What is a covalent bond?

The theory of chemical bonding can be very complex and is worthy of several volumes in itself, but a basic knowledge will be sufficient for our purposes (see Box 1.2). The modern understanding of chemical bonding derives from quantum mechanics. A bond is formed when an atom shares its electrons with another atom. According to quantum mechanics, the electron behaves like a light wave and when electrons are shared they can either reinforce each other or cancel each other out in the way that light waves do when they destructively interfere. The waveform of an electron is known as a molecular orbital. The simplest atom is the hydrogen atom and when two hydrogen atoms join together they provide the simplest situation in which two molecular orbitals combine. Figure 1.2 shows the overlap between in-phase and out-of-phase hydrogen orbitals to form a bond.

Figure 1.2 Bond formation between hydrogen atoms.

The in-phase orbitals produce a viable σ−bond which holds the atoms together and the out-of-phase orbitals produce an σ* anti-bonding orbital in which there is no overlap between the atomic orbitals. Since the bonding between the two hydrogen atoms, or indeed any pair of atoms, is due to an electronic wave phenomenon one way of examining bonds is by observing their interaction with electromagnetic radiation.

A pair of hydrogen atoms requires high energy/short wavelength radiation at about 150 nm (8.3 eV) to break the bond between them. Formation of the anti-bonding orbital can be represented as shown in Figure 1.3 where the two electrons making the bond are moved apart. The use of curly arrows to represent movement of electrons is shown in more detail in Box 1.3. Because bonding energy is quantised, the energy required is an exact figure for a given bond. The anti-bonding state is unstable and short lived so either the bond will re-form or a new bond will form with some suitable chemical species in the environment.

Figure 1.3 The effect of UV radiation on the electrons in a σ bond.

Box 1.3 Curly arrows

Throughout the book, extensive use is made of curly arrows to represent the movement of electrons. This convention is quite simple to understand.

1. Any bond is composed of two electrons.

2. A double-headed arrow indicates two electrons moving and can indicate the formation or the breaking of a bond.

3. A single-headed arrow indicates the movement of a single electron.

Unpaired electrons (free radicals) usually have a very short time of independence since they will rapidly pair with another electron to form a bond. This is why free radicals are very damaging to living tissues.

Weaker bonds such as those found in DNA bases only require radiation around 260 nm (4.8 eV) to break them. This radiation is present in sunlight and hence the potentially damaging effects of overexposure to sunlight. A simple example of a weaker type of covalent bond is found in ethylene (Fig. 1.4). In this molecule the molecular orbitals have a more complex (hybrid) waveform than the simple 1S orbital of the hydrogen atom. The most important point to note is that the π bond in ethylene is formed by only partial overlap of the orbitals that form it. The stronger σ-bond in the molecule results from more complete overlap of the orbitals involved in it. Thus the energy required to excite the electrons in the π bond in ethylene to their excited (anti-bonding) state is lower than that required to excite those in the σ bond.

Figure 1.4 The carbon–carbon bonds in ethylene.

Table 1.2 shows the strength of some covalent bonds commonly found within drug molecules and within the biomolecules composing the human body, derived from calorimetric measurements.

Table 1.2 The energy of some covalent bonds

A method for estimating the strength of a bond is by using calorimetry, which measures the heat of a reaction. For example, the reaction between fluorine and hydrogen

shown above in theory generates 542 kJ mol⁻¹ of heat. Using the values shown in Table 1.2, on the deficit side, the hydrogen–hydrogen bond (431 kJ mol⁻¹ ) and the fluorine–fluorine bond (155 kJ mol⁻¹) are broken but on the credit side two H–F bonds (2 × 564 kJ mol⁻¹) are formed. Thus the total energy derived from the reaction is 1128−586 = 542 kJ mol⁻¹. The heat given out by a reaction is called the enthalpy of reaction and given the symbol Δ H, it is also given a minus sign, indicating that heat is produced by the reaction; thus the theoretical Δ H for the above reaction is −542 kJ mol⁻¹. The heat of such a reaction can be measured using a calorimeter, which measures the amount of heat released by measuring the increase in temperature of water surrounding the sample being reacted. In this case, if 2 moles (40 g) of hydrogen fluoride were formed from hydrogen and fluorine the heat generated would be enough to raise the temperature of 1 litre of water from 0°C to beyond its boiling point (1 mole = 18 g of water requires 75.4 J to raise its temperature by 1°C). To break the bonds between H and F in 2 moles of HF, 1128 kJ of heat energy would be required. Thermal dissociation of bonds within molecules occurs, for example, when the analytical technique of inductively coupled plasma emission spectroscopy is used. In this technique temperatures of ca. 6000°C, provided by a heated plasma of argon gas, are used to disrupt the bonds in molecules so that the elements making them up can be analysed by exciting their electrons.

Dipole moments of bonds

The strengths of covalent bonds (Box 1.4) are less important in appreciating the relationship between drug structure and activity than the weaker interactions which occur between molecules as opposed to the forces within molecules. An example of a weak bonding force is that generated by the dipole moment of a bond. Linus Pauling produced a method for estimating the electronegativity of elements using equation 1 shown below.

1

X is the electronegativity of elements A and B.

D_AB, D_A2 and D_B2 are the bond dissociation energies between AB, A₂ and B₂.

By arbitrarily defining the most electronegative element, fluorine, as having an electronegativity value of 4.0 the values for all the other elements can be defined and this led to the values shown in Table 1.1.

Many covalent bonds have dipoles and dipole moments within drug molecules are important for their interaction with the biomolecules within cells which transmit the drug's action. The larger the dipole, the stronger the dipolar interaction with proteins and receptors within the cell. In addition, the strength of the dipole moment gives an indication of how likely the drug is to be degraded by reactants with an affinity for positive or negative charge. As indicated above, in most cases drug action is not based on covalent bond formation. There are exceptions (see Box 1.5), and thus the importance of covalent bond strength in pharmaceutical chemistry relates mainly to the stability of drugs in formulations, although it usually relates to the lifetime of the drug in vivo as well. The likelihood of a drug degrading depends on a combination of bond strength and the ease with which it is attacked at either positive or negative centres of charge. Thus the carbon–carbon bond within biomolecules is one of the least susceptible to degradation, certainly in vitro and to some extent in vivo, since it does not have a large dipole despite the fact that it stores a large amount of energy.

In Table 1.2 above it can be seen that the carbon–carbon double bond is stronger than the carbon–carbon single bond but listing its additive strength in this way belies the fact that one of the carbon bonds, the π bond, within the double bond is relatively weak and thus is quite susceptible to chemical attack.

Some types of degradation reaction are more important with regard to drug molecules than others. Most often it is the polarised bonds within drug molecules which are most susceptible to chemical degradation, i.e. have the lowest activation energies in this respect. Figure 1.5 shows the partial charge distribution within aspirin and paracetamol.

Figure 1.5 The hydrolysis of (A) aspirin and (B) paracetamol.

Under conditions of acidic hydrolysis of an ester or amide the positively charged protons in the acidic medium will attack the most negative point within the group undergoing hydrolysis. In the case of aspirin the greatest negative charge is on the two oxygens within the ester group and protonation can occur at either of these positions, thus promoting attack by water at the positively charged carbon, which results in the breakdown of aspirin into acetic acid and salicylic acid. The activation energy for this reaction is low and hydrolysis occurs slowly at room temperature. It can also be seen from the polarised distribution of charge that the ester bond in aspirin is also susceptible to attack by HO⁻ which will attach itself to the positively charged carbon and thus promote hydrolysis. The hydrolysis of esters is covered in Chapter 2. As pharmaceutical chemists, we are interested in what conditions are likely to cause our medicine to degrade. Obviously, aspirin should be kept dry.

The amide bond in paracetamol is also polarised and the charges on oxygen and carbon are not very different from those on the oxygens in the ester group in aspirin. However, amides hydrolyse much less readily than esters. One way of explaining this is that the carbonyl carbon is not as positively charged in the amide as in the ester and thus is less liable to attack by water or by the hydroxyl ion. However, calculation of charge distribution by chemical modelling software has its limitations and sometimes a less sophisticated view yields useful information. Without changing the charge on the amide group it is possible to write a resonance form for the amide where the proton is on the oxygen. The concept of resonance can be applied throughout organic chemistry and in simple terms all it means is that electron density is shared more evenly throughout a structure or group within a structure. As was discussed at the beginning of this section, the most stable bonds are those where electron density is distributed evenly and thus there is likelihood of degradative attack. If the charge distribution is calculated for the resonance form of paracetamol the charge distribution across the oxygen, carbon and nitrogen is evened out. Thus, activation energy required for hydrolysis of an amide is greater than that required for hydrolysis of an ester. It is important to note that once the activation energy for hydrolysis has been overcome the acidic hydrolysis of an amide proceeds rapidly and irreversibly, i.e. it is thermodynamically favourable and releases more energy than hydrolysis of the equivalent ester. In this case a non-polar molecule activation energy can be high but this is followed by a high release of energy (Box 1.6).

Box 1.6 Activation energy versus thermodynamic energy release

The absence of a dipole moment means that the activation energy required for degradation of a carbon–carbon bond to occur is high; for example, propane gas is a rich source of thermal energy (3671 kJ mol⁻¹ is enough to boil several kettles of water when camping under the most arctic conditions). However, it has to be heated to a high temperature, e.g. with a match, gently warming it does not work, before it ignites and reacts with oxygen. Thermodynamically, this reaction lies far in the direction (to the right) of the formation of carbon dioxide from the carbon atoms and water from the hydrogen atoms in the propane. In contrast, the energy required to activate the degradation of an ester by acidic hydrolysis described below is minimal but the energy released by the reaction is small and thermodynamically the reaction lies only slightly in the direction of the reaction products.

Thus in chemical reactions we have two ideas: the ease with which the reaction can be triggered and the amount of energy released by the reaction.

Self Test 1.2

From the charge distribution in the following esters, arrange them in order of increasing rate of hydrolysis caused by HO⁻.

How much do you think the rate of acid hydrolysis might vary?

Under acidic conditions (pH 1) at room temperature, aspirin hydrolyses in a few hours whereas paracetamol can be stored at room temperature under acidic conditions (pH 1) for many days without much hydrolysis occurring. We will return to the properties of esters and amides in a little more detail in Chapters 3 and 4.

As can be seen in Figure 1.5 aspirin and paracetamol have many bonds in their structure but it is only the bonds where the charge distribution is polarised which are susceptible to hydrolytic attack. Ester bonds are among the most labile bonds which are found in drug molecules. Another type of bond which limits shelf life is found in the lactam ring of penicillins. This bond may look like an amide bond, and calculation of charge density on the C, N and O atoms in the bond does not reveal any great difference from the atoms in the amide bond of paracetamol. At this point an additional concept is required, which is that ring strain renders the lactam ring much more susceptible to nucleophilic attack (i.e. where a negative species attacks a positive) or electrophilic attack (where a positive species attacks a negative) and this is indeed also responsible for its biological activity (Ch. 22). Ring strain arises from the fact that carbon atoms have preferred angles between the bonds attached to them. The preferred angles for the four bonded carbons in the ring (A and B) is 109°28′ (Box 1.7) and for the three bonded carbonyl carbons 120°; in the lactam ring the carbons are forced to have angles approximating to 90°. The strained ring reduces the activation energy required for the hydrolysis reaction to proceed (Fig. 1.6).

Box 1.7 Ring strain

Four-bonded carbon prefers tetrahedral geometry. Any deviation from the preferred angle produces strain. Rings with six atoms or more are not strained. In five-membered rings the strain is fairly minimal but four- and three-membered rings are quite strained. Cyclopropane is the simplest three-membered carbon structure. In fact, the orbitals of the carbon atoms in this ring can be represented as being bent away from the ring with an angle of 9.4°, indicating that they do not overlap as extensively as in an unstrained structure.

Nevertheless, cyclopropane is quite stable compared to the lactam ring in a penicillin because its even charge distribution presents no point for nucleophilic or electrophilic attack and thus the activation energy for its reaction is high. Indeed, it was even used as a volatile anaesthetic, although it has fallen out of favour since a spark overcoming the activation energy for the molecular reaction results in a large thermodynamic change, i.e. an explosion.

Figure 1.6 Hydrolysis of the lactam ring.

The distortion of the ring geometry is further reflected in the IR absorption of the carbonyl carbon within the lactam ring which is at high energy, ca. 1770 cm⁻¹. This is due to the lack of interaction between the nitrogen and the carbonyl oxygen, which normally occurs in amides, resulting from the distorted geometry. IR provides a simple method for examining bond strength (Box 1.8).

  Page 9 

Box 1.8 Resonance and infrared spectrophotometry as a simple tool for observing bond strength

The concept of resonance was first applied to the benzene ring where all the bonds were found to be identical. In the structure drawn on the left it looks as if there are two different types of bond and the structure is more correctly drawn as on the right with the electron density from the double bonds evenly distributed throughout the molecule.

There are a number of ways of observing the electron density within double bonds. One simple method is infrared spectrophotometry (IR). IR observes interaction of IR radiation (heat radiation) with bonds in molecules. IR radiation causes bonds to stretch: the lower the wavenumber (1/λ) of the IR radiation required to stretch a bond the weaker a bond. The energy of the radiation used in IR is reported as 1/wavelength in centimetres. The higher the number, the higher the energy. Carbonyl groups absorb IR radiation between 1650 cm⁻¹ and 1800 cm⁻¹. The carbonyl group of the ester in aspirin has an absorption at 1760 cm⁻¹ whereas the carbonyl group in the amide of paracetamol absorbs at 1650 cm⁻¹. This indicates that the carbonyl in paracetamol has lost some of its electron density and thus some of its strength relative to the carbonyl in aspirin due to the resonance shown in Figure 1.5.

Introduction

The actions of drugs on biological molecules involve intermolecular forces. There are a number of types of interactions between molecules, ranging from strong ionic interactions to weak but additively strong van der Waals interactions. The different strength of the different types of intermolecular force can be observed in the melting points of the series of compounds shown in Table 1.3, all of which have similar molecular weights but which exhibit different types of intermolecular interactions. The high degree of precision with which a naturally occurring ligand binds to its site of action depends on a delicate balance of the different intermolecular forces (see also Box 1.9). Drugs usually do not fit a given receptor with the same degree of precision as a natural ligand but have actions which originate in strong binding to parts of the receptor. With a knowledge of receptor structure, the intermolecular forces required for binding can be used to model ideal drug structures which will bind strongly.

Table 1.3 The effect of different intermolecular forces on melting point and boiling point

Ionic bonds

When two atoms which are widely different in electronegativity are brought into contact, rather than a covalent bond forming, the more electronegative atom will take an electron or electrons from the less electronegative atom, thus becoming negatively charged while the less electronegative atom becomes positively charged. A simple example is the reaction between a sodium atom (a soft metal) and a chlorine atom (gas). Having gained positive and negative charges, the two atoms are now attracted to each other, and if they are considered to be point charges the magnitude of the energy of the attraction would be 694 kJ mol⁻¹ if they were separated by a distance of 200 picometres in a vacuum. Such an energy is greater than the strength of most covalent bonds. However, in practice many ionic bonds can be readily broken using water as a solvent because of the strength of the bonds formed between the water and the positive and negative ions.

When a salt is dissolved in a solvent, the force of attraction between its ions is governed largely by the dielectric constant of the medium in which it is dissolved and the distance of separation between the two ions. The force of attraction between two ions can be calculated according to equation 2.

2

Where Q1 and Q2 are the charges on the ions, r is the distance of separation and ε is the dielectric constant of the medium.

The dielectric constant of water is defined as 80 relative to that of a vacuum. Thus energy of the bond between two point charges separated by two picometres in water would be 80 times less than the energy of the bond in a vacuum and have a value of ca. 8 kJ mol⁻¹. Water is the best solvent for ionic compounds because of its high dielectric constant. Table 1.4 shows the dielectric constants of some common solvents. Sodium chloride dissolves in water at 1 g in 2.8 mL and in glycerol, which has half the dielectric constant of water; 1 g dissolves in 10 mL. Sodium chloride is almost insoluble in ethanol; in this case the dielectric constant is too low for the inter-ionic attraction to be overcome.

Table 1.4 Dielectric constants of some common solvents

In living systems, ionic associations take place in aqueous media and the energies involved in these types of interaction are between 4 and 40 kJ mol⁻¹. The reason for the wide range in energies of ionic interactions is that the strength of the interaction depends on the dielectric constant of the environment and on how closely the two ions approach. Strong interactions between ions may occur because within cell membranes and in the hydrophobic cavities of proteins the amounts of water present may be greatly reduced, resulting in a decrease in the dielectric constant of the environment. The binding of an ionic drug with hydrophobic groups attached may further reduce the water present in a binding site. Since the force of attraction between ions varies as the square of the distance of their separation ionic forces act over a greater distance than the other intermolecular forces described below. The long-range attraction of ions and the consequent speed of association between oppositely charged ions means that ionic forces are important in driving much of the mechanical work of the cell. Table 1.5 shows the principal ionic groups which are found in proteins, along with their pKa values and percentage ionisation of these groups at the physiological pH of 7.4. These charged groups within protein structures provide sites for the binding of ionic drugs through formation of ion pairs and non-ionic drugs through ion–dipole interactions.

Table 1.5 Ionised groups found on proteins

Acidic groups	pKa	% Ionisation at pH 7.4
Terminal carboxyl (–COO−)	1.8–2.4	100
Aspartic acid side chain (–CH2CH2COO−)	3.7	99.98
Glutamic acid side chain(–CH2CH2CH2 COO−)	4.3	99.9
Basic Groups
Terminal ammonium (–NH3+)	7.5–10.3	55.7–99.9
Arginine (–NH−(NH2)C=NH2+)	12.5	100
Lysine (–CH2CH2CH2CH2NH3+)	10.5	99.9
Histidine	6.0	4

Where ionic interaction is an important component in the interaction of a drug with a receptor or G-protein, other forces may come into play in reinforcing the ionic interaction and these include particularly dipole–dipole interactions, hydrogen bonding and van der Waals forces (as illustrated in Fig. 1.1 for adrenaline). Thus the thermodynamics of drug action is based on the interplay between several forces which are relatively weak compared to the covalent bonds which make up the scaffolding of living systems. That is not to say that these weaker forces do not result indirectly in the breaking or formation of covalent bonds, but this is a consequence of the initial interaction of the drug with its receptor, which may trigger off enzyme action which results the making or breaking of covalent bonds.

Dipole–dipole interactions

The strength of a dipole (Fig 1.7) within a bond depends on the relative electronegativities of the atoms composing the bond. Where a molecule contains a dipole, it can interact with other molecules containing a dipole. The force between two dipoles is give by equation 3.

Figure 1.7 Dipole–dipole interactions.

3

Where μ_A and μ_B are the dipole moments of the molecules involved in the interaction, ε is the dielectric constant on the medium in which the interaction is taking place. r is the distance between the two dipoles.

Equation 3 holds if the dipoles interact in a linear fashion as shown in Figure 1.7A: in this case the interaction is maximal. Where the interaction is at an angle as in Figure 1.7B, the interaction is given by equation 4 (cos θ is <1).

4

Dipolar interactions are very dependent on temperature since an increase in temperature promotes random motion in molecules, which makes the linear orientation of the dipoles less likely. The interaction between dipoles falls off with the fourth power of distance and thus acts over a much smaller distance than ionic interactions, which are dependent of r². The strength of these interactions is within an energy range of 0.4–4 kJ mol⁻¹. When brought into close contact within a receptor, for example, ionic interactions, dipolar interactions with a drug become significant since contact is both close and at the optimum angle. If dipole–dipole interactions are important for activity at a receptor, any slight spatial adjustment in order to accommodate of a molecule could result in loss of activity.

Ion–dipole interactions

The interaction between an ion and a dipole (Fig. 1.8) depends on the cube of the distance between them (equation 5). Like dipole–dipole interactions, these interactions depend on the orientation of the dipole with respect to the ion.

Figure 1.8 Ion–dipole interaction.

Charged dipole interactions tend to be stronger than dipole–dipole interactions and have energies in the range 1–10 kJ mol⁻¹. These interactions are responsible for the solvation of ions in solution and are responsible for promoting the dissolution of ionic solids.

5

Van der Waals forces

Van der Waals forces (or London dispersion forces) arise in molecules where the charge appears to be evenly distributed because vibrations within molecules cause transient dipoles. These transient dipoles induce transient attractive dipoles in adjacent molecules with energies in the range 0.3–1.9 kJ mol⁻¹. As can be seen from equation 6, these forces only operate at very close range and fall off at the rate of the seventh power of the separation of the interacting species.

6

Where I is the first ionisation potential of the atom or molecule involved in the interaction, α is its polarisabliity and r is the separation distance.

However, at close range the additive effect of the interactions can be high and, for instance, it has been estimated that the additive effect of attraction between the atoms involved in the hydrocarbon chain of stearic acid is 33.5 kJ mol⁻¹ (Fig. 1.9).

Figure 1.9 Van der Waals interaction between the chains of stearic acid.

The strength of the forces involved depends on the polarisablility of the structural component in which the transient dipole is induced. The increased polarisability is related to increased molecular weight and large atoms within a structure increase the van der Waals interactions considerably because of their high polarisabilities. This can be seen in the case of C₂H₅Cl (B.P. 12°C), C₂H₅Br (B.P. 38°C) and C₂H₅I (B.P. 72°C). The dipole moment of these molecules increases from iodide to chloride but this is more than offset by the increasing van der Waals interactions from chloride to iodide.