Alkanes Free- Radical Substitution

Chapter 3

Alkanes Free- Radical Substitution

Classification by structure: the family

• The basis of organic chemistry, we have said, is the structural theory. We separate all organic compounds into a number of families on the basis of structure. Having done this. We find that we have at the same time classified the compounds as to their physical and chemical properties. A particular set of properties is thus characteristic of a particular kind of structure.

• Within a family there are variations in properties. All members of the family may, for example, react with a particular reagent, but some may react more readily than others. Within a single compound there may be variations in properties, one part of a molecule being more reactive than another part. These variations in properties correspond to variations in structure.

• As we take up each family of organic compounds, we shall first see what structure and properties are characteristic of the family. Next, we shall see how structure and properties vary within the family. We shall not simply memorize these facts, but. whenever possible, shall try to understand properties in terms of structure, and to understand variations in properties in terms of variations in structure.

• Having studied methane in some detail, let us now look at the more complicated members of the alkane family. These hydrocarbons have been assigned to the same family as methane on the basis of their structure, and on the whole their properties follow the pattern laid down by methane. However, certain new points will arise simply because of the greater size and complexity of these compounds.

Structure of ethane

• Since each carbon atom is bonded to four other atoms, its bonding orbitals (sp* orbitals) are directed toward the corners of a tetrahedron. As in the case of methane, the carbon-hydrogen bonds result from overlap of these^ 3 orbitals with the s orbitals of the hydrogens. The carbon-carbon bond arises from overlap of twos orbital.

• In ethane, then, the bond angles and carbon-hydrogen bond lengths should be very much the same as in methane, that is, about 109.5 and about 1.10 A, respectively. Electron diffraction and spectroscopic studies have verified this structure in all respects, giving the following measurements for the molecule: bond angles, 109.5; C H length, 1.10 A; C C length, 1.53 A. Similar studies have shown that, with only slight variations, these values are quite characteristic of carbon-hydrogen and carbon-carbon bonds and of carbon bond angles in alkanes.

Free rotation about the carbon-carbon single bond. Conformations. Torsional strain

• This particular set of bond angles and bond lengths still does not limit us to a single arrangement of atoms for the ethane molecule, since the relationship between the hydrogens of one carbon and the hydrogen of the other carbon is not specified. We could have an arrangement like I in which the hydrogens exactly oppose each other, an arrangement like II in which the hydrogens are perfectly staggered, or an infinity of intermediate arrangements. Which of these is the actual structure of ethane. The answer is: all of them.

FREE ROTATION ABOUT THE CARBON-CARBON SINGLE 

• We have seen that a bond joining the carbon atoms is cylindrically symmetrical about a line joining the two carbon nuclei; overly strength should be the same for all these possible arrangements. If the various arrangements do not differ in energy, then the molecule is not restricted to any one-off them but can change freely from one to another. Since the change from one to another involves rotation about the carbon-carbon bond, we describe this freedom to change by saying that there is free rotation about the carbon-carbon single bond.

• The picture is not yet complete. Certain physical properties show that rotation is not quite free: there is an energy barrier of about 3 kcal/mole. The potential energy of the molecule is at a minimum for the staggered conformation, increases with rotation, and reaches a maximum at the eclipsed conformation. Most ethane molecules, naturally, exist in the most stable, staggered conformation; or, put differently, any molecule spends most of its time in the most stable conformation.

• The picture is not yet complete. Certain physical properties show that rotation is not quite free: there is an energy barrier of about 3 kcal/mole. The potential energy of the molecule is at a minimum for the staggered conformation, increases with rotation, and reaches a maximum at the eclipsed conformation. Most ethane molecules, naturally, exist in the most stable, staggered conformation; or, put differently, any molecule spends most of its time in the most stable conformation.

• How free are ethane molecules to rotate from one staggered arrangement to another. The 3-kcal barrier is not a very high one; even at room temperature the fraction of collisions with sufficient energy is large enough that a rapid interconversion between staggered arrangements occurs. For most practical purposes, we may still consider that the carbon-carbon single bond permits free rotation.

• The nature of the rotational barrier in ethane is not understood or what is not exactly the same thing -is not readily explained. It is too high to be due merely to van der Waals forces (Sec. 1.19): although thrown closer together in the eclipsed conformation than in the staggered conformation, the hydrogens on opposite carbons are not big enough for this to cause appreciable crowding. The barrier is considered to arise in some way from interaction among the electron clouds of the carbon-hydrogen bonds. 

• As the hydrogens of ethane are replaced by other atoms or groups of atoms, other factors affecting the relative stability of conformations appear van der Waals forces, dipole-Dipta interactions, hydrogen bonding. But the tendency for the bond orbitals on adjacent carbons to be staggered remains, and any rotation away from the staggered conformation is accompanied by torsional.

 Propane and the butanes

• The next member of the alkane family is propane, C3H8. Again, following the rule of one bond per hydrogen and four bonds per carbon, we arrive at structure I. Here, rotation can occur about two carbon-carbon bonds, and again is essentially free. Although the methyl group is considerably larger than hydrogen, the rotational barrier (3.3 kcal/mole) is only a little higher than for ethane. Evidently there is still not significant crowding in the eclipsed conformation, and the rotational barrier is due chiefly to the same factor as the barrier in ethane: torsional strain.

• one-carbon branch. There can be no doubt that these represent different structures, since no amount of moving, twisting, or rotating about carbon-carbon bonds will cause these ->nurtures to coincide. We can see that in the straight-chain structure each carbon possesses at least two hydrogens, whereas in the branched chain structure one carbon possesses only a single hydrogen; or we may notice that in the branched-chain structure one carbon is bonded to three other carbons, whereas in the straight-chain structure no carbon is bonded to more than two other carbons.

Conformations of /r-butane Van der Waals repulsion

• Let us look more closely at the /z-butane molecule and the conformations in which it exists. Focusing our attention on the middle C- C bond, we sec a molecule similar to ethane, but with a methyl group replacing one hydrogen on each carbon. As with ethane, staggered conformations have lower torsional energies and hence are more stable than eclipsed conformations. But, due to the presence of the methyl groups, two new points are encountered here: first, there are several different staggered conformations; and second, a factor besides torsional strain comes into play to affect conformational stabilities.

• The anti-conformation, it has been found, is more stable (by 0.8 kcal/mole) than the gauche. Both are free of torsional strain. But in a gauche conformation, the methyl groups are crowded together, that is, are thrown together closer than the sum of their van der Waals radii; under these conditions, van der Waals forces are repulsive (Sec. 1.19) and raise the energy of the conformation.

• Van der Waals strain can affect not only the relative stabilities of various staggered conformations, but also the heights of the barriers between them. The energy maximum reached when two methyl groups swing past each other rather than past hydrogensis the highest rotational barrier of all and has been estimated at 4.4-6.1 kcal/mole. Even so, it is low enough that at ordinary temperatures, at least the energy of molecular collisions causes rapid rotation; a given molecule exists now in a gauche conformation, and the next instant in the aim conformation.

Higher alkanes. The homologous ser

• If we examine the molecular formulas of the alkanes we have so far considered, we see that butane contains one caflki and two hydrogens more than propane, which in turn contains one carbon and two hydrogens more than ethane, and so on. A series of compounds in which each member differs from the next member by a constant amount is called a homologous series, and the members of the series are called homologs. The family of alkanes forms such a homologous series, the constant difference between successive members being CH2. We also notice that in each of these alkanes the number of hydrogen atoms equals two more than twice the number of carbon atoms, so that we may write as a general formula for members of this series, Cn H2n + As we shall see later, other homologous series have their own characteristic general formulas. 

• In agreement with this general formula, we find that the next alkane, pentane, has the formula C5 H 12, followed by hexane, C6 H14, heptane, C7 HJ6, and so on. We would expect that, as the number of atoms increases, so does the number of possible arrangements of those atoms. As we go up the series of alkanes, we find that this is true: the number of isomers of successive homologs increases at a surprising rate. There are 3 isomeric pentanes, 5 hexanes, 9 heptane, and 75 decancs (Ci); for the twenty-carbon eicosane, there are 366,319 possible isomeric structures! The carbon skeletons of the isomeric pentanes and hexanes are shown below.

• It is important to practice drawing the possible isomeric structures that correspond to a single molecular formula. In doing this, a set of molecular models is especially helpful since it will show that many structures which appear to be different when drawn on paper are actually identical.

Nomenclature

• We have seen that the names methane, ethane, propane, butane, and pentane are used for alkanes containing respectively one, two, three, four, and five carbon atoms. Table 3.2 gives the names of madjjplarger alkanes. Except for the four members of the family, the name is simply derived from the Greek (or Latin) prefix for the particular number of carbons in the alkane; thus, pentane for five, hexane for six, heptane for seven, octane for eight, and so on.

• The student should certainly memorize the names of at least the first ten alkanes. Having done this, he has at the same time essentially learned the names of the first ten alkenes, alkynes, alcohols, etc., since the names of many families of compounds are closely related. Compare, for example, the names propane, propene, and propane for the three-carbon alkane, alkene.

• As organic chemistry has developed, several different methods have been devised to name the members of nearly every class of organic compounds; each method was devised when the previously used system had been found inadequate for the growing number of increasingly complex organic compounds. Unfortunately for the student, perhaps, several systems have survived and are in current use. Even if we are content ourselves to use only one system, we still have to understand the names used by other chemists; hence it is necessary for us to learn more than one system of nomenclature. But before we can do this, we must first learn the names of certain organic.

Alkyl groups 

• In a similar way names are given to certain groups that constantly appear as structural units of organic molecules. We have seen that chloromethane, CH3C1, is also known as methyl chloride. The CH* group is called methyl wherever it appears, CH3 Br being methyl bromide, CH3 I, methyi iodide, and CH3OH, methyl alcohol. In an analogous way, the C2 H5 group is ethyl; C3 H7 , propyl; C4H9 , butyl; and so on.

• Among the alkyl groups we again encounter the problem of isomerism. There is only one methyl chloride or ethyl chloride, and correspondingly only one methyl group or ethyl group. We can see, however, that there are two propyl chlorides. I and II, and hence that there must be two propyl groups.  

• propyl bromides, iodides, alcohols, and so on in the same way. We find that there are four butyl groups, two derived from the straight-chain /f-butane, and two derived from the branched-chain isobutane. These are given the designations it- (normal), sec- (secondary), iso-, and tert- (tertiary), as shown below. Again the difference between w-butyl and sec-butyl and between isobutyl and /erf-butyl lies in the point of attachment of the alkyl group to the rest of the molecule.

• If the branching occurs at any other position, or if the point of attachment is at any other position, this name does not apply. Now that we have learned the names of certain alkyl groups, let us return to the original problem: the naming of alkanes.

 Common names of alkanes

As we have seen, the prefixes //-, iso-, and neo- are adequate to differentiate the various butanes and pentanes, but beyond this point an impracticable number of prefixes would be required. However, the prefix //- has been retained for any alkane, no matter how large, in which all carbons form a continuous chain with no branching:

IUPAC names of alkanes

• To devise a system of nomenclature that could be used for even the most complicated compounds, various committees and commissions representing the chemists of the world have met periodically since 1892. In its present modification, the system so devised is known as the IUPAC system (International Union of Pure and Applied Chemistry). Since this system follows much the same pattern for all families of organic compounds, we shall consider it in some detail as applied to the alkanes.

• from propane by the replacement of a hydrogen atom by a methyl group, and thus may be named methylpropane.

Classes of carbon atoms and hydrogen atoms

Each hydrogen atom is similarly classified, being given the same designation of primary, secondary, or tertiary as the carbon atom to which it is attached. We shall make constant use of these designations in our consideration of the relative reactivities of various parts of an alkane molecule.

Physical properties

• The physical properties of the alkanes follow the pattern laid down by methane and are consistent with the alkane structure. An alkane molecule is held together entirely by covalent bonds. These bonds either join two atoms of the same kind and hence are non-polar or join two atoms that differ very little in electronegativity and hence are only slightly polar. Furthermore, these bonds are directed in a very symmetrical way, so that the slight bond polarities tend to cancel out. As a result, an alkane molecule is either non-polar or very weakly polar.

• As we have seen ^Sec. 1.19), the forces holding non-polar molecules together (van der Waals forces) arc weak and of very short range; they act only between the portions of different molecules that are in close contact, that is, between the surfaces of molecules. Within a family, therefore, we would expect that the larger the molecule and hence the larger its surface area the stronger the intermolecular forces.

• Except for the very small alkanes, the boiling point rises 20 to 30 degrees for each carbon that is added to the chain; we shall find that this increment of 20-30 per carbon hofds not only for the alkanes but also for each of the homologous series that we shall study.

• The first four i-alkanes are gases, but, as a result of the rise in boiling point and melting point with increasing chain length, the next 13 (C$Cn) are liquids, and those- containing 18 carbons or more are 

• There are somewhat smaller differences among the boiling points of alkanes that have the same carbon number but different structures. On pages 77 and 80 the boiling points of the isomeric butanes, pentanes, and hexanes are given. We see that in every case a branched-chain isomer has a lower boiling point than a straight-chain isomer, and further, that the more numerous the branches, the lower the boiling point. Thus w-butane has a boiling point of and isobutane 12. w-Pentane has a boiling point of 36, isopentane with a single branch 28, and neopentane with two branches 9.5. This effect of branching on boiling point is observed within all families of organic compounds. That branching should lower the boiling point is reasonable: with branching the shape of the molecule tends to approach that of a sphere; and as this happens the surface area decreases, with the result that the intermolecular forces become weaker and are overcome at a lower temperature.

Industrial source

• The principal source of alkanes is petroleum, together with^th^accc^p^r^jng natural gas. Decay and millions of years of geologicaTstresses have transformed the complicated organic compounds that oncejnade up living plants or animals into a mixture of alkanes ranging in size from one carbon to 30 or 40 carbons. Formed along with the alkanes, and particularly abundant in California petroleum, are cycloalkanes (Chap. 9), known to the petroleum industry as naphthenes.

• Natural gas contains, of course, only the more volatile alkanes, that is, those of low molecular weight; consists chiefly of methane and progressively smaller amounts of ethane, propane, and higher alkanes. For example, a sample taken from a pipeline supplied by a large number of Pennsylvania wells contained methane, ethane, and propane in the ratio of 12:2: 1, with higher alkanes making up only 3of the total. The propane-butane fraction is separated from the more volatile components by liquefaction, compressed into cylinders, and sold as bottled gas in areas not served by a gas utility.

• In addition to being used directly as just described, certain petroleum fractions are converted into other kinds of chemical compounds. Catalytic isomerization changes straight-chain alkanes into branched-chain ones. The cracking process (Sec. 3.31) converts higher alkanes into smaller alkanes and alkenes, and thus increases the gasoline yield; it can even be used for the production of "natural" gas. In addition, the alkenes thus formed are perhaps the most important raw materials for the large-scale synthesis of aliphatic compounds. The process of catalytic reforming (Sec. 12.4) converts alkanes and cycloalkanes into aromatic hydrocarbons and thus helps provide the raw material for the large-scale synthesis of another broad class of compounds

Industrial source vs. laboratory preparation

• We shall generally divide the methods of obtaining a particular kind of organic compound into two categories: industrial source and laboratory preparation. We may contrast the two in the following way, although it must be realized that there are many exceptions to these generalizations. An industrial source must provide large amounts of the desired material at the lowest possible cost. A laboratory preparation may be required to produce only a few hundred grams or even a few grams; cost is usually of less importance than the time of the investigator.

• For many industrial purposes a mixture may be just as suitable as a pure compound; even when a single compound is required, it may be economically feasible to separate it from a mixture, particularly when the other components may also be marketed. In the laboratory a chemist nearly always wants a single pure compound. Separation of a single compound from a mixture of related substances is very time-consuming and frequently does not yield material of the required purity.

• In our study of organic chemistry, we shall concentrate our attention on versatile laboratory preparations rather than on limited industrial methods. In learning these we may, for the sake of simplicity, use as examples the preparation of compounds that may actually never be made by the method shown. We may discuss the synthesis of ethane by the hydrogenation of ethylene, even though we can buy all the ethane we need from the petroleum industry. However, if we know how to convert ethylene into ethane, then, when the need arises, we also know how to convert 2-methyl-l-hexene into 2-methylhexane, or cholesterol into cholestanol, or, for that matter, cottonseed oil into oleomargarine.

Preparation

• Preparation Each of the smaller alkanes, from methane through -pentane and isopentane, can be obtained in pure form by fractional distillation of petroleum and natural gas; neopentane does not occur naturally. Above the pentanes the number of ironers of each homolog becomes so large and the boiling point differences become so small that it is no longer feasible to isolate individual, pure compounds; these alkanes must be synthesized by one of the methods outlined below.

• In writing these generalized equations, however, we must not lose sight of one important point. An equation involving RC1, to take a specific example, has meaning only in terms of a reaction that we can carry out in the laboratory using a real compound, lii e methyl chloride or tert-butyl chloride. Although- typical of alkyl halides, a reaction may differ widely in rate or yield depending upon the particular alkyl group actually concerned. We may use quite different experimental conditions for methyl chloride than for ten- butyl chloride; in an extreme case, a reaction that goes well for methyl chloride might go so slowly or give so many side products as to be completely useless for tert-butyl chloride.

The Grignard reagent: an organometallic compound

• When a solution of an alkyl halide in dry ethyl ether, (C2H5)2O, is allowed to stand over turnings of metallic magnesium, a vigorous reaction takes place: the solution turns cloudy, begins to boil, and the magnesium metal gradually disappears. The resulting solution is known as a Grignard reagent, after Victor Grignard (of the University of Lyons) who received the Nobel prize in 1912 for its discovery. It is one of the most useful and versatile reagents known to the organic chemist.

• The Grignard reagent is the best-known member of a broad class of substances, called organometallic compounds, in which carbon is bonded to a metal: lithium, potassium, sodium, zinc, mercury, lead, thallium almost any metal known. Each kind of organometallic compound has, of course, its own set of properties, and 'its particular uses depend on these. But, whatever the metal, it is less electronegative than carbon, and the carbon-metal bond like the one i.

• The reaction with water to form an alkane is typical of the behavior of the Grignard reagent and many of the more reactive organometallic compounds toward acids. In view of the marked carbanion character of the alkyl group, we may consider the Grignard reagent to be the magnesium salt, RMgX, of the extremely weak acid, R H. 

Coupling of alkyl halides with organometallic compounds

• To make an alkane of higher carbon number than the starting material requires formation of carbon-carbon bonds, most directly by the coupling together of two alkyl groups. The most versatile method of doing this is through a synthesis developed during the late 1960s by E. J. Corey and Herbert House, working independently at Harvard University and Massachusetts Institute of Technology

• The choice of organometallic reagent is crucial. Grignard reagents or organolithium compounds, for example, couple with only a few unusually reactive organic halides. Organosodium compounds couple, but are so reactive that they couple, as they are being formed, with their parent alkyl halide; the reaction of sodium with alkyl halides (Wurtz reaction) is thus limited to the synthesis of symmetrical alkanes, R R.

• Organocopper compounds were long known to be particularly good at the formation of carbon-carbon bonds but are unstable. Here, they are generated in situ from the organolithium, and then combine with more of it to form these relatively stable organometallics. They exist as complex aggregates but are believed to correspond roughly to R2Cu~Li+. The anion here is an example of an ate complex, the negative counterpart of an onium complex (ammo/jfffm, oxonium).

Reactions

• The alkanes are sometimes referred to by the old-fashioned name of paraffins. This name (Latin: parumqffinis, not enough affinity) was given to describe what appeared to be the low reactivity of these hydrocarbons.

• But reactivity depends upon the choice of reagent. If alkanes are inert toward hydrochloric and sulfuric acids, they react readily with acids like HF-SbF5 and FSOjH-SbFs ("magic acid") to yield a variety of products. If alkanes are inert toward oxidizing agents like potassium permanganate or sodium dichromate, most of this chapter is devoted to their oxidation by halogens. Certain yeasts feed happily on alkanes to produce proteins certainly a chemical reaction. As Professor M. S. Kharasch used to put it, consider the "inertness" of a room containing natural gas, air, and a lighted match.

• In its attack, the reactive particle abstracts hydrogen from the alkane; the alkane itself is thus converted into a reactive particle which continues the reaction sequence, that is, carries on the chain. But an alkane molecule contains many hydrogen atoms and the particular product eventually obtained depends upon which of these hydrogen atoms is abstracted. Although an attacking particle may show a certain selectivity, it can abstract a hydrogen from any part of the molecule, and thus bring about the formation of many isomeric products.

Halogenation

• As we might expect, halogenation of the higher alkanes is essentially the same as the halogenation of methane. It can be complicated, however, by the formation of mixtures of isomers.

• Under the influence of ultraviolet light, or at 250-400, chlorine or bromine converts alkanes into chloroalkanes (alkyl chlorides) or bromoalkanes (alkyl bromides); an equivalent amount of hydrogen chloride or hydrogen bromide is formed at the same time. When diluted with an inert gas, and in an apparatus designed to carry away the heat produced, fluorine has recently been found to give analogous results. As with methane, iodi nation does not take place at all.

• Depending upon which hydrogen atom is replaced, any of a number of isomeric products can be formed from a single alkane. Ethane can yield only one halomethane; propane, /z-butane, and isobutane can yield two isomers each; -pentane can yield three isomers, and isopentane, four isomers. Experiment has shown that on halogenation an alkane yields a mixture of all possible isomeric products, indicating that all hydrogen atoms are susceptible to replacement. For example, for chlorination:

• Although both chlorination and bromination yield mixtures of isomers, the results given above show that the relative amounts of the various isomers differ markedly depending upon the halogen used. Chlorination gives mixtures in which no isomer greatly predominates; in bromination, by contrast, one isomer may predominate to such an extent as to be almost the only product, making up 97- 99% of the total mixture. In bromination, there is a high degree of selectivity as to which hydrogen atoms are to be replaced. (As we shall see in Sec. 3.28, this characteristic of bromination is due to the relatively low reactivity of bromine atoms, ancl is an example of a general relationship between reactivity and selectivity.)

Mechanism of halogenation

• This in turn depends upon the alkane and which hydrogen atom is abstracted from it. For example, w-propyl halide is obtained from a w-propyl radical, formed from propane by abstraction of a primary hydrogen; isopropyl halide is obtained from an isopropyl radical, formed by abstraction of a secondary hydrogen.

• How fast an alkyl halide is formed depends upon how fast the alkyl radical is formed. Here also, as was the case with methane, of the two chain propagating steps, step is more difficult than step, and hence controls the rate of overall reaction. Formation of the alkyl radical is difficult, but once formed the radical is readily converted into the alkyl.

Orientation of halogenation

• With this background let us turn to the problem of orientation; that is, let us examine the factors that determine where in a molecule reaction is most likely to occur. It is a problem that we shall encounter again and again, whenever we study a compound that offers more than one reactive site to attack by a reagent. It is an important problem, because orientation determines what product we obtain.

• As an example, let us take chlorination of propane. The relative amounts of /? -propyl chloride and isopropyl chloride obtained depend upon the relative rates at which w-propyl radicals and isopropyl radicals are formed. If, say, isopropyl radicals are formed faster, then isopropyl chloride will be formed faster, and will make up a larger fraction of the product. As we can see, w-propyl radicals are formed by abstraction of primary hydrogens, and isopropyl radicals by abstraction of secondary hydrogens.

• In spite of these differences in reactivity, chlorination rarely yields a great preponderance of any single isomer. In nearly every alkane, as in the example \vc have studied, the less reactive hydrogens are the more numerous; their lower reactivity is compensated for by a higher probability factor, with the result that appreciable amounts of every isomer are obtained.

Relative reactivities of alkanes toward halogenation

• The best way to measure the relative reactivities of different compounds toward the same reagent is by the method of competition, since this permits an exact quantitative comparison under identical reaction conditions. Equimolar amounts of two compounds to be compared are mixed together and allowed to react with a limited amount of a particular reagent. Since there is not enough reagent for both compounds, the two compete with each other. Analysis of the reaction products shows which compound has consumed more of the reagent and hence is more as reactive as each hydrogen of methane.

• Data obtained from similar studies of other compounds are consistent with this simple generalization: the reactivity of a hydrogen depends chiefly upon its class, and not upon the alkane to which it is attached. Each primary hydrogen of propane, for example, is about as easily abstracted as each primary hydrogen in H-butane or isobutane; each secondary hydrogen of propane, about as easily as each secondary hydrogen of w-butane or w-pentane; and so on. 

Ease of abstraction of hydrogen atoms. Energy of activation

• We have seen that the larger the E*ct of a reaction, the larger the increase in rate brought about by a given rise in temperature. We have just found that the differences in rate of abstraction among primary, secondary, and tertiary hydrogens are due to differences in aet . We predict, therefore, that a rise in temperature should speed up abstraction of primary hydrogens (with the largest act) most, and abstraction of tertiary hydrogens (with the smallest act) least; the three classes of hydrogen should then display more nearly the same reactivity.

• This leveling-out effect has indeed been observed: as the temperature is raised, the relative rates per hydrogen atom change from 5.0:3.8:1.0 toward. At very high temperatures virtually every collision has enough energy for abstraction of even primary hydrogens. It is generally true that as the temperature is raised a given reagent becomes less selective in the position of its attack; conversely, as the temperature is lowered it becomes more selectin.

• How can we account for the effect of structure on ease of abstraction of hydrogen atoms Since this is a matter of fact, we must look for our answer, as always, in the transition state. To do this, however, we must first shift our focus from the hydrogen atom being abstracted to the radical being formed.

Stability of free radicals

• We are not attempting to compare the absolute energy contents of, say, methyl and ethyl radicals; 'we are simply saying that the difference in energy between methane and methyl radicals is greater than the difference between ethane and ethyl radicals. When we compare stabilities offree radicals, it must be understood that our standard for each radical is the alkane from which it is formed. As we shall see, this is precisely the kind of stability that we are interested in.

• Relative to the alkane from which each is formed, then, the order of stability of free radicals is:

                                                          3 > 2 > 1 > 

Ease of formation of free radical

• Let us return to the halogenation of alates. Orientation and reactivity, we have seen (Sec. 3.23), are governed by the relative ease with which the different classes of hydrogen, atoms are abstracted. Jut by definition, the hydrogen being abstracted, and the radical being formed belong to the same class. Abstraction of a primary hydrogen yields a primary radical, abstraction of a secondary hydrogen yields a secondary radical, and so on. For example:

• This is an extremely useful generalization. Radical stability seems to govern orientation and reactivity in many reactions where radicals reformed. The addition of bromine atoms to alkenes, for example, is a quite different sort of reaction from the one we have just studied; yet, there too, orientation and reactivity are governed by radical stability. (Even in those cases where other factors steric hindrance, polar effects are significant or even dominant, it is convenient to use radical stability as a point of departure.)

Transition state for halogenation

• Is it reasonable that the more stable radical should be formed more easily? We have already seen that the differences in reactivity toward halogen atoms are due chiefly to differences in E^: the more stable the radical, then, the lower the act for its formation. This, in turn, means that the more stable the radical, the more stable the transition state leading to its formation both stabilities being measured, as they must be, against the same standard, the reactants. (Remember: fact is the difference in energy content between reactants and transition state.

• Examination of the transition state shows that this is exactly what we would expect as we saw before, the hydrogen-halogen bond is partly formed, and the carbon-hydrogen bond is partly broken. To the extent that the bond is broken. 

• The energy difference between isobutane and the tert-butyl\ radical, for example, to be smaller than between propane and the isopropyl radical. It is not unreasonable that this same factor should cause the energy difference between isobutane and the incipient tert-built radical in the transition state to be smaller than between propane and the incipient isopropyl radical in its transition.

Orientation and reactivity

• Throughout our study of organic chemistry, we shall approach the problems of orientation and reactivity in the following way.

• Both problems involve comparing the rates of closely related reactions: in the case of orientation, reactions at different sites in the same compound; in the case of reactivity, reactions with different compounds. For such closely related reactions, variations in rate are due mostly to differences in act; by definition, eat is the difference in energy content between reactants and transition state.

• We shall examine the most likely structure for the transition state, then, to see what structural features affect its stability without at the same time affecting by an equal amount the stability of the reactants; that is, we shall look for factors that tend to increase or decrease the energy difference between reactants and transition state. Having decided what structural features affect the act, we shall compare the transition states for the reactions whose rates we wish to compare: the more stable the transition state, the faster the reaction.

• In many, if not most, reactions where a free radical is formed, as in the present case, tile transition state differs from the reactants chiefly in being like the product. It is reasonable, when, that the factor most affecting the act should be the radical character of the transition state. Hence, we find that the more stable the radical.

Reactivity and selectivity

• In its attack on alkanes, the bromine atom is much more selective than the chlorine atom (with relative rate factors of 1600:82: 1 as compared with 5.0:3.8: 1). It is also much less reactive than the chlorine atom (only 1/250,000 as reactive toward methane, for example, as we sa\s in Sec. 2.19). This is just one example of a general relationship: in a set of similar reactions, the less reactive the reagent, the more selective it is in its attack.

• To account for this relationship, we must recall what we learned in Sec. 2.23. In the attack by the comparatively unreactive bromine atom, the transition state is reached late in the reaction process, after the alkyl group has developed considerable radical character. In the attack by the highly reactive chlorine atom, the transition state is reached early, when the alkyl group has gained very little radical character.

• Now, by "selectivity" we mean here the differences in rate at which the various classes of free radicals are formed; a more stable free radical is formed faster, we said, because the factor that stabilizes it delocalization of the odd electron (Sec. 6.28) also stabilizes the incipient radical in the transition state. If this is so, then the more fully developed the radical character in the transition state, the more effective delocalization will be in stabilizing the transition state. The isopropyl radical, for example, is 3 kcal more stable than the w-propyl radical; if the radicals were completely formed in the transition state, the difference in act would be 3 kcal. Actually, in bromination the difference in act is 3 kcal: equal, within the limits of experimental error, to the maximum potential stabilization, indicating, as we expected, a great deal of radical character. In chlorination, by contrast, the difference in act is only 0.5 kcal, indicating only very slight radical character.

Non-rearrangement of free radicals. Isotopic tracers

• Our interpretation of orientation (Sec. 3.21) was based on an assumption that we have not yet justified: that the relative amounts of isomeric halides we find in the product reflect the relative rates at which various free radicals were formed from the alkane. From isobutane, for example, we obtain twice as much isobutyl chloride as te/7-butyl chloride, and we assume from this that, by abstraction of hydrogen, isobutyl radicals are formed twice as fast as /erf-butyl radicals.

• H. C. Brown (of Purdue University) and Glen Russell (no\\ of luau State University) decided to test the possibility that free radicals, like carbonium ions, might rearrange, and chose the chlorination of isobutane as a good test case, because of the large difference in stability between w/-butyl and isobutyl radicals. If rearrangement of alkyl radicals can indeed lake place, it should certainly happen here.

• What the problem comes down to is this: does every abstraction of primary hydrogen lead to isobutyl chloride, and every abstraction of tertiary hydrogen lead to tert-butyl chloride? This, we might say, we could never know, because all hydrogen atoms are exactly alike. But are they? Actually, three isotopes of hydrogen exist: 'H, profit, ordinary hydrogen: 2 H or D, deuterium* heavy hydrogen; and 3 H or T, tritium. Protium and deuterium are distributed in nature in the ratio of 5000: 1. '(Tritium, the unstable, radioactive isotope, is present in traces, but can be made by neutron bombardment of 6 Li.) Modern methods of separation of isotopes have made very pure deuterium available, at moderate prices, in the form of deuterium oxide. D2O, heavy water. 

Combustion

• The reaction of alkanes with oxygen to form carbon dioxide, water, and most important of all heat* is the chief reaction occurring in the internal combustion engine; its tremendous practical importance is obvious.

• The mechanism of this reaction is extremely complicated and is not yet fully understood. 'I here seem to be no doubt, however, that it is a free-radical chain reaction. 'I he reaction is extremely exothermic and yet requires a very high temperature, that of a flame, for its initiation. As in the case of chlorination, a great deal of energy is required for the bond-breaking that generates the initial reactive - particles; once this energy barrier is surmounted, the subsequent chain-carrying steps proceed readily and with the evolution of  energy.

• A higher compression ratio has made the modern gasoline engine more efficient than earlier ones but has at the same time created a new problem. Under certain conditions the smooth explosion of the fuel-air mixture in the cylinder is replaced by knocking, which greatly reduces the power of the engine.

Pyrolysis: cracking

• Decomposition of a compound by the action of heat alone is known as pyrolysis. This word is taken from the Greek pry, fire, and lysis, a losing, and hence to chemists means "cleavage by heat"; compare hydrolysis, "cleavage by water.

• The pyrolysis of alkanes, particularly when petroleum is concerned, is known as cracking. In thermal cracking alkanes are simply passed through a chamber heated to a high temperature. Large alkanes are converted into smaller alkanes, alkenes, and some hydrogen. This process yields predominantly ethylene (C2 H4) together with other small molecules. In a modification called steam cracking, the hydrocarbon is diluted with steam, heated for a fraction of a second to 700- 900, and rapidly cooled. Steam cracking is of growing importance in the production of hydrocarbons as chemicals, including ethylene, propylene, butadiene, isoprene, and cyclopentadiene. Another source of smaller hydrocarbons is hydrocracking, carried out in the presence of hydrogen at high pressure and at much lower temperatures (250-450).

• e synthesis of aliphatic compounds. Most cracking, however, is directed toward the production of fuels, not chemicals, and for this catalytic cracking is the major process. Higher boiling petroleum fractions (typically, gas oil) are brought into contact with a finely divided silica-alumina catalyst at 450-550 and under slight pressure. Catalytic cracking not only increases the yield of gasoline by breaking large molecules into smaller ones, but also improves the quality of the gasoline: this process involves carbon Wm ions), and yields alkanes and alkenes with the highly branched structures desirable in gasoline.

Determination of structure

• One of the commonest and most important jobs in organic chemistry is to determine the structural formula of a compound just synthesized or isolated from a natural source.

• 
  

• The compound will fall into one of two groups, although at first we probably shall not know which group. It will be either, (a) a previously reported compound, which we must identify, or (b) a new compound, whose structure we must prove

• If the compound has previously been encountered by some other chemist who determined its structure, then a description of its properties will be found somewhere in the chemical literature, together with the evidence on which its structure was assigned. In that case, we need only to show that our compound is identical with the one previously described.

• First, we purify the compound and determine its physical properties: melting point, boiling point, density, refractive index, and solubility in various solvents. In the laboratory today, we would measure various spectra of the compound (Chap. 13), in particular the infrared spectrum and the nmr spectrum; indeed, because of the wealth of information to be gotten in this way, spectroscopic examination might well be the first order of business after purification. From the mass spectrum we would get a very accurate molecular weight.

• Now the question is: which alkane is it? Or which alkene, or which aldehyde, or which ester? To find the answer, we first go to the chemical literature and look up compounds of the particular family to which our unknown.

• If we find one described whose physical properties are identical with those of our unknown, then the chances are good that the two compounds are identical. For confirmation, we generally convert the unknown by a chemical reaction into a new compound called a derivative and show that this derivative is identical with the product derived in the same way from the previously reported compound.

Analysis of alkanes

• Upon qualitative elemental analysis, an alkane gives negative tests for all elements except carbon and hydrogen. A quantitative combustion, if one is carried out, shows the absence of oxygen; taken with a molecular weight determination, the combustion gives the molecular formula, -Cn H2n + 2 which is that of an alkane.

• An alkane is insoluble not only in water but also in dilute acid and base and in concentrated sulfuric acid. (As we shall see, most kinds of organic compounds dissolve in one or more of these solvents.)

• An alkane is unreactive toward most chemical reagents. Its infrared spectrum lacks the absorption bands characteristic of groups of atoms present in other families of organic compounds (like OH, C~O, ,C C, etc.)

• On the basis of its physical properties boiling point, melting point, density, refractive index, and, most reliable of all, its infrared and mass spectra it may be identified as a previously studied alkane of known structure.

• If it turns out to be a new alkane, the proof of structure can be a difficult job. Combustion and molecular weight determination give its molecular formula. Clues about the arrangement of atoms are given by its infrared and nmr spectra. (For compounds like alkanes, it may be necessary to lean heavily on x-ray diffraction and mass spectrometry.)