of materials of lower dimensionality (i.e., in one and two dimensions) and of a finite extent (as against an infinite lattice) has a profound influence in technology since some of these materials are known to have unusual properties such as high temperature superconductivity. One of the earliest methods of studying the structure of crystalline solids is X-ray diffraction. In this method, X-rays of varying wavelengths and angles of incidence are made to interact with the surface of a crystal. The diffracted light emerging from the surface is analysed by plotting its intensity as a function of the angle of the incident beam. The results can be used to identify the type of the crystalline face on which the X-ray beam is incident and also calculate the distances between the atoms

or the ions of the adjacent layers, and thereby, the unit cell dimensions. In Fig. 5.2, the arrangement of atoms in a cubic lattice and the diffraction pattern arising from it are shown. Rays incident at two angles are shown. The rays R, and R, are incident on the surface at an incident angle 8. Each plane reflects only a part of the incident light. The first two layers, layer-1 and layer-2 contribute significantly to the intensity of the diffracted light and the contribution of the inner layers to the intensity of the diffracted light progressively diminishes.

Non-CovalentIntcradbm~:

layer
1

platform

a) X-ray diffraction from the surfaa o f a cubic lattice.

b) Schematic diagram of an X-ray diffractamrcs.

Figure 5.2

Consider the interference pattern between rays abc and defgh. Upto the points b and e, the two rays are in phase. The phase difference between the two rays at the points c and h respectively is due to the path difference efg. Both e f and fg are equal to d sin 8, where d is the distance between the two adjacent layers. If this difference is equal to nh, where n is an integer and h is the wavelength of the X-ray, i.e., 2dsin8 = n h (Bragg's Law) [5.la] then, a maximum in the intensity of the diffracted light I( 8 ) appears at the angle 8 . At other angles such as 8 ' for the ray R,, the intensity is not a maximum unless the Bragg condition (Equation 5.1) is satisfied. From the maxima in the curves of 1 ( 8 ), the value of d can be ascertained.

In the single crystal X-ray diffraction, the crystal is mounted on a platform and the angle 8 is varied by rotating the platform. Crystals can be cut along varius planes and a method is needed for labelling various planes of the crystal. These labels are the Millerindices of a plane.

Figure 5.3 :a) A few planes in a lattia. b) 'lhe coordinate system C) 'The Miller Indices ot a few planes.

Atom & Molecules

Consider the plane labelled 1 in Fig. 5.3(a). It cuts the x , y and z axes at 2a, a, and infinity respectively. The plane is parallel to the z-axis which is perpendicular to the plane of the paper, and therefore, the intersection of this plane with the z-axis is at infinity. The reciprocals of these numbers multiplied by the respective edge le~~gths (a, b and c).are 112, 1 and 0. To obtain the smallest integer set, multiply each number by 2 to obtain by (1,2,0). These are the Miller indices of this plane. The Miller indices of plane 2 are also (1.2.0). The Miller indices of planes 3 and 4 are (1,3,0) and (1,1,0) respectively. In Fig. 5.3 (c), the 111 plane is shown. It is seen that this plane intersects all the three axes at the distances a, b and c respectively from the origin, which are the dimensions of the unit cell. There are useful formulae which relate the distances between adjacent hkl planes and the unit cell edge-lengths a, b and c of the lattice. From the interplanar distance, the unit cell dimensions and the distance between the atoms can be derived. As an example, the distance between the adjacent (hkl) planes, h,,,, of a cubic lattice is given by

5.3 IONIC BONDING AND LATTICE ENERGIES

When sodium and chlorine react, the most stable product that can be obtained is the ionic NaCl lattice. In this lattice, the constituent units are Na' and C1- ions. The lattice is more stable than a diatomic NaCl. The process of the formation of the lattice from the elements Na and C1, in their normal states at room temperature may be decomposed into the following steps. Na (s) + M C12 (g) Na(g)+Cl(g)

Nu+ (g) + C t (g) + NaCl (s) The resultant of the above three steps is

-+

1 NO (8) + Cl (g) ;A H = AHvap(Nu) + j D (C12)

; A H = I(Na)-A(Cl)
;A?I=U

L5.2~1 [5.2b] [5.2c]

Na+(g)+Ct(g)

and the enthalpy change for this process is the enthalpy of formation of NaCl from its elements. Since enthalpy is a state function (i.e., depends on the initial and final states and not on the path), the following relation holds.

The formation of NaCl (s) from the elements is an involved process and the visualization of this process as a sum of steps brings out the role of the lattice energy U , defined here as the energy change in the formation of the lattice from the constituents of the lattice. In the present case, the constituents are Na' and C1-. AHvq(Na) is the energy required to vapourise metallic sodium, D is the dissociation energy of Cl,. I is the ionization energy of Na, i.e.. the energy change for the process Na -+ Na' + e, A is the electron affinity, which is the energy change for the process X---+ X + e. All the quantities may be taken in units of ~callmol or kJ/mol. Substituting the respective values in the above equation for the NaCl lattice, the above equation becomes (in kcallmol)
A Hf (NaCl) = 24.14 + 28.56 + 118.4 - 83.4 + U
.

[5.51

How do we obtain the lattice energy? To get an estimate of this value, consider the intkraction energy between a positive ion and a negative ion, with Na' and C1- as an example. At large separations between the ions, the only interaction is the Coulombic attraction (or repulsion between two positive ions) - e2/r where r is the separation between the ions. At short distances, interionic repulsion dominaies because the electronic clouds of the'two ions overlap 'too closely'. When two elerons ceme yery close to one another, there is a very strong repulsion between them in addition to the Coulombic interaction and this is a quantum mechanical effect. When two

I I

1I

electrons are very close to each other, it is not possible to distinguish them, because such a distinction implies that their positions are known more precisely than permitted by the uncertainty principle. Furthermore, if the two electrons occupy the same region of space, they will have all quantum numbers identical and this will be a violation of the Pauli exclusion principle. This repulsive interaction is referred to as the exchange interaction and at short distances, it is severe1 orders of magnitude larger than the Coulomb repulsion. At large separations, the exchange interaction decays to zero very rapidlyand the main i.~ter;ction at large distances is the Coulomb interaction. The attractive, repulsive and the total energy of interaction between two ions is shown in Fig. 5.4.

Non-CovalentI n h sctions

Figure 5.4 :Interactionenergy between two oppositely charged ions as a function of the interionic separation.

At the distance r,, there is a minimum in the potential energy and if only one Na' and one C1- were present, they would be found at this separation. This distance is found to be 2.38 A for gaseous NaCl. This means that the gaseous NaCl will have an energy of -e2/2.38 = -139.3 kcal/mol relative to the separated ions, taking only the Coulombic energy into account.

SAQ 2 :
a) b) c) Why is a factor of 112 used in Eq. [5.2a]? Draw the interaction energy as a function of r between two like charded ions. Using Eq. [5.2b], obtain an estimate for the formation of a diatomic NaCl from gaseous Na and C1. In this calculation, assume that the short range interactions between Na' and C1- are negligible.

A question that now naturally arises is that, when the diatomic NaCl is even more stable than H,, then, why is it that NaCl occurs in nature as a lattice? How does the extra stabilization of the lattice come about? To understand this, consider a one dimensional NaCl lattice with a unit cell of length r,.

Figure 5.5 :A one dimensional lattice of NaCl.

Atom & Molefulea

Consider the interaction enzrgy of the central Na' ion with all the other ions using a point charge model. The two nearest neighbours contribute an energy of -22/ro. The two next nearest neighbours contribute +2e2/2r0.The -ve ions contribute towards attraction and the +ve ions contribute towards repulsion. Summing up all the terms, we have,

The sum in the square brackets is greater than 112 since the first term is 112 and every one of the remaining terms is positive. The value of U can be shown to be = -1.4 e2/rowhich is larger in magnitude than the Coulombic attraction in the Na' C1pair at r,. Therefore, the one dimensional crystal is stabler than the diatomic NaCl. The calculations of the sum of the Coulombic interactions can be extended to three dimensions and the sum may be expressed as -M e2/ro.The constant M is different for different lattices and is characteristic of a given lattice. This constant is called the Madelung constant. For the FCC lattice of NaCl, the value of M is 1.75. The lattice energy of NaCl, taking into account the Coulombic interactions alone is

- 1.75 x (4.8 x 10-loesu)'

2.8 x 1@ cm

- 1.44 x 10-l1 ergs = - 207 kcaVmol

To these Coulomb interactions, we need to add the contributions due to the short range repulsions. The repulsion terms, as shown in Fig. 5.4, mqy be expressed either as exponential terms, Beq". or as power law repulsions as CIS. Here, B, C and n are constants which are to be determined for each system. The values of n range from 9 to 12. The values of B and C may be expressed in terms of M and r, as shown below. At the distance r = r,, when the lattice is r ~ ~ ostable, st the total potential energy is dU minimum and therefore, -= 0. Taking CIS as the repulsive term, U becomes, dr

and therefore, C =

Mey-I
n

,and

Since the value of n (commonly referred to as the Born exponent) is approximately equal to 10, the lattice energy U differs from the Madelung energy U, by about 10%. We would expect the U of NaCl to be about - 207 + 20.7 = - 186.7 kcallmol and the experimental energy is - 183 kcallmol.

In most iolric solids, the ions are packed very close to one another so as to maximize
the lattice energy. What is the fraction of the total available space that is occupied by the ions in a close packed lattice? To answer this question consider the faces of the three commonly occurring close packed lattices: (a) a simple cubic lattice (b) a body centered cubic lattice and (c) an FCC lattice. The faces of the three unit cells are shown in Fig. 5.6. The calculation of the packing fraction in the FCC lattice is shown below.

Example 1:
Calculate the packing fraction in a FCC lattice.

Solution: In the FCC unit cell, there are six atoms at the six faces of the unit cell. At each comer of the unit cell, there is an atom. A face is shared between two adajacent unit cells and.therefore each unit cell conhins half the volume of the atom from this face. Since there are six faces. they contribute a volume corresponding to three atoms to the unit cell. Each comer is shared by eight adjacent unit cells. Thus, the eight comer atoms together contribute a volume equal to that of one atom to the unit cell. Therefore, an FCC unit cell contains four atoms. The volume of the unit cell is equal to a3, where a is the edge length of the unit cell. The packing fraction of a lattice is defined as the ratio of the volume occupied by the atoms in the unit cell to the volume of the unit cell. To calculate the volume of the atoms in the unit cell, we need to know the relation between the radius of the atoms and the unit cell edge length. From Fig. 5.6(c), if r is the radius of the atom, then, using the Pythagoras theorem, (4r)' = a2 + a2 = 2 2 Therefore, The volume of four atoms

The volume of the unit cell = a3 The packing fraction is therefore = 0.74

(a)
Figure 5.6 :

(b)

Faces of unit d s showing h e closest packing in' (a) a simple cubic lattice. @) a BCC lattice and (c) a FCC lattice

Example 2: Atomic A1 (at. wt. = 26.98 glmol) crystallizes into an FCC structure with a density of 2698 kg/m3. X-rays of wavelength, 0.1537 nm, when diffracted from the (1 11) planes of this lattice, gave a maximum intensity at an angle of 19.2'. Calculate the Avogadro number, using the above information. Solution: To solve this problem, we need to calculate (a) the edge length of the unit cell, (b) the volume of the unit cell, (c) volume of one mole of Al and (d) the number of unit cells in the volume occupied by one mole of Al. From Bragg's law,

From Eq. 15.lbI. d =

$; Therefore, o = (3)'" x d = 0.40475 nm

Atom & Molecules

density =

mass = volume

p = 2698 kg/m3
kglmol

molecular weight of A1 = 26.98 x

Volume of I mol of Al = 0.02698 m3 = 0.00001 m3. 2698 The volume of a unit cell = (4.0475 x 10 m3 The number of unit cells in 1 mol of A1 = 1.5081 x loz3 Each unit cell in an FCC lattice contains 4 atoms. Therefore, the number of atoms of Al in one mol, or the Avogadro number = 4 x 1.5081 x loz3 = 6.0326 x loz3

r
(

SAQ~:

Calculate the packing fraction of a simple cubic lattice.

discuss the nature of molecular crystals Before golng to the next section, let us br~efly or van der Waals crystals. In molecular crystals, the individual units in the structure are molecules or atoms; not ions. The forces that bind the molecules are not as strong as in metallic solids (c.g., Cu) or as in covalent solids (e.g., diamond). Common examples of molecular solids are the crystalline phases of the inert gas elements, crystalline iodlne and crystalline benzene. The arrangement of the lattice is determined largely by packing considerations. The attractive as well as the repulsive forces between the molecules are very weak and the molecules largely retain their identity, very much as in the case of a liquid. Since the bonding forces are weak, the temperatures at which these solids melt or sublime are low in comparison with the corresponding values for the metallic or covalent solids. Molecules which are nearly spherical, such as HBr, H,S. and CH, form cubic close packed structures with each molecule having 12 neighbours. The arrangements are thus largely determined by packing considerations. The rare gas atoms form FCC close packed structures. In this arrangement, layers of atoms are stacked in the following way. First, a close packed layer of atoms is formed in a plane. Each spherical atom is surrounded by six atoms in a hexagonal arrangement. The second layer is placed on the interstitial spaces of the first layer. The third layer of atoms is placed at the interstices of the second layer. If the locations of the atoms of the third layer are exactly on top of thc atoms of the first layer, then we have a hexagonal close packed (HCP) lattice. If the third layer is placed on alternate sites above the second layer, not corresponding to the sites vertically above the atoms of the first layer, then we get a face centered cubic (FCC) close packed structure. Both the HCP and FCC closed packed structures have the same packing fraction. The packing of the rare gas . atoms is thus similar to the packing of hard spheres. In a simple cubic lattice, the second layer will be placed directly on top of the atoms (and not at the interstices) of the first. The interstitial space 1s not used for accommodating the atoms, thereby, resulting in a lower packing fraction in the case of a simple cubi'c lattice. The crystals of diatomics such as I, and triatomics such as CO, are not closely packed and these fotm varied structures. Among large molecules, rod like molecules like polyenes tend to lie in parallel rows close to one another, forming sheets. Flat molecules tend to stack into columns giving rise to needle-like crystals. Large

molecules tend to maximise the area of contact with the adjacent molecules. In aromatic compounds, the hydrogens of one molecule are near the K-electroncloud of its neighbours. The vibrational and electronic properties of molecular crystals reveal a number of properties of the individual molecules themselves.

Non-Covalent Interactions

5.4

INTERMOLECULAR FORCES

At room temperature (25" C or 298 K ), a very large number of molecules are stable and mobile and this is one of the reasons why our life span is around one hundred years. At high temperatures such as 3000 K, the thermal energies are of the order of (312) kBT which is equal to 0.39 eV or 45 kJ/mol. At this temperature, molecules with strong bonds such as Hz and N, are stable and will exist as diatomics, although the percentage decomposition of these molecules into atoms begins to increase. However, at such high temperatures, very few liquids will exist. In fact liquids exist only between the triple point of the material and its critical temperature. This is because, above the critical temperature, the thermal energies are too strong to be countered by intermolecular forces. We may define intermolecular forces as the interactions between otherwise stable molecules, wherein, the individuality or the identity of the molecules is intact. They are generally in the range of 0 to 1 kcal/mol, which is an order of magnitude less than the intramolecular forces such as the bonding interactions between the H atoms in H,. In bonding interactions, the chemical properties of the species change considerably. It is these intermolecular forces that are responsible for the formation of liquids and nonionic and non-covalent solids. The reason for excluding ionic solids is only to point out that the ionic interactions are not only as strong as the covalent interactions but also because, the ionic interactions extend to very large interparticle separations.

the formation of liquids and

SAQ 4 :
What are the critical temperatures, critical pressures and the triple point temperatures of (a) water, (b) ammonia, (c) benzene and (d) CCI,F, (freon) ?

In Units 3 and 4, we learnt that the binding energies between molecules may be determined experimentally and that the origin is quantum mechanical. These interactions (best exemplified by the plot of the electronic energy vs internuclear distance shown in Fig. 3.6 ) may be contrasted with several other familiar interactions that play an important role in chemistry. These are the (a) ion - ion, (b) ion - dipole, (c) dipole - dipole, (d) ion induced dipole, (e) dipole - induced dipole and (f) induced dipole - induced dipole or the van der Waals or the dispersion interaction. The strength of these interactions decreases from (a) to (0.In addition, there are interactions which are popularly referred to as hydrogen bonding. The interactions (a) to (e) are classical (electrostatic) in their origin. The interactions in chemical bonds or in the dispersion forces are quantum mechanical in origin. It should be really emphasized at this point that the distinction between covalent and non-covalent interactions is actually arbitrary and the interactions between any two molecules at small separations should actually be calculated quantum mechanically. Nevertheless it has proved to be quite fruitful to study the interactions with classification such as the covalent and the noncovalent interactions such as (a) to (e) which are mentioned above. The classification such as (a) to (e) above is useful as well as more exact at large separations between the molecules. The interactions (a) to (0 play an essential role in the formation of liquids and solutions. We have already seen in Sections 5.2 and 5.3 how ionic forces contribute to the formation of solids. Now, we analyse the

Atom &Molecules

other forces in some detail and then find out how these forces can be used to study the structure as well as the dynamics of the liquid state.

In an electrically neuual system, the sum of all positive charges is equal to the sum o i all negative charges. If one of the two is in excess, then we get an ionic species such as ca2' or Even when the system is neuual. the centers of the positive and negative charges need not coincide. This may be readily seen by considering examples of molecules such as HC1 and H20. In HC1, the center of positive charge is closer to the H atom and the center of the negative charge is closer to the C1 atom since the later is more elecuonegative. The molecule is said to be dipolar or simply polar (i.e., it has two 'poles' corresponding to positive and negative charges) and possesses a dipole moment represented by $The magnitude of the dipole moment is given by the product of the magnitude of the charge and the charge separation. If a charge of +e and -e are separated by 1 A, then the magnitude of the dipole moment is e x 1 A = 4.8 x 10-l8 esu.cm. The quantity lo-'' esu. cm is called 1 Debye. The dipole moments of molecules range from 0 Debye to a few Debyes. The values of the dipole moments and some other molecular properties of a few molecules are given in Table 5.1. The dipole moment is a vector quantity. The direction of the vector is taken from the positive to the negative charge. The vector $is given by + $ = xi qi ri [5.10]

G-.

where q, is the algebraic charge and c i s the vector from the center of mass of the molecule to the location of q,. Instead of point charges, if we have a charge distribution represented by p(x, y, z), whose dimensions are charge per unit volume. the dipole moment is given by

+ Pr
where

I $1,

Y,2 ) p(x,y,z) h d y dz.

[5.11]

?=r(~,~,z)=xi+~j+zk

Analogous to the dipole moment zitwhich is a moment of first order in r, higher moments such as the quadrupole moment (second order in r), the octupole moment (third order in r) etc., can be defined. The net charge on a molecule or an atom is ~iothin but ~ the zeroth moment of the charge density, i.e., the integral of the charge density over the whole space. Given below is the xy component of the quadrupole moment Q,.

Q ,

= N x Y P(~,YJ) h d~ dz

[5.12]

There are six components of the quadrupole moment, namely the xx, xy, xz, yy, yz and zz. For a diatomic molecule, with the z axis as the molecular axis, only one component among the six is independent and is referred to as the quadrupole moment and is given by
Q = Zi qi ( 3

2 - t)/e.

r5.131

The division by e is so as to make Q have the dimensions of an area. The values of the quadruple moments of some molecules are given in Table 5.1 Another property of a molecule that plays a crucial role in intermolecular interactions is the polarizability. In a uniform electric field, a molecule is polarized by the field (i.e.. there is a displacement of the centers of the positive and negative charges of the molecule) and the induced dipole moment (to be distinguished from the permanent dipole moment) is given by

c=a 2

[5.14 ]

where a is called the polarizability. For small elecuic fields, the polarizability may be taken to be independent of the field. The polarizability is anisotropic in general, i.e., it is not the same in all directions. For example, the polarizability along or parallel to a bond in a diatomic, a,, is different from the polarizability in a direction perpendicular to the bond, a,. The polarizability is a measure of the extent to which the charge cloud of an atom or a molecule can be distorted by an external field. Tightly bound clouds of systems like rare gases and positively charged ions have low poluiubilities. Loosely bound or relatively diffuse clouds such as alkali metals or

negative ions have large values of polarizabilities. Molecular polarizabilities correlate fairly well with molecular volumes. Typical values of a are given in Table 5.1. The three values of the quadrupole moments for water correspond to Q,, Qyy,and Q,, respectively, with the molecule placed in the xz-plane with the z-axis being the molecular axis. The various interactions between two molecules will now be considered. We first consider the dominant interactions between two charged particles or molecules. If the particles carry charges q, and q, respectively, the main interaction energy between them is q, q,l ri,. This term has already been discussed in the section on ionic solids. The next interaction to be considered is the ion-dipole interaction. The interaction + energy between an ion of charge q, and a dipole of moment CL/ is given by - p , E, where E is The electric field due to the ion at the dipole $ This energy depends on the angle 0 that the dipole vector makes with the axis between the ion and the center of the dipole. The energy is given by

V(r,O)iomiiPok = - p , .E = -

++

qi cos 0/?

[5.15]

Table 5.1 : Values of dipole moments, quadrupole moments and polarizabilities for a few systems
-

A tom/Molecule

Dipole Moment

Quadrupole Moment

Polarizability

LiH

HF
Hz0 HCI
1.109

3.7

2.63

Atom & Molecules

SAQ 5 :
Prove the validity of the equation C5.151.

The next interaction to be considered is the ion-induced dipole interaction. If an ion i and species j are separated by a distance r, the interaction between the ion i and the . induced dipole moment in species j is given by

SAQ 6 :
Will the quadrupole moment of a molecule multiplied by the molecular size (e.g., the length) give the molecular polarizability? You may want to refer to Table 5.1; .

The next interaction to be considered is the interaction between two dipoles pi and p? In this case, the interaction energy depends on the distance r between the dipoles as well as the angles defining the orientation of each dipole. For the dipole i, the angles are 8, and cp,. For the dipole j, the angles are 8, and + The angle 8 is measured with -3 + +The centers of the two dipoles z a n d $are given by respect to the vector rij = ri - rj. <and ~respectively. The angles cpi and cp, are measured in a plane perpendicular to the plane of the paper containing the centers of the two dipoles. All the angles are measured in the anti-clockwise direction. The formula for evaluating this interaction is
) =

(7).

( - cos Oi cos 8, + sin 8, sin 8, cos (cpi - cp,) )

[5.17]

It may be verified from this formula that when two dipoles lie along a straight line and are parallel to each other, the interaction energy is negative. That is, the dipoles attract each other. This is similar to magnets or magnetic dipoles. It was already pointed out that an ion polarizes another particle whether the other particle is charged or not. In the same manner, a dipole also polarizes a molecule near it. It is, after all, the electric field of the ions or the dipoles that polarizes the surrounding molecules and induces in them a charge separation corresponding to an and the induced induced dipole moment. The interaction energy between a dipole (k) dipole caused by it is given by
Vdrpoltrnduceddipole

2r: This'interaction depends on the angle Ii of the dipole pi with respect to the distance vector r, or Gbetween the center of the dipole and the molecule which has acquired . an induced dipole moment. It is to be carefully observed and noted that the interactions gradually diminish in magnitude as we go in the sequence: ion-ior.,

pr

,(3cos28,+ 1)

L5.181

ion-dipoIe, dipole-dipole, ion-induced dipole and dipole-induced dipole. The other interactions such as the dipole-quadrupole, quadrupole-quadruple and so on, are even smaller in magnitude and are not considered in this discussion. All the above interactions were electrostatic in nature, in which atleast one of the species was charged. Now we consider an interaction - referred to as the dispersion interaction. The dispersion interaction is present between all molecules and atoms even when they are not charged and do not possess an electric moment of any order. These forces have a quantum mechanical origin and can not be explained by classical electrostatics. The dispersion force between atoms/molccules i and j arises from a correlated motibn of electrons of the species i and j. Although this interaction can be understood in terms of the wavefunctions of the systems 1 and j, a classically intuitive argument may be given as follows. Consider two molecules 1 and j that do not possess any electric moment. At any instance of time, although each of the system is neutral, i t will possess a transient dipole moment due to the instantaneous positions of its electrons relative to the nuclei. Of course, the time average ( i.e., an average taken over a long interval of time) of this instantaneous dipole moment is zero. This instantaneous dipole of molecule i can "polarize" the electronic motion of molecule j due to the field that it creates at the molecule j. Thus, the electronic clouds of the two molecules will execute a correlated motion. An uncorrelated or an independent motion of the two electronic clouds will not lead to the dispersion interaction. This argument is only pictorial and it should be remembered that a quantum mechanical result can not always be justified by a classical argument. You may want to draw pictures of instantaneous dipole moments of two nearby molecules to follow the above arguments. Try with two He atoms. The detailed calculations of the dispersion energy are quite difficult. Therefore, the result of a simple model will be given below.

Non-Covalent lnc;;?rl!ms

This formula brings out the role of polarizabilities aiand a, of the molecules involved in the interaction. The interaction is an inverse sixth power law. The coefficients of the interaction are system specific. Thbfrcquencies vi and v , are the i and j and are nearly equal to the frequencies of the characteristics of n~olccules lowcst clectronic transition energies of the molccules i and j. These attractive forces between molccules are known as the van der Waals forces. A quantitative estimation of the intermolecular forces ( ds agalnst the covalent or the chem~cal forces ) can be made in the studies involving molecular collisions and in the studles ol the liquid state. The magn~tude of these interactions may be calculated for a few typ~cat values of interparticle separations. The parameters taken in the following calculations are some of the typical values of parameters that occur in molecules. Thesc calculated values are given in Table 5.2. In this table, the ionic charges q, and 9, are taken to be +e and -e, where, e is the electronic charge. The dipole moments F~ and p , are taken to be 1.5 D. The polarizabilities are assigned the values 3 A3. The d~poles are all aligned along the vector r,,, so that all the angles 0, to cp, are zero. For b be discussed below, the parameters taken are &/k, = the Lennard-Jones potentla1 t 120 K and o = 3.4 A. The various intermolecular potentials discussed in this section are shown as a function of r in Fig. 5.7, and the values at r , = 5 and 15 A are shown in Table 5.2. Table 5.2 : Values of typical non-covalent interactions at a few interparticle separations. All values in kcallmal.
,

Type of interaction ion-ion ion-dipole dipole-dipole ion-induced dipole dipole-induced dipole Lemard Jones potential
-.

r,=5A

r,, = 10 A

r , = 15 A
- 22.14

- 66.4 1
- 4.15
-

0.52

0.019

- 0.79 - 0.01 24 - 0.0212

- 0.0098

- 0.000017
- 0.0000032

Atom & Mokcules

ion-induceddipole

Figure 5.7 : Various intermolecular interactions as a function of the intermolecular separation.

~ S A 7 Q:

Fill in the second blank column of Table 5.2

Two argon atoms or two methane molecules will not bind chemically to one another. Precise calculations of the interaction energies between such systems is very difficult; it is certainly harder than the calculation of the energy levels of the constituent individual molecules! The accurate calculation of these intermolecular energies is becoming possible only in the last quarter of the twentieth century. However, one may choose an empirical potential based on the potentials of related systems or by using typical values of molecular sizes and the van der Waals radii. Using these potentials, one may then test whether this potential is capable of predicting certain experimental features. What are the experimental properties that are related to intermolecular forces? This is indeed a very poor question! Indeed, every property of a collection of molecules (an isolated molecule does not react with anything) or the reaction between molecules is dependent on and is governed by the intermolecular forces. The question can now be framed better : Which are the most direct measures or the criteria that can be used in determining intermolecular forces? Consider the case of an ideal gas first. In this case, as is well known, there are no intermolecular forces. The equation of state is given by PV = nRT. The role of intermolecular forces is to give rise to deviations from the ideal gas behaviour. The first deviation from the ideal gas behaviour is contained in the second virial coefficient which is defined in the equation

OD

Non-Covalent Interadions

B = 2 n N,
0

[ 1 - a-"(r)'knT ]

dr

[5.21]

Thus, the deviation from the ideal gas behaviour, which is obtained from the pressure volume data, can be used to obtain an estimate of the intermolecular potential u(r). There are several forms of the intermolecular potentials and here, only the simplest forms that are useful in atomic and molecular systems will be given. The hard sphere potential is defined by UHS(r) =
00.

if r I a , [5.22]

=O,ifr>a The square well potential is given by U,&r)=m if r I a =-~,ifa<rIa+8 =O,if r > a + & The Lemard-Jones potential is given by
12

[5.23] [5.24]

Uii(r)=4~[[~)

(?)I

This is also called a 12 - 6 potential since r-l2 and r4 terms appear in it. In this between the interacting equation, a represents an estimate of the contact d~stance particles and E is the depth of the potential. The attractive term ( a ~ rincorporates )~ the correct form of the asymptotic, long range, dispersion interaction. The repulsive term ( c ~ l r )incorporates '~ the short range steep repulsion. This form is very widely used and the parameters of this potential for a few systems are given in Table 5.3. Table 5.3 :The parameters for the Lennard-Jones potentials. System Ar-Ar Kr-Kr Xe-Xe Na-Ar Li-Xe CH4-CH4

ElkB.(K)
145 20 1 288 72 175 145

a lA
3.8

3.6
3.9 3.8 3.8 4.2

It is noticed that the contact distances amongst various systems are very similar.

SAQ 8 :
Draw sketches of the hard sphere and the square-well potentials. In the Lennard-Jones potential, what is the value of r at which the potential is minimum?

Ingases and liquids there is a continuous movement of molecules due to molecular translations, vibrations and rotations. Rotations cause a significant reduction or

Atom & Molecules

damping of those intermolecular forces which depend on the relative orientations of the molecules. For example, if two dipoles A and B are parallel to each other, then the interaction energy is negative. Let us denote it by -X. If the two dipoles are antiparallel. then the energy is + X. If there was a free rotation of the molecules, then the angle averaged dipole-dopole interaction would have been zero! But we do know that the lower, or the more negative values of energy are more favoured than the higher values of energy at all finite temperatures. In fact, the probability of the occurrence of a given value of energy U, is proportional to exp (-U,/kBT). This factor is commonly referred to as the Boltzmam factor. (The absolute temperature is denoted by T and kBis the B o l t z m a ~ constant )The angle averaged potentials are temperature dependent, and for the dipole-dipole interaction, the angle averaged potential is given by
(UdiP-d,dav =

-1.5

PB)~/~BT~~B

[5.25]

Here ( , ) refers to angle averaging using the B 0 l t z m a ~ probability distribution. This averaging is also referred to as thermal averaging. In this averaging, each energy is weighted by the Boltzmann factor corresponding to that energy. While averaging over the orientations, the distance between the dipoles rAB is not changed. This orientationally averaged dipole-dipole interaction may be contrasted with the angle averaged (not Boltzmann average) dipole-induced dipole interaction which is given by, PB)/'AB [5.26] This later average is independent of temperature because the induced dipole moment is always parallel to the inducing field. The angle averaged dipole-dipole interaction (Boltzmam weighted), the averaged dipole-induced dipole interaction and the dispersion interaction possess the same functional form. They all vary as r4.
Udipoltinduceddipo[~a v
I

-( a ~ PA+

5.5 HYDROGEN BONDING

The term hydrogen bonding is invoked to explain certain special properties of some systems containing (1) a hydrogen atom bonded to a highly electronegative atom and (2) other electronegative atoms present in the vicinity of this hydrogen .atom which can significantly alter the electron density in the vicinity of the hydrogen atom without actually breaking the parent bond. Hydrogen bonding is present in a variety of systems ranging from water molecules to the DNA molecule and is responsible for a variety of properties of systems ranging from the strengths of fibers to even the process of genetic coding and human memory. To get an insight inlo the nature of hydrogen bonding, let us begin by considering the boiling points of a few liquids. The boiling point of a liquid is the temperature at which its vapour pressure is equal to the atmospheric pressure. A higher boiling point implies that the molecules of the liquid are held together more strongly. This is because a higher temperature is required to evaporate a sufficient number of molecules for the boiling to start. Molecules with large values of dipole moments and other higher order moments and molecules possessing large values of molecular polarizabilities are generally known to have relatively large values of boiling points. The polarizabilities generally increase with an increase in the number of electrons in the system. As we go from He to Ne, Ar, Kr and Xe, the larizabilities increase respectively as 0.203.0.392. 1.63,2.41 and 4.01 ( in 10-%O cm3units). The boiling points of these systems are 4.22,27.3,87.3, 119.9 and 165.1 K respectively. It is seen that the boiling points increase with an increase in the polarizability. Next consider a few relatively small molecules such as propane, dimethyl ether and ethylene oxide. Their polarizabilities are 6.4.6.0 and 5.0 (in lo-" cm3) respectively, which are quite close to one another. The dipole moments of these molecules are 0, 1.3 and 1.9 D and their boiling points are 23 1,248 and 284 K respectively. We observe that an increase in the values of the dipole moments of the molecules also leads to an increase in the boiling points. If the size of the molecules increases considerably, then the dispersion interactions become more significant and the role of the dipole moments and polarizabililies tends to become less important. In compounds containing hydroxyl, amino or imino groups and also in molecules such as HF, the boiling points are considerably higher than for the molecules

possessing similar values of dipole moments and polarizabilities. As examples. consider C,H,OH (ethanol) and C2H40(ethylene oxide). Their dipole moments are 1.7 D and 1.9 D and they have similar polarizabilities. We may expect ethylene oxide to boil at a higher temperature. But its boiling point T , is 284 K while the boiling point of ethanol is 352 K. HF with a value of a = 0.84 x lo-% cm3 and a p = 1.91 D has a boiling point of 291 K. Similarly methanol (a= 3.0 x cm3, p = 1.7 D and T, = 338 K), ammonia (a = 2.34 x lo-%cm3, p = 1.467 D and T, = 240 K) and water (a= 1.59 x 10-" cm3, p = 1.86 D and T, = 373 K ) show very high boiling points in comparison with other molecules with similar values of a and p. This additional interaction present in these molecules is much weaker than the usual covalent bonds and is slightly stronger than the van der Waals forces. The strengths of these interactions range from 3 to 10 kcallmol. This interaction is attributed to the association between the molecules by the formation of a weak bond, called the hydrogen bond that results when a hydrogen atom is sandwiched between two very highly electronegative atoms. The lengths of these bonds are in the range of 1.5 to 2 A. Some common examples are given in Fig. 5.7.

Non-Covalent Interactions

Figure 5.8 : Some examples of hydrogen bonds.

.li
'

Hydrogen bonding is not restricted to liquids alone. The fact that urea or boric.acid are solids at room temperature is attributed to the formation of extensive hydrGgen bonding. Aqetic acid molecules pair as dimers due to hydrogen bonding. In salicylic acid, an intramolecular hydrogen bond is formed between the hydroxyl H and the adjacent 0 of the COOH group. Hydrogen bonding is not restricted to small 2 . molecules alone. Indeed the utility of nylon as a fiber is due to hydrogeh bonding between adjacent chains of molecules. The dase pairing in the genetic material, DNA is also due to hydrogen bonding.

5.6 CLASSICAL MOLECULAR DYNAMICS

Having discussed the various forms of the potentials, it is worthwhile to consider how these forces inf-luencc the physics and the chemistry of the condensed matter. We shall use classical mechanics or Newton's laws, and not quantum mechanics in this

section. In a gaseous system. the average distance between two molecules is quite large, typically 50 A at room temperature. This large separation enables a nearly random motion between the molecules of the gas. If we consider solids and liquids, the molecules therein are very close to one another. The intermolecular distances are typically 3 to 5 A. In a solid, whether it is ionic. covalent, metallic or molecular, the lattice particles do not move away from heir equilibrium sites. Vv'hatever mobility remains, it is due to the presence of vacancies or defects.

In liquids, there is a continuous motion of all the particles and this motion may be
studied in a very simple manner if the form of the intermolecular forces is known. In the last two decades, an emormous effort has been put in developing the forms of the potentials between several systems, particularly those systems relevant to life such as proteins and nucleic acids. Several computer packages are available to study the molecular motions in such systems. In the present unit, the major features of such an approach will be illustrated by considering the system of liquid argon. Consider a collection of N argon atoms in a cubic box of edge length L. The number N is in the range 50 to 1000. A cross section of the box is shown in Fig. 5.9(a). The number N is so chosen that the value NIL^ corresponds to the actual density of liquid Ar at the temperature of interest. The initial configuration of the atoms in the system, i.e., the initial values of the coordinates of all the particles is determined by placing all the atoms randomly in h e box. They are placed in such a way that no two particles are too close to one another. Typically each particle is at least 2.5 A. away from every other particle. The random placement can be done by using a random number generator such as those present in most scientific calculators today.
'

a) A simul~~iorrcell f a m o l ~ l a dynamics. r Figure 5.9

b) Demonstn~ion of periodic bgundary condtions in two dimensions.

SAQ 9 :
Why are the initial coordinates so chosen that no pair of atoms is closer than 2.5 A? From the knnard-Jones potential energy diagram of Ar, see what will happen if two atoms of Ar come as close as, say, 1 A.

Having chosen an initial set of coordinates of all atoms, we need the initial velocities of all the atoms. The initial velocities are obtained from a Maxwell-Boltzmann distribution of velocities at the temperature of interest. According to this distribution, the fraction of molecules having velocities between v and v+dv is given by According 1 0 this distribution function, velocities which are very small or very large occur very rarely. This result is from the kinetic theory of gases. In thisdistribution formula, m is the mass of the gas molecule and T is the absolute temperature. The

assignment of velocities is similar to the assignment of positions except that the velocities are chosen from the plot offlv) vs v. This distribution function will again be used in Units 9 and 10.

Non

ralent Interactions

SAQ 10 :
Plot a graph'offlv) vs v for Ar at 150 K.

Given the initial positions and velocities, we would like to follow the path or the trajectory of all the atoms in the system. This is very much like following the future locations of all the planets in our solar system knowing the present positions and velocities of all the planets and the forces of interaction between the planets. Or even more commonly, the evolution of a satellite in the vicinity of the earth. The equations of classical mechanics or the Newton's laws of motion can be used for this purpose. If you begin wondering how we have come back to the use of Newton's laws suddenly, your worry is more than justified. But it has been found that Newton's laws can be used quite effectively to monitor the motion of atoms and molecules in a liquid. They can not be used to describe the motion of smaller objects such as electrons and other tinier particles. The method that is being described here is known as the molecular dynamics simulation. Having obtained the initial velocities and positions of all the atoms, we want to know the positions and velocities of the atoms after a time 6t. The value of 6t is in the range of 1 to 10 fs (femtosecond). In the fall of bodies from a height onto the earth's surface, we may recall having used the familiar formula, s = ur + at2/2. In this formula, s is the distance travelled by the object, u is the initial velocity, a , the constant value of the acceleration due to gravity and r is the time of travel. On and very near the surface of the earth, the acceleration due to gravity may be assumed to be a constant. But in the case of molecular motions, the acceleration on each of the molecules is continuously changing. This acceleration depends on the relative distances between the molecules which are continuously changing in a liquid. Consider the particles i and j. The interaction energy between the particles is given by
12

[5.28] where, rij is the distance between the particles i and j and ?= (xi. yi. zi) is the coordinaie of the particle i and ?the coordinate of particle j. The distance r, is given by [(xi - x,)~+ (yi - $)+ (zi - z,) 2] 1/2 The force on the particle i due to its interaction with particle j is given by the gradient of the potential. Recall that force multiplied by distance or the integral of F.ds is potential energy.

The particle i interacts with all the remaining particles in the box. The total force on particle i can be obtained by adding the forces on it due to all the other particles. ari The summation excludes the interaction of the particle i with itself; i.e., j is not equal to i in the summation.
j+i

F, (total) = C

au

SAQ 11 :
What is the formula for the total force on particle j ?

The new position of the particle i after a time &. (i,e., at the time t + 61) can be obtained from its positions at times t and t - & by the Verlet algorithm which is given below. r,(t+61) = 2 ri(t) - ri(t-61)

+ [Fl/m] 6 ?

[5.31]

and the velocity of the particle at time t is given by vi(t) = [ r1(1+6t)- ri(t - 61) ]/2& [5.32] Using this algorithm, positions and velocities of all the particles can be generated for several time steps (usually in tens of thousands). The configurations (collection of all the coordinates in a given time step) at all the time steps is referred to as the molecular dynamics trajectory. Molecular dynamics simulations have become an industry now and they provide a very powerful method for studying the structure and dynamics of liquids and solutions. Its relevance in the present unit is to bring out the critical role of intermolecular forces.

Example 3 : The Verlet algorithm, as outlined in the above discussion, allows one to calculate the new (or future) position of the particle if two of its previous positions (i.e., the present and the immediate past) are known. When one starts the simulation, only the present positions and velocities are known. How are the new positions determined in this case? Solution : In the beginning of the simulation I = 0, and only r,(O)s and vi(0)s are known. To obtain r,(0+6t) =r,(6t) use the familiar Newton's formula
r,(&) = r, (0) + vi (0) 61 + 0.5 Fi (0) 6t2/m

Example 4 : For a three particle system of Ar atoms, choose an initial set of configurations and velocities, take & = 0.05 ps, and calculate the first four steps in the molecular dynamics simulation of these three particles using the Verlet algorithm. Solution : The three particles are placed in the xy-plane as follows. One particle is at the origin, another particle is on the x-axis, 4 A from the origin and the third particle is on the y-axis, again 4A from the origin. The (x,y) coordinates of the particles are (0,0), (4,O) and (0.4) respectively. The initial velocities are arbitrarily chosen to be (0.0). (0.1) and (1.0) respectively. The unit of velocity is Alps. The unit of force is 10 JI(mo1. A). The MKS unit of force is Newton or Jlm. The values,of the components of the forces, positions and velocities are given in table 5.4. The time step of the simulation is 0.05 ps. Mass is taken in dmo! units.Tor Ar, take m = 39.95. The units of mass, length and time which are taken here as drnol, A and ps respectively are very convenient in these calculations. Verify that these units give energy in units of 10 Jlmol.

Non-Covalmt Interactions

Components Step Number 0 Forces Positions

Particle No. 1
X

Particle No. 2
X

Particle No. 3
X

55.47 0.0

55.47

- 68.3

12.8

12.8

- 68.3

Velocities 0.0 Step Number 1 Forces + 55.56 Positions 0.00174 Velocities 0.0347 Step Number 2 Forces +54.3 Positions 0.00695 Velocities Step umber 3 Forces Positions Velocities Step Number 4 Forces Positidns Velocities 0.0695 51.4 0.0156 0.1383 +46.3 0.0274 0.2044

SAQ 12 :
Repeat the calculation of the example for a two particle Ar system with starting velmities of (0,O) and (0, I) and ~nitial positions of (0,O) and (4.0) respectively~~ositions are in 8, and velocities in bps. The force on each particle at a given time is calculated from the positions of all the particles at that lime.

For the interaction bctween molecules, several empirical potentials have been constructed. Only one example of such a potential, namely the water-water potential is given here. It is very convenient to think of a molecule as a collection of interacting sites. The sites on a molecule do not interact amongst themselves, but interact only with the interacting siles of all the other molecules that are present in the system. The model presented here for water is known as the SPC (simple point charge) model. Three point charges are placed on each molecule. The charge on hydrogens is + 0.41 e and that on oxygen is - 0.82 e. Here, as usual, e is the magnitude of the electronic charge. In addition to this Coulombic interaction, there is a Lennard-Jones center on each oxygen. This means that the interaction betwecn oxygens of any two water molecules is given by where r is the oxygen-oxygen distance,'^ = 0.3428 nm (k~/mol)"'~ and A = 0.37122 meter. In addition to this nm ( k ~ / m o l ) ' Here, ~ . nm refers to nanometer, i.e., Lennard-Jones interaction between the two molecules, all lhc charges on one molecule interact with all th25harges on all the other molecules. In each water

molecule, the 0-H distance is taken to be 1 A and the H-O-H bond angle is taken to be the tetrahedral angle of 109"28'.

Figure 5.10 : The SPC model of water.

This model is found to yield the properties of bulk water such as the diffusion coefficient (which measures the extent of random motion in the liquid), the liquid structure and the water bulk density. The dipole moment of this model of water is 2.351 D. This is much higher than the dipole moment of an isolated water molecule which is 1.86 D. This indicates that the models that yield good bulk properties (or macroscopic properties) need not correspond to the exact results for the isolated molecules or the isolated pairs. This is an indication that the interactions in the bulk medium are significantly different from those of isolated molecules or isolated pairs. This is precisely the reason for studying the three-body and other higher n-body interactions. These are discussed a little later in the present unit.

SAQ 13 :
Show that the dipole moment of water in the SPC model discussed in this section is 2.35 D. How will you change the charges on H and 0 to get a dipole moment of 1.86 D?

You may have been wondering by now- "How does the simulation on a system containing at most a few' hundred particles correspond to the behaviour of a liquid?" Even a drop of a liquid contains about 1019molecules. It was found that if the central cubic box containing about 100 to 500 molecules is surrounded on all sides by identical boxes, extending to infinity, then we have a large system representing a liquid. If a particle of the cenual box enters a box to its right, then the particle in the left box, which is the image of the particle that has left the cenual box, enters the central box from the left. Such conditions are called periodic boundary conditions and are applied in all the three directions. Periodic boundary conditions in two dimensions are demonstrated in Fig. 5.9(b). It was discovered by 1960, that calculations performed on a hundred molecule system with periodic boundary conditions yielded properties that matched those of real liquids quite well. For . example, the average potential energy of the system which may be taken as the sum of all pairwise interactions present in the system, corresponded to the internal energy of a macroscopic system which will be considered in the next unit. It is thus possible to make the vital connection between molecular properties and properties of the bulk systems (i.e., thermodynamics and chemical kinetics to be studied in Units 5 to 10) containing an extremely large number of particles. The important features of the structure of liquids is illustrated in Fig. 5.11. Fig. 5.1 1(a) describes the structure of fused (liquid) sodium (labelled L) and compares it with the suucmre of solid sodium (labelled S). In a liquid, the atoms/molecules are in continuous motion. One typical arrangement in fused sodium

is shown in the upper part of Fig. 5.1 l(a). Let us call the atom R as the reference atom (any other atom can be used as the reference). We may think of the neighbours of R as constituting a set of nearest neighbours, a set of next nearest neighbours and so on. Starting from the atom R, let us count the number of atoms enclosed in a sphere of 4 radius r of volume - x?. This number, designated as n(r), is called the running 3 coordination number. The area under the thick curve in the lower part of Fig. 5.1 1(a) (labelled L) gives this n(r) for liquid sodium. For example, n(4.0 A) = 6.6 and this is the area shown by the dots in the figure. As r increases, n(r), the area under the curve 47tr2 pg(r) also increases, because more particles get included in spheres of larger radii. Here, p is the bulk density of the liquid in units of particles per A3. The dashed line labelled U is the plot of 4 x 3 vs ~ r. This is a parabola, as expected. Deviations from this curve are the deviations from a fully uniform or random distribution of particles about the reference atom R. These deviations, shown by the Line L specify the structure of the liquid. The vertical dashed lines in figure with the heights of 8,6, 12, 24, 8, 6 and 24 at values of r near 3.7,4.4, 6.1,7.1, 7.4, 8.6 and 9.3 A respectively indicate the number of atoms present at these distances from a reference atom in the solid. At the remaining distances, e.g., between 3.7 and 4.4, between 4.4 and 6.1 and so on, no particles are present in the solid. This is due to the rigid, regular arrangement of atoms in the solid. In a liquid, although this rigidity is lost due to the movement of molecules, a structure is still prevalent and this is shown by the peaks of the curve L. The quantity 4 x 3 pg(r) is similar to 4 x 3 studied ~ in Units 2 and 3. The quantity 47tr2 pg dr gives the number of atoms, dn(r) contained in a spherical shell of radius r and thickness dr. A cross-section of such a shell is shown in Fig. 5.1 1(b). The center of the shell labelled 0 corresponds to the reference atom R of Fig because the distance r is measured 5.1 l(a). A reference atom has to be present at 0, with respect to a reference atom. The quantity 1 1 g(r) = - [n (r + dr) - n (r)]/4xr2dr = - [dn (r)/4m2 dr] P P is called the pair distribution function of the liquid. The bulk density of the liquid is given by p = N/V where N is the number of particles in the liquid and V the volume. For very small values of r, the spherical shells about the reference atom do not contain any particles and therefore, dn(r) = 0 and so is g(r) = 0. For large values of r, the spherical shell is very much like the bulk of the liquid and is not influenced by the reference atom and the 'local density' of the shell, namely

Non-Covalent Interaction

[n (r + dr) - n (r)]/(4x? dr) becomes the bulk density p and thus g(r) becomes unity. The pair distribution functions in liquids are extensively studied and they give considerable information about the structure of liquids and they can be related to the macroscopic (or bulk) properties of the liquid. In Fig. 5.1 1(c), the oxygen-oxygen, the oxygen-hydrogen and the hydrogen-hydrogen pair distribution functions (i.e., g,(r), go,,(r) and g,,(r)) in liquid water are shown. In gdr), i refers to the atom i in a reference molecule, and j refers to the atom j at a distance r in any other molecule. You are asked to give the ang go&) in analogy to g(r) for liquid physical interpretations of goo(r), gHH(r) sodium which could have been explicitly written as g, ,, (r). The results of the molecular dynamics simulations on water are shown in Fig. 5.11(c). The 0-0.0-H and H-H radial distribution functions are shown in the figure. These results are obtained by averaging over tens of thousands of configurations in a system containing more than 100 water molecules. The graphs represent the average structure of water. The 0-0 running coordination number has the following meaning. If we take the oxygen atom on one of the water molecules as the center of the coordinate frame and count the number of oxygen atoms in a spherical shell of radius r aroulrd it, the number obtained is the running coordination number noo(r). This is obtained for various values of r. Since all the water molecules in a liqu~d are of equal importance, the value of n(r) is averaged over all the water molecules.

Figure 5.11 :

a) The structure of liquid sodium. b) A section of a spherical shell of radius rand thickness dr. c) The 0 - 0 . 0 - H and H-H pair distribution functions in water.

SAQ 14 :
What is the meaning of the 0-H and H-H coordination numbers? Define them in analogy to the 0 0coordination number. From the discussion above, give a simple definition of the pair distribution function in tern's of the local density in a spherical shell of radius r. In the graph 5.1 1(a), what are the number of sodium atoms surrounding a reference atom in a spherical volume around it of radius 8 A in (a) liquid sodium, (b) solid sodium and (c) sodium of the same density as the liquid-but behaving as an ideal gas ? For this part, compute the areas under the dashed curve (U, or ideal gas) and the smooth curve L (for the liquid). For the solid, count the total number of atoms (vertical dashes) at distances (from the reference atom) less than 8 A.

SAQ 15 :
Develop the formula for the total energy of an N partiele system. Start with 2, 3 and 4 particles and then extend to an arbitrary N.

Non-Covalent Interadions

In this section, as well as in the previous sections, an underlying assumption that is made is that the total energy is the sum of all distinct pairwise interactions. This is correct for point charges alone. In this situation, the force on a particle i can be calculated as the sum of the contributions from all the remaining particles. But pairwise addititivity of forces or energies is only an approximatio~ for molecules zn6 this needs to be corrected by the inclusion of the three body d3d the other higher n-body ( n > 3 ) forces. The three body potential for a system of 3 particles is defined by the following equation.
I/rold

u12(r12)

+ ~ 2 3 ( ~ 2+ 3 )~ 1 3 ( ~ 1 + 3 )u123(r12?r23*r13)

[5.35]

The last term above is the three-body potential. It is that part of the potential that can not be expressed in terms of the pairwise interactions alone. Consider the example of three particles i , j and k. If only pairwise interactions are present, then the total energy is given by

We have already seen that molecules have non-zero values of polarizability. The polarization energy of the molecule i when only molecule j is present, added to its polarization energy when only k is present, is not the same as the polarization energy of i when both j and k are simultaneously present. This is because, the presence of molecule k alters the very moment of the moleculej which is to polarize the molecule i. Similarly, j alters the moment of k which polarizes i. The same argument holds for the polarization energies of molecules j and k. The dispersion energy of three molecules or atoms is thus given by

where,
cGk = u ( 3 ~ 0 Oi s cos 8, cos Ok + 1)

Here, a is a constant whose values are of the order of molecular polarizabil~ties and c,, has a value similar to the values of c,)s. The formula for U , , , is called the Axilrod-Teller formula. It gives the contribution that can not be expressed in tsrms of the sum of pairwise interactions alone. The angles 8,9, and 8, are the internal angles formed by the three atoms and are shown in Fig. 5.12. The analysis presented above can be extended to higher terms such as the 4-body interactions and so on. But these are progressively smaller in magnitude. In general, the three body contributions are of the order of 10%of the total energy of the system. The n-body interactions are important especially when the distances between the interacting molecules are small. At large distances, the two body apptcximation is quite good. The computation of three body forces is quite laborious and therefore, it has been a common practice to look for effective two-body potentials which inclade

i
Figure 5.12 : Intemai angles formed
by the three atoms

Atom & Molecules

the contributions or tne n-uouy potentials in an average way. For very accurate calculations, we have to include the 3-body forces.

5.7 SUMMARY
In this unit, our studies began with common non-covalent interactions such as the ionic interactions that are present in crystalline lattices. This lead to a characterization of the common types of crystal lattices. The structure of these lattices can be obtained from the diffraction experiments. The dimensions of the unit cells can be obtained by analysing the diffraction patterns arising from the various faces or the planes of the crystal. The study of the non-ionic interactions lead to the forms of different types of intermolecular forces that are important in determining the collective behaviour of molecular systems. It is well known that if intermolecular forces were absent, then the earth would be a collection of ideal gases, and among other things, this unit would not have been there! Different intermolecular interactions have been discussed, their formulae have been provided and examples have been given so that you are able to use them in calculations. All of technology (other than nuclear technology) is approaching the molecular level and a good understanding of intermolecular forces is becoming more and more critical with the passage of time. As a special case of intermolecular interactions, hydrogen bonding was discussed and several examples of hydrogen bonding were given. They range from liquid water to DNA. Molecular properties, along with inter molecular forces determine the macroscopic properties of bulk systems such as solids and liquids. The use of intermolecular forces in understanding the behaviour of liquids was illustrated through the technical details of the molecular dynamics simulations. The Verlet algorithm for classical molecular dynamics was given and examples of the use of the algorithm for oversimplified systems were given. For larger systems a computer is necessary, but often not sufficient!

5.8 GLOSSARY
Ionic Solids Unit Cell Crystalline Lattice : Lattice Energy Bragg's Law
:

Solid form of matter wherein the constituent species are ions. The smallest arrangement of atoms/molecules, which on repetition in space yields the lattice. The structurelarrangement that results when.the unit cell is repeated in space. The energy change for the process of ions leading to.the. formation of the Lattice.
2d sin 8 = nh where d is the distance between layers of a crystal, 8 is the angle of incidence of X-rays of wave length h and n is the order of reflection which is usually
1.

Packing Fraction

The fraction of volume of the unit cell occupied by atomslmolecules of the unit cell.

the

Intermolecular Forces :

The weak forces between molecules such as the dipole-dipole, ion-induced dipole, and dispersion forces which do not lead to the formation of a chemical bond between the molecules. where q, is the charge at the ith site in a molecule and 3 s the distance vector of that site from the center of mass.
= 2, q,

The extent of distortion of a molecular charge distribution in the presence of an electric field defined

through ~ , , , , = , , a E, where $is the induced dipole moment and ?is the applied field. Hydrogen Bonding
I

-9

Non-CovalentInteraci;;?c

i
I

i
Molecular Dynamics

: The interaction of the order of 3 to 10 kcallmol, resulting when a hydrogen atom is bonded to an electronegative atom (such as 0, N, F) and placed near another electronegative atom about 2 A away from it. : The motion of molecules determined by the nature of forces acting on the molecules : The algorithm x,,,
=2x ,
- x,,,

11

Verlet Algorithm

+Fx (602 where the m

new position is determined in terms of the current and old positions, the force at the current position and the time step 6t

SPC Model of Water

: A model of water wherein the charges on 0 and H are -0.8476e and 0.4238e respectively, r,, = 1 A, and the

HOH angle is 109" 28'. The oxygen atom of one water molecule interacts with the oxygen atoms of the other water molecules through a 12-6 potential. The interaction energy between two SPC water molecules consists of one 12-6 potential between the oxygens and 9 Coulombic terms.

5.9

ANSWERS TO SAQs
To draw the BCC unit cell, add an atom at the center of the simple cubic unit cell. To obtain an FCC unit cell, start with a simple cubic (SC) unit cell and add a particle at the center of each face of the SC unit cell. Start with a tetragonal unit cell and add a particle at its center to obtain a body centered tetragonal unit cell. A lattice is an infinite array of points in space. If at each lattice point of a periodic lattice, a unit cell of atoms/molecules (which is the repeating unit) is placed, then the result is a crystalline lattice.

SAQ 2.
a) b) C) The factor of 112 is used because we need one mole of chlorine atoms for which we need only 112 mole of C1, molecules. For two like charged ions of charge Z,e and Z,e, the function tc plot is ~ , ~ , e ~This / r . is similar to plotting the function llr vs r. Assume a bond length of 2.8 A. The energy change for the process Na (g) + C1 (g) ---9 Nu+ .... C t is

e2 1394 4 kJ/mol = - 368 U/mol I (Na) - A (Cl) - -= 494 - 364 2.8 A 2.8
To get the energylenthalphy of formation of gaseous NaCl from Na (s) and C1, we have to add the heat of vapourization of Na and the corresponding dissociation energy of Cl,.

SAQ 3.
In a simple cubic lattice, there is one atom per unit cell. If the radius-of the sphere is r, and the edge length of the cube is a, then, the volume of the sphere and the volume of the unit cell = a3.The packing fraction
R- -0.523.

.6

.-

Atom & Molecules

S A Q 4.
Molecule
Water ammonia benzene CCl,F,

Critical Temperature, K 647.3 405.5 561.6 384.2

Critical Pressure (atm) 218.5 111.5 47.9 39.6

Triple Point, K 273.16 195.5 278.5 118.3

S A Q 5.

The dipole moment = q d. The interaction energy U = - -++ y . E and R = distance between the ion and the center of the dipole.

(R +;6 '( , cos 9)' where -Q = charge on the ion. Expand the denominatoras a Taylor series.
+

u=-

Qq ( R - (6/2) cos 9 )

Since - << R, keep only the first term 2

L

The force on the dipole with its moment presumed to be at its center of charge

S A Q 6.
No. Two molecules can have the same quadruple moment and size and yet have different polarizabilities.

SAQ 7.
Coulombic, ++ : + 33.2 kcallmol, ion-dipole : -1.04 kcallmol, dipole-dipole: -0.065 kcallmol, ion - induced dipole : -0.05 kcallmol, dipole-induced dipole: 0.0002 kcallmol, dispersion : -0.002 kcallmol, LJ potential :-0.00147 kcallmol.

S A Q 8.
The sketches follow directly from eqn. 5.22 to 5.24. Confirm your sketches by referring to a reference book.
+ -

dU r = 2'16 a when U = U,,, = -&. Obtain this by setting -= 0 for the Lemarddr Jones potential.

SAQ 9.
The values of a for most systems are over 3 A. If the initial coordinates of two Lhcn the forces 011 ihc molecules will be particles are too close (say 1 or 2 very large and in the next time step, these molecules may come closer to some other moleculc, resulting in larger forces and the molecular dynamics trajectory will deviate far from equilibrium.

a),

SAQ lo.
Choose the units of (112) r n 2 the same as that of kT.Find the value of v for which f (v) is maximum and calculate f (v)for values of v in steps of 0.25 v, wheref (v,) =f*,,

Non-CovalentInteractions

F on particle.j = - C
i= 1

= ,
arj

a u,

for pairwise forces.

SAQ 12.
Components
Step 0 Forces Positions
e

Particle No. 1

Particle No. 2

Velocities Forces Positions Velocities

Step 1

Step 2

Forces Positions Velocities

Step 3 .

Forces Positions Velocities

'

SAQ 13.
The OH bond moment in SPC water is 0.4238 x 4.8D = 2.034 D.The dipole moment of water = 2 x 2.034 x cos - x 109.5~ = 2.35 D. To reproduce the 1 86 =0.3354 e actual dipole moment of water, the charge on H has to be 0.4238 x ' 2.35 and that on 0 has to be - 0.6708 e. Repeat this calculation with the charge o n . oxygen = -0.82e and the charge on hydrogen 0.41e..
\ /

r:

SAQ 14.
The pair distribution function g(r) is defined as the ratio of the'lccal density of particles at a distance r (from a given particle) and the bulk density. As r -+ 0, g(r) + 0 and as r + -, g(r) + 1. In fact, the structure of most normal liquids is short ranged and g(r) + 1 beyond 10 to 15 A. For ideal gases, g(r) + 1 for all values of r. The OH running coordination number no, (r) gives the number of 4 H atoms in a volume of - ~c3 surrounding an oxygen atom. 3

SAQ 15.
Label the particles as 1,2, ...N. Assume only pairwise interactions. For a two particle system the interaction is u12 (r). For a three particle system, it is u12(rI2)+ u13(r13)+ UZ3 (r23).For an N particle system, there are N (N - 1)/2 terms and the formula is

SUGGESTED READINGS
Atomic Spectra and Atomic Structure, G. Herzberg. Dover Publications, Inc. ,New York (1944). Textbook of Physical Chemistry (Second Edition), S. Glasstone, Macmillan Student Edition, New Delhi (1974). Fundamentals of Physical Chemistry, S.H.Maron and J.B.Lando, Macmillan Publishing Co. , Inc. ,New York. (1974). Introduction to Modern Physics. F.K.Richunyer ,E.H.Kennard and J.N.Cooper, Tata-McGraw-Hill Publishing Co. Ltd. , New Delhi (1976). University Chemistry (Third Edition), B. H. Mahan, Narosa Publishing House, New Delhi (1985). Coulson's Valence (Third Edition). R. McWeeny, ELBS and the Oxford University Press, Oxford (1979). Physical Chemistry (Fourth Edition), P. W. Atkins, ELBS and the Oxford University Press. Oxford (1991). Physical Chemistry (Fifth Edition), G. M. Barrow, Narosa Publishing House, New Delhi (1990). Fundamentals of Molecular Spectroscopy (Third Edition), C. N. Banwell. Tata McGraw-Hill Publishing Company. New Delhi (1985). The Nature of the Chemical Bond (Third Edition), L. Pauling, Oxford and IBH Publishing Company, Calcutta (1967) Chemistry of the Elements, N. N. Greenwood and A Earnshaw, Pergamon Press, New York (1989). Valence Theory, J. N. Murrell, S. F. A. Kettle and J. M. Tedder, ELBS and John Wiley & Sons, London (1977). Quantum Chemistry, D. A. McQuarrie, University Science Books and the Oxford University Press, Oxford (1983). Physical Chemistry. R. S. Berry, S. A. Rice and J. Ross, Academic Press, New York (1980). Intermolecular Forces: Their Origin and Determination, G. C. Maitland, M. Rigby, E. B. Smith and W. A. Wakeham, Clarendon Press, oxford (1981). Physical Chemistry (Third Edition), G. W. Castellan. Narosa Publishing House, New Delhi (1990). Principles of the Solid State, H. V. Keer. Wiley Eastern, New Delhi (1993). Computer Simulation of Liquids, M. P. Allen and J. Tildesley, Oxford University press (1987).