首页   按字顺浏览 期刊浏览 卷期浏览 The infra-red and Raman spectra of hydrocarbons. Part II. Paraffins
The infra-red and Raman spectra of hydrocarbons. Part II. Paraffins

 

作者: Norman Sheppard,  

 

期刊: Quarterly Reviews, Chemical Society  (RSC Available online 1953)
卷期: Volume 7, issue 1  

页码: 19-55

 

ISSN:0009-2681

 

年代: 1953

 

DOI:10.1039/QR9530700019

 

出版商: RSC

 

数据来源: RSC

 

摘要:

THE INFRA-RED AND RAMAN SPECTRA OF HYDROCARBONS. PART II. PARAFFINS By NORMAN SHEPPARD M.A. PH.D. and DELIA M. SIMPSON M.A. PH.D. (Mrs. J. N. AGAR) (DEPARTMENTS OF COLLOID SCIENCE AND PHYSICAL CHEMISTRY FREE SCHOOL LANE CAMBRIDGE) IN Part I of this Review,l dealing with the acetylenes and the olefins attention was focused on those features of the spectra which originate from relatively small and rigid parts of on the whole much larger mole- cules. This made the discussion of the spectra a comparatively simple matter and allowed the identification of most of the expected frequencies. To some extent a similar approach is helpful in the interpretation of the spectra of the paraffins in that many compact groups such as CH, CH, CHMe, and CMe, give rise to characteristic frequencies which can be recognised and assigned in the spectra of series of related molecules.How- ever it is not usually possible to obtain in this way complete interpretations of the spectra of the paraffins because the vibrations of the different parts of a molecule are so closely coupled as to defy analysis by such a method. This is true for instance in the case of the polymethylene chain of the n-paraffins. Further most of the data available in the infra-red and Kaman spectra of the paraffins refer to the liquid or the vapour phase. As will be shown in a later section under these conditions the substances are usually examined as mixtures of different rotational isomers. This produces other complications since each isomer of a given paraffin has its own set of characteristic frequencies. It should also be noted that the spectra of even the smallest paraffins where rotational isomerism is absent (such as propane) are relatively complex so complete assignments of their frequencies are not available to assist in the interpretation of the spectra of the more complex molecules.The study and correlation of the spectra of series of related paraffins is clearly of paramount importance and must form the first stage in attempt- ing the assignment of their frequencies. Most of the correlations in the infra-red spectra of paraffins discussed in this Review were noted in the course of a programme of research carried out during the last war on the infra-red spectra of hydrocarbons in aviation fuel a t the Universities of Oxford and Cambridge under the direction of Dr. H. W. Thompson and Prof.G. B. B. M. Sutherland. The results have been published in report form and have appeared in a more general review of group frequencies.3 They were also used in a more detailed discussion of the vibrations of some Sheppard and Simpson Quart. Reviews 1952 6 1. a Fellgett Harris Simpson Sutherland Thompson Whiffen a,nd Willis Institute of Petroleum Report XI 1946. Thompson J . 1948 328. 19 20 QUARTERLY REVIEWS branched paraffins by Simpson and S~therland.~ D. C. Smith 5 carried out' an extensive analysis of the data obtained in Washington and an article by Barnes and his co-workers lists various paraffin correlations in a general tabulation of group frequencies. Since the present Review was first written a paper by McMurry and Thornton 7 has appeared which includes a dis- cussion of the infra-red paraffin correlations between 1500 and 700 cm.-l.Their analysis is presented mainly in tabular form and is particularly valu- able in that intensities are also listed. The approach is empirical with little discussion of the assignment of the various frequencies and none of the analogous Raman data. Many of the correlations in the Raman spectra of paraffins have already appeared in articles by Stepanov 8 and by Sheppard.1° It should also be ernphasised that the earlier work of Kohlrausch l1 and Mecke l2 is of considerable importance. The experimental data on the infra-red spectra used in this Review have as far as possible been taken from curves contributed by the Naval Research Laboratory to the Catalogue of Spectra published by the American Petroleum Institute; l3 in most cases no observations are available below 450 cm.-I.The Raman data employed are those published in two papers by Fenske and his collaborators.14~ l5 Together these cover more than a hundred substances including observations available on all paraffins up to the dodecanes that have hitherto been investigated. The Raman data on the paraffins up to 1939 have been reviewed by Hibben,lG and most of the references up to 1943 as well as tables of fre- quencies will be found in the most recent book by Kohlrausch; l7 further references may be obtained from an article by Glockler.ls The following papers on the Raman spectra of series of paraffins may be noted n- paraffins,19 octanes,20 hexanes and heptanes ; 21 the work of Bazhulin and his collaborators is also extensive.22 The pioneer studies by Lambert and (a) J .Chem. Phys. 1947 15 153 ; ( b ) Proc. Roy. SOC. 1949 A 199 169. 5 D. C. Smith Report on the infra-red spectra of hydrocarbons C 3274. 6 Barnes Gore Stafford and Williams Analyt. Chem. 1948 20 402. 7 Ibid. 1952 24 318. J . Phys. Chenz. U.R.S.S. 1946 20 017. 9 Acta Physicochim. U.R.S.S. 1947 22 238. 10 J . Chem. Phys. 1948 16 690. 11 Kohlrausch " Ramanspektren " Hand- und Jahrbuch der Chemischen Physik 1s American Petroleum Institute Research Project 44. Catalogue of infra-red Carnegie Institute of Technology Pittsburgh Pa. U.S.A. 1 4 Fenske Braun Wiegand Quiggle McCormick and Rank Analyt. Chem. 1947 16 " The Raman Effect and its Chemical Applications " Reinhold New York 1939. 17 Op. cit. 1943 p. 209. 19 Herz Kahovec and Wagner Monatsh. 1946 76 100.20 von Grosse Rosenbaum and Jacobson Analyt. Chem. 1940 12 191. 21 Cleveland and Porcslli J . Chem. Phys. 1950 18 1469. 22 ( a ) Bazhulin Plate Solovova and Kazanskii Bull. Acad. Sci. U.R.S.S. (SBr. Chim.) 1941 13 ; (b) Bazhulin Bokshtein Liberman Lukina Margolis Solovova and Kazanskii ibid. 1943 198 ; (c) Bazhulin Ukholin Bulanova Kopernina Plate and Kazanskii ibid. 1949 481. Naval Research Laboratory Washington D.C. U.S.A. 1948. Reeker und Erler Leipzig 1943. a,bsorption spectral data. 19 700. la 2. phys. Chem. 1937 B 36 347. l5 Braun Spooner and Fenske ibid. 1950 22 1074. l8 Rev. Mod. Phys. 1943 15 111. SHEPPAR-D AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 21 Lecomte on the infra-red spectra of paraffins should be mentioned,23 24 as well as the wartime reports published under the auspices of the British Institute of Petroleum 2 ,5 26 and the investigation of the octanes in the liquid and the vapour state by Oetjen and Randall.27 Other spectra are to be found in the collection of data published by the American Petroleum Institute,13 and further references will be given in their appropriate contexts.1. Introduction In considering the spectra of the paraffins it is convenient to divide the vibrations into two main types those involving mainly the hydrogen atoms and those concerned with the nuclear framework although in a few cases such a distinction is too drastic. it is clear that the skeletal frequencies will differ considerably from one molecule to another whereas some at least of the hydrogen vibrations may be ex- pected to remain nearly invariant.It is necessary in the first place to describe briefly the various types of hydrogen vibrations associated with the CH, CH, and CH groups. These fall into two categories (i) internal vibrations are those that would appear even if the group in question were isolated (for example the carbon-hydrogen stretching modes) ; (ii) ex- ternal vibrations (such as many of the carbon-hydrogen deformation modes) are only present if there is some mechanical interaction between the group itself and the rest of the molecule. As might be expected the former give rise to very constant frequencies whereas those corresponding to the external vibrations are much more dependent on the structure of the molecule as a whole. These externa'l vibrations may couple extensively with other carbon-hydrogen deformation modes and with the skeletal vibrations.(a) Carbon-Hy drogen Stretching Vibrations .-T he carbon - h y d r og en stretching vibrations usually appear in the range 3000-2700 cm.-l and a,re generally intense in both Raman and infra-red spectra. It is possible as shown by Fox and 29 to use the infra-red frequencies to dis- tinguish between the different modes of the CH, CH, and CH groups ; all are internal vibrations. CH Symmetrical stretching vibration 2872 cm.-l 2962 cm.-] From the earlier discussion These assignments are listed below Asymmetrical doubly degenerate stretching vibration CH Symmetrical stretching vibration Asymmetrical stretching vibration 2853 cm.-l 2926 cm.-l CH Stretching vibration (semi-schematic) 2890 cm.-l 2 3 (a) Andant Lambert and Lecomte Compt.rend. 1934 198 1316 ; ( b ) Lambert 2 4 Lambert and Lecomte Ann. Physique 1938 10 503. z 5 Sutherland and Thompson Petroleum Board Enemy Oils and Fuels Committee 26 Idem Institute of Petroleum Report X 1945. 37 Rev. Mod. Phye. 1944 16 165 and (with Anderson) 260. 28 PTOC. Roy. SOC. 1938 A 167 257. *9 Ibid. 1940 A 175 208. and Lecomte ibid. 1938 206 1174. Report 1943. 22 QUARTERLY REVIEWS The above are mean values and are to some extent dependent on the number and environment of each type of group present (see also refs. 30-32). Other weaker frequencies are found near 2934 cm.-l (infia-red) 2912 cm.-l (Raman and infra-red) and in the range 2800-2700 cm.-l (Raman). These are interpreted as overtone or combination frequencies of the internal deformation modes of the CH and CH groups and appear with enhanced intensity through interaction with the neighbouring carbon-hydrogen stretching fundamentals of the same symmetry classes.( b ) Carbon-Hydrogen Deformation Vibrations.-The deformation modes involving changes in the angles associated with carbon-hydrogen linkages occur in the range 1500-600 cm.-l. (1) The CH Group.-An isolated CH group attached to a heavy frame- work has three internal deformation vibrations (one symmetrical and two approximately doubly degenerate asymmetrical modes) which are observable as nearly constant frequencies. The two doubly degenerate vibrations appear in the range 1470-1440 cm.-l and are intense in both Raman and infra-red spectra. The symmetrical deformation mode occurs as an intense infia-red band near 1380 crn.-l but the corresponding Raman line is usually missing.As has already been noted,l the frequency of this mode shifts to somewhat higher values if the CH group is directly attached to a double or triple bond and its intensity in the Raman spectrum is then enhanced. In the paraffins splitting of some of the infra-red bands occurs when a plurality of CH groups is attached to a single carbon atom. The regu- larities which have been noted are summarised below and are quoted as average values (in cm.-l) 5 9 7 / \ CHa-C- CH H isolated 1380 isopropyl 3 3-dimethyl tert.-butyl 1385 1370 1384 1367 1397 1370 The splitting patterns in more complex paraffins with several of these groups have not yet been fully worked out. Each CH group has also 30 Sushinskii; Bull. Acad. Sci. U.R.S.S. (SBr.Phys.) 1947 11 341. 31 Saier and Coggeshall Analyt. Chem. 1948 20 812. 3a Hastings Watson Williams and Anderson ibicl. 1052 24 612. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART II 23 three external vibrations viz. two rocking (or wagging) modes and a tor- sional motion of the group about the CH,-C linkage. The frequencies of the CH rocking modes vary considerably in different classes of paraffins and may be coupled strongly with the skeletal vibrations. They appear to be confined to the range 1250-800 ~m.-1,~3 and may give rise to both prominent Raman lines and intense infra-red bands. The CH torsional mode corresponds in most cases to some sort of restricted rotation and so may be expected in the low-frequency region < 300 cm.-I. This mode seems to be weak in both Raman and infra-red spectra and usually cannot be located directly (2) The CH Qroup.-An isolated CH group attached to a heavy frame- work has four deformation modes as shown in Fig.1.l0$ 34 The bending vibration is internal in type and occurs in the narrow range 1470-1440 cm.-l (Le. it coincides with the asymmetrical CH deformation vibration) ; it is strong in both Raman and infra-red spectra. Bending Wagging Twisting Rocking FIG. 1 Approximate diagrams of the deformation modes of an bolated CH group. The three remaining CH modes are all external and so are more variable in frequency. Investigations of the spectra of the long-chain paraffins 36 have suggested that the CH wagging and twisting frequencies are to be found in the range 1350-1150 cm.-l ; this range is also indicated by cal- culations for a variety of paraffin^.^ 34* 36-41 Some of these modes give rise to prominent Raman lines near 1300 cm.-l; in general the infra-red absorption bands are only of medium or weak intensity.There has been some discussion about the exact interpretation of the various series,l0* 83-35 but no general agreement has been achieved. Further investigations of the spectra of propane and its deuterated derivatives by McMurry and Thornton 4 2 9 43 (see later) have shown that for these molecules the CH wagging vibration is coupled with other modes of the same symmetry class. It might therefore be thought unrealistic to attempt to assign frequencies to the various external modes of CH groups. However in the opinion 3 3 Torkingt,on J. Chem. Phys. 1950 18 768. 34 Rasmussen ibid. 1948 16 712.35 Brown Sheppard and Simpson Discz~ss. Faraday SOC. 1950 9 261. 3* Eliashevitch and Stepanov Conapt. rend. Acad. Sci. U.R.S.X. 1941 8S 481. 37 Stepanov Bull. Acad. Sci. U.R.X.S. (SBr. Phys.) 1947 11 357. 38 Kellner Nature 1949 163 877. 3* Idem Proc. Phys. Soc. 1951 A 64 521. 40 Simanouti b. Chem. Pkys. 1949 17 734. 4a Ibid. 1960 18 1515. 4 3 Ibid. 1951 19 1014. 4l Barrow ibid. 1951 19 346. 24 QUARTERLY REVIEWS of the Reviewers such a procedure is still useful as a first approximation since it is not impossible that the effects of interaction are less drastic in larger molecules. Such a classification has in fact been used by McMurry and Thornton in their recent analysis.; The CH rocking modes occur in the range 1100-700 cm.-l as sug- gested by theoretical treatments 37 and investigations of the long-chain paraffins.35 These vibrations cannot be identified in the Raman spectra but give rise to intense infra-red absorption bands particularly in the region 770-720 ~ m .- l . ~ ~ 45 (3) The CH Group.-The two hydrogen deformation modes of a lone CH group appear in symmetrical surroundings as a doubly degenerate pair as in isobutane 46 near 1335 cm.-l. In other circumstances splitting may occur but one component can usually be recognised as a Raman line be- tween 1360 and 1330 cm.-l; lo owing to the proximity of the intense absorption due to the CH deformation vibration the corresponding infra- red frequencies cannot usually be identified (c) Skeletal Vibrations.-The range of skeletal stretching vibrations in branched paraffins is probably 1300-650 ~ m .- l ~ ? 47 though it is more restricted for the n-paraffins. It should however be noted that some authors 3 4 7 48 still favour an upper limit near 1100 cm.-l as suggested originally by Kohlrausch and Koppl. 4g The skeletal deformation modes and the torsional vibrations of the nuclear framework occur together below 600 cm.-l. From the above summary it may be concluded that the carbon-hydrogen stretching vibrations and the internal deformation modes of the CH and CH groups are relatively insensitive to changes in structure. In the follow- ing discussion attention will be concentrated on the remaining types of vibration which occur a t frequencies below 1350 cm.-l the torsional modes about which too little is known being omitted. 2. Rotational Isomerism in the Paraflfins As already noted the interpretation of the spectra of many paraffins is complicated by the fact that such substances exist as mixtures of two or more rotational isomers in the liquid or vapour state.Thus to consider a very simple example n-butane (which may be regarded as symmetrical dimethylethane) can occur in two distinct forms shown in Fig. 2 The trans-form corresponds to the planar zig-zag configuration ; the gauche forms (spectroscopically indistinguishable) are derived from this by rotation through approximately 120" of one half of the molecule with respect to the other about the central carbon-carbon linkage. Although restricted rotation of the CH group can also occur about the terminal C-C linkages 4 4 Sheppard and Sutherland Nature 1947 159 739. 4 5 Vallance Jones and Sutherland ibid.160 567. .t6 Pitzer and Kilpatrick Chem. Reviews 1946 39 435. 47 Ahonen J . Chem. Phys. 1946 14 625. 48 Rank Saksena and Shull Discuss. Fcrrnday Soc.. 1950 9 187. 49Z. phys. Chem. 1934 B 26 209. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART II 26 the high symmetry of this group causes the resulting configurations to be indistinguishable. The existence of such mixtures of rotational isomers in paraffins was first demonstrated spectroscopically by Kohlrausch and K o ~ p 1 ~ ~ who investigated Raman spectra in the liquid state and showed that more lines than predicted were present Since then considerable evidence both spectroscopic and thermodynamic has accumulated to confirm these observations. Of major importance is the fact that the spectroscopic observations indicate that each rotational isomer of a given molecule has its own set of vibrational frequencies.Me H H Me Me trans ‘Gauche’ FIO. 2 The conjguralions of the rotational isomers of n-butane. The number of possible rotational isomers of a particular paraffin may be readily deduced. Systematic discussions have been given by Pifzer,s0 and by Simanouti and Mizushima 51 who have invented a convenient notation for the description of the different forms. Since all the rotational isomers of a given molecule may theoretically be present in the liquid substance it is obviously of considerable import- ance to know their relative stability since this will determine the propor- tions of the different forms in the mixture. A direct method of tackling this problem is based on spectroscopic measurements.The spectra (Raman or infra-red) of the crystalline solid and the liquid at different temperatures are compared. The solid material shows the frequencies of one form (often but not always the most stable isomer) ; the extra frequencies in the liquid arise from isomers of different energy content. In favourable circumstances the liquid may contain only one species other than the most stable isomer ; by comparison of the variation with temperature of the intensities of cor- responding frequencies of the two forms the energy difference involved may be determined directly. This method in the hands of Rank and his col- laborators has proved very successful for some of the 12-paraffins 52-55 (see also ref. 56). Alternatively it is possible to calculate energy differ- ences from thermodynamic data.50 At present data on the solid-state spectra are available for a rather 50 J.Chem. Phys. 1940 8 711. 5 1 Sci. Papers Inst. Phys. Chem. Res. Tokyo 1943 40 467. 5 2 Szasz Sheppard and Rank J. Chenz. Phys. 1948 16 704. 53 Rank Sheppard and Szasz ibitl. 1949 17 83. 5 1 Sheppard and Szasz ibid. p. 86. “ ( a ) Axford and Rank ibid. p. 430; (b) ibid. 1950 18 51. 56Mizushima and Okazaki J . Amer. Chem. SOC. 1949 71 3411. 26 QUARTERLY REVIEWS limited range of molecules. The isomer present in the crystalline n-paraffins has the planar zig-zag configuration? as is also known from the X-ray investi- gation of the solids ; 57 the stability of the other isomers which are formed by one or more rotations about the different carbon-carbon bonds away from the planar configuration decreases with the number of these rotations.50 Thus the gauche form of n-butane has a greater energy content than the trans-isomer.A few branched paraffins have also been studied but the interpretation of the data is still under discu~sion.55~~ 58-61 Since each rotational isomer of a particular molecule has its own distinct set of vibrational frequencies? it might be thought that in order to discuss the characteristic group frequencies of a series of related paraffins? it would be essential first of all to distinguish between the frequencies arising from each type of isomer and then to compare those of each type in turn. Such a procedure might be feasible in principle but would in practice be extremely laborious and time-consuming. Fortunately? it is found empirically that Characteristic group frequencies may be recognised in various classes of paraffin even when these substances are investigated in the liquid state.There are a number of reasons for this; it might so happen that one or more vibrations of the different isomers of a given type of molecule all appear near the same frequency as for instance for some of the CH wagging and CH rocking modes of the n-paraffins. Alternatively? if a symmetrical unit such as the tert.-butyl group forms the common feature of a series of paraffins the frequencies of the vibrations characteristic of this group may remain relatively unaffected by what is happening in the rest of the molecule so that they too are nearly the same for different isomers. Even in less favourable circumstances it is still possible to pick out frequencies charac- teristic of particular structures.Purther details of the examples already discussed and other cases of rotational isomerism will be found in the appropriate sections. 3. The n-Paraffi In the discussion of the frequencies characteristic of paraffins the spectra of the n-paraffins form an appropriate starting point for a number of reasons. In the first place most paraffins possess within their structures a carbon- carbon chain of varying length so it is necessary to know the characteristic frequencies of such an arrangement? in order that the frequencies of other groupings may be identified. Further the data on the spectra of the n-paraffins have provided most of the information on the CH and CH hydrogen vibrations discussed above and it is desirable at an early stage to show how this has been derived.Moreover as mentioned previously only in the case of the n-paraffins are extensive data available for a series 57 Muller Proc. Roy. Sroc. 1928. -4 120 437. 55 Mizushima Morino and Taketla. Sci. Papers Inst. Phys. Chon. Res. Tokyo 1941 59 Avery and Ellis J . Chem. Phys. 1942 10 10. 60 Szasz and Sheppard ibid. 1949 17 93. 6 1 Brown and Sheppard ibid. 1951 19 976. 38 437. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART 11 27 of rotational isomers of a particular type (in this case with the trans-planar zig-zag configuration) and the data on the liquid substances are also much more complete than those for any other class of paraffin. Comparison of the spectra of the trans-isomers with those of mixtures of the isomers as found in the liquids is very instructive (Fig.3). Finally much attention Infra-red spectra between 1500 and 700 cm.-l of n-octane in the liquid (A) and the crystalline solid (B) phase.87 ( Vertical lines indicate characteristic absorption frequencies listed below.) has been given to the theoretical treatment of the skeletal vibrations of chains of repeating units of which the trans-isomers of the n-paraffins form examples. Comparison of the theoretical expectations with the experi- mental results for these molecules gives some indication as to how far such treatments are likely to be helpful guides in the interpretation of spectra. (a) Methane and Ethane.-The smallest paraffin methane is in a class by itself since it has only one carbon atom and therefore no skeletal vibra- tions. The frequencies of the four hydrogen modes are well-known ; those listed by Herzberg 62 have been confirmed by more recent work.63 6 4 Extensive investigations and assignments have also been made for the various deuterated derivatives.62 The structure of ethane has been the subject of some discussion; free rotation about the carbon-carbon linkage is definitely absent and on the whole the staggered configuration with a centre of symmetry is ~referred.~5> The frequencies as given by Herzberg 6 5 have been slightly revised by the recent work of L.G. Smith.66 The CH stretching modes and the internal CH deformation vibrations give frequencies corresponding to those listed on pp. 21 22. The single skeletal mode appears as a polarised Raman line 6 2 " Infra-red and Ra,man Spectra of Polyatoinic Molecules " Van Nost,rand New Pork 1945 p.306. 63 Holden Taylor arid Johnston J. Chen~. Phys. 1949 17 1356. 64 Rank Shull and Axford ibid. 1950 18 116. 6 6 Op. cit. 1945 p. 342. 6 6 J . Chem. Phys. 1049 17 130. 28 QUARTERLY REVIEWS a t 993 cm.-l. The two doubly degenerate CH rocking vibrations appear a t 1190 cm.-l (infra-red active) and 822 cm.-l (Raman active depolarised). The frequency of the torsional mode has been located at 290 cm.-l; this is in good agreement with the thermodynamic data. The molecule C2D6 has also been investigated ; references and the generally accepted assign- ments of the frequencies may be found in Herzberg’s book.65 ( b ) Propane.-Although propane and several of its deuterated derivatives have been extensively studied,17 2 6 9 4 2 9 439 67-70 the assignment of its frequen- cies below 1350 cm.-l has proved unexpectedly trouble~ome.~~~ 3 6 9 429 439 67-72 The main reason for this difficulty is that although it is formally possible to draw up a scheme classifying the vibrations in this region as CH torsion CH rocking CH deformation and skeletal modes yet in this molecule (because of the small number of atoms) vibrations of the same symmetry type become more or less mixed so that such clear-cut distinctions become unrealistic.2 9 43 For the present purpose it is not necessary to discuss in detail the eight CH stretching modes and the six internal deformation vibrations of the CH groups. These occur in the expepted frequency regions as set out on pp. 21 22. With regard to the remaining vibrations the following seems to be the most plausible assignment on the data available (Y in cm.-l).It should be emphasised however that more experimental work is necessary before it can be considered established. In particular the polarisations of the Raman lines of propane itself are required as well as further investigations on the deuterated derivatives. B C-C stretching (asymmetrical) 1053 A C-C stretching (symmetrical) 870 A CH rocking (in-plane symmetrical) 1155 B CH rocking (in-plane asymmetrical) 923 B A CH rocking (out-of-plane asymmetrical) 940 ? CH rocking (out-of-plane symmetrical) A CH bending 1460 B CH wagging 1336 A CH twisting 1278 B CH rocking 748 A C-C-C deformation 372 The above assignment is essentially that of P i t ~ e r ~ ~ as modified by Rasrnu~sen,~~ and is based on the earlier infra-red and Raman studies of the molecule.l79 67 The two CH torsional modes have been put at 283 and 202 cm.-l by Pitzer,Y2 using indirect methods.More recent investi- gations have not modified this interpretation though some difficulties remain. The infra-red band corresponding to the A CH rocking mode b7 Wu and Barker J . Che?iz. Phys. 1941 9 487. 68 A.P.I. Infra-red Spectrograms Nos. 60 99 629 791 792 793 1141. 69 McMurry Thornton a,nd Condon J . Chem. Phys. 1949 17 918. 70 Friedman and Turkevitch ibid. p. 1012. 7 1 Wagner 2. phys. Chem. 1939-1940 B 45 69. 78 Pitzer J . Chem. Phys. 1944 12 310. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART 11 29 has a somewhat anomalous contour under low resolution the reality of the weak Raman line corresponding to the A CH rocking mode needs confirmation and the position of the B CH rocking mode has yet to be established .Infra-red data are available for a number of deuterated derivatives of propane. The results for the unsymmetrical molecules are not easy to interpret.693 70 From the investigations of the molecules CH,*CD,CH3,42 CD3*CH,*CD, and C3DB,43 it has been established that the external carbon- hydrogen deformation vibrations and the skeletal modes of the same sym- metry classes interact significantly so that in propane the A frequencies at 1155 and 870 em.- and the B frequencies a t 1336 1053 and 923 crn.-l are respectively coupled together. Although the Reviewers agree with the conclusion that it is not strictly correct to attribute a given frequency solely to one particular type of group vibration yet they feel that such a classification is valuable as a first approximation even for propane.It should also be noted that for the undeuterated molecule the separation of some of the interacting frequencies of the same symmetry class is quite large. Further it appears not impossible that owing to the large shifts in the carbon-hydrogen deformation frequencies that occur on deuteration the degree of coupling of vibrations of the same symmetry type may alter significantly and is possibly more important the more deuterated the molecule.43 (c) The trans-Isomers of the n-ParafEns.-As has already been noted. in the crystalline solids the n-paraffins CnHZn+2 assume a planar zigzag truns- configuration. This implies that molecules with an even number of carbon atoms have a centre of symmetry so that vibrations are allowed in either Raman or infra-red spectra ; on the other hand many of the vibra- tions in molecules with an odd number of carbon atoms may appear in both types of spectra.These differing selection rules have proved most valuable in assigning the various series of frequencies that are observed. Further guidance for this purpose is provided by the theoretical dis- cussions (of varying degrees of complexity) of the skeletal vibrations of a regular chain of repeating units such as that represented by the truns- isomers of the n-paraffins.12 33 35-41 47 49 50p 73-86 The results of these investigations indicate that the (n - 1) skeletal stretching modes of a n-para& CnH2n+2 are confined to a definite range of frequencies depending mainly on the magnitude of the carbon-carbon stretching force constant.73 Bartholom6 and Teller 2. phys. Chem. 1932 B 19 366. i4 Bauermeister and Weizel Physikal. Z. 1936 37 169 7 5 Barriol J. Phys. Radium 1939 10 215. 7 G Barriol and Chappelle ibid. 1947 8 8. i i Parodi ibid. 1941 2 58. 7 8 Stepanov J. Phys. Chem. U.R.S.S. 1940 14 474. Eliashevitsch and Stepenov ibid. 1943 17 146. Kellner Trans. Faraday SOC. 1945 41 217. Whitcomb Nielsen and Thomas zbid. 1940 8 143. 81 Kassel J. Chem. Phys. 1935 3 326. 8 4 Mizushima and Simanouti J. Amer. Chern. SOC. 1949 71 1320. 85 Simanouti and Mizushima J. Chem. Phys. 1949 17 1102. 86 Gates ibid. p. 393. 8 2 Kirkwood ibid. 1939 7 506. 30 QUARTERLY REVIEWS The widest possible range may be put as 1200-700 cm.-l. Further as the value of n increases the highest and lowest frequencies approach definite limiting values ; the lower limit corresponding to the most symmetrical modes should appear as highly polarised Raman lines whatever the value of n.The (2% - 5) skeletal deformation modes are probably restricted to the region below 500 cm.-l. The most symmetrical of these are expected t o give rise to prominent polarised Raman lines which do not however correspond to the upper limit. Each trans-n-paraffin has two pairs of CH rocking vibrations cor- responding t o in-plane and out-of-plane motions. The data on ethane and propane (p. 28) suggest that their frequencies should lie between 1190 and 820 cm.-l Le. they fall in the same frequency range as the skeletal stretching modes. Interaction between the CH rocking modes and the skeletal stretching vibrations of the same symmetry type may therefore be expected.It should be emphasised that the value of these theoretical treatments lies in the guidance they have provided in the interpretation of the spectra. The simple methods do not give good agreement between the calculated and the observed frequencies. More ambitious discussions can overcome this difficulty but the effort involved is considerable. Complete treatments of both skeletal and hydrogen vibrations such as those given by Stepanov for n-butane and n-~entane,~l 36 could in principle be carried out for higher members of the series but it is doubtful whether such laborious calculations would add a proportionate amount to the understanding of the spectra. Experimental data for the trans-isomers of the n-paraffins are available for all molecules up to n-decane and for some higher members of the series.The Raman spectra were obtained by Mizushima and his c o - ~ o r k e r s . ~ ~ ~ 85 Raman and infra-red spectra have been published for the lower members by Rank and his collaborators,52-55 and further work on the infra-red spectra has been carried out in 87 A fairly complete assign- ment of the frequencies of trans-n-butane has been given 35 and sufficient data are also available to permit adequate discussion of the gauche isomer. Table 1 gives the characteristic frequencies that can be picked out in the spectra of the n-paraffins in the solid state; infra-red data are not available below 700 cm.-l. The series of prominent Raman lines near 1300 cm.-l (col. 3) which have rather weak infra-red counterparts cor- responds to CH wagging or twisting modes as already indicated in the introduction.Molecules with an even number of carbon atoms should on account of their centre of symmetry give rise to frequencies that appear only in the Raman or the infra-red spectra. Since the series near 1300 cm.-l shows up consistently in both types of spectra it must correspond to at least two modes with different selection rules but coincident frequencies. Four series (cols. 4-7) occur in the region 1200-800 cm.-l appropriate to the skeletal stretching and CH rocking frequencies. Like the 1300 cm.-l series that near 1060 cm.-l (col. 6) must correspond to at least two modes of vibration for it is observed in both types of spectra. Specific assign- 87 Brown and Sheppard unpublished dat,a. TABLE l.* Characteristic frequencies (in cm.-l) of the n-parafins in the crystalline state Paraffin __- n-Butane .n Pentane 17-Hexane . n-Heptane n-Octane . n-Nonane . ?a-Decane . n-Undecane n - D o decane n-Tridecane n-Tetra- decane ?a- Hexa - decane n - Nona - decane Infra-red mtensity Raman mtensity I R R I R R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R - 1350 1300 1308 1303 1307 1300 1305 1296 1303 1297 1298 1297 1303 1295 1290 vc-c or CH rockmg f 1 965 1151 f 1138 1025 1145 1031 f ' 1057 1143 f 1140 1139 f 1138 1138 1136 f 1075 - 1088 f 1094 - 1103 I 984 ' f I 732 1059 837 1 - i 1066 ' f 726 1064 1 898 1 - [lo571 910 ' 723 [I0561 1 905 1 - I 1058 f ' 722 1062 ~ 899 1 - [lo631 889 ' 721 [lo601 I 888 - 1062 f 720 1060 1 886 - 1136 j f 1137 1 1106 [lo641 1 889 1 720 <- No observations 1303 1112 1062 f 722 1297 1 1:36/ f 1061 892 - a, Ref.f 426 NI 406 f 373 NI 31 1 f 283 NI 249 f 231 NI --+ f 194 N I 1304 1 1138 1110 [lo641 893 723 - No observations ____I__ 1306 ~ f 1114 ' 1060 f 1 722 1 f P No observations --r t-------- No observations -+ 1135 f 1068 888 - 160 1300 1296 ~ 1134 1121 [lo561 890 720 NI - No observations -j I 56a 84 65b 84 56b 84 87 84 35 84 35 84 36 84 87 - 36 84 87 35 - - - 84 87 - [ ] indicates two modes with coincident frequencies * Explanation of nomenclature employed in Tables 1 4 and 6-9 API American Petroleum Institute IR Infra-red R Raman NI No data available No entry no frequency can be recognised VC-C skeletal stretchmg mode 6~ skeletal deformation mode f forbidden st strong m.st medium strong m medium w weak v.w very weak abs.absent not observed w./t. = waggmg or twwting. - r = roolung * no vibration exists 32 QUARTERLY REVIEWS ments for each of these series have been suggested,35 but only in the case of the highly polarised Raman lines near 890 cm.-l (coI. 7) is there general agreement-they correspond to the lower limiting skeletal stretching modes.10 359 84 85 Prominent infra-red absorption bands which have no observable Raman counterparts are found in the region 1100-700 cm.-1; these correspond to the CH rocking modes. All molecules for n > 5 show the intense bands near 720 cm.-l (col. 8 ) which are characteristic of this type of vibration ; these can almost certainly be assigned to the lower limiting vibrations. In addition bands which spread to higher frequencies (up to 1050 cm.-l) split off as the chain length increases ; 35 they seem to be approaching an upper limit as is also suggested by the theoretical treat- m e n t ~ .~ ~ The series of polarised Raman lines below 500 cm.-l (col. 9) cor- responds to the most spmetrical skeletal deformation vibrations.lO+ 35 g49 85 Polythene must be mentioned here since it represents an infinite paraffin chain with the [-CH,*CH,-] repeating unit. For detailed discussions reference must be made to the original papers ; 35 38-41 4 5 85 88 89 it is sufficient to note that the assignments of the polythene frequencies can be correlated in a satisfactory way with those given above for the truns-isomers of the n-paraffins. ( d ) Spectra of the n-Parafhs in the Liquid State.-Experimental data on the Raman and infra-red spectra for the n-paraffins in the liquid state are available for all molecules up to and including n-tetradecane and for some higher members.The prominent regularities already noted in the spectra of the trans-n-paraffins are also apparent in the spectra of the liquids which consist of mixtures of rotational isomers. Some modifications in intensity and activity are observed because in many cases the non-planar isomers may have identical frequencies but different selection rules as e.g. in the 1140 cm.-l series. Other series may be recognised which must be due only to non-planar isomers.49 Since all these are of some practical importance for the identification of unbranched chains they are given in Table 2 together with brief comments on their behaviour. Most of these series have effectively constant frequencies.Others in which progressive shifts are observed have'been noted among the weaker bands in the range 1320-1220 cm.-l; these must correspond to additional CH wagging and twisting modes.' I n the spectra of the liquids the region 800-700 cm.-l is somewhat complex. The intense band at 720 cm.-l is accompanied by subsidiary bands at higher frequencies which can also be arranged in shifting sequences. It may be noted in this connection that paraffins with an unbranched chain of four carbon atoms show an intense band near 740 cm.-l (compare n-butane 747 and 731 cm.-l) and those with a chain of three carbon atoms i.e. an ethyl group one usually near 770 cm.-l; these are to be attributed to CH rocking modes of the shorter chains (see also ref. 23b). From the above discussion it may be concluded that the spectra of the liquid n-paraffins are closely related if there are five or more carbon atoms 88Thompson and Torkington PTOC.Roy. SOC. 1945 A 184 3. 89 Elliott Ambrose and Temple J. Chem. Phys. 1948 16 877. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 33 present. In the Raman spectra there is some alternation of frequency for 11 even and n odd but this is not distinguishable among the infra-red bands.5 The most prominent frequencies in the infra-red spectra are those associated with tlhe hydrogen deformation modes. It follows that the features that TABLE 2. Characteristic frequencies (in cm.-l) of the n-parafins in the liquid state v approx.* E l . . 13401 1305 . . . 1140 . . . 1120-965 . 1070 . . . 890 . . . 870-840 840-810) ' 720 .. . 425-150 . . Infra-red spectra Moderately prominent Moderate Moderate Moderate Difficult to Weak ; Weak ; distinguish difficult to distinguish difficult to distinguish Intense No data Raman spectra Absent Prominent Prominent Weak Prominent Prominent ; polarised Prominent ; polarised Absent Prominent ; polarised Assignments and comments Hydrogen deformation modes of non- planar isomers ; nearly constant n 2 5 CK wagging or twisting modes ; nearly constant n> 5 ; intensities of the bands increase as n increases Nearly constant n 9 4 Intensities and frequencies of the bands increase as n increases ; gives indica- tion of the chain length Nearly constant n 2 4 Lower limiting skeletal stretching modes ; nearly constant n > 5 Lower limiting skeletal stretching modes of non-planar isomers ; alter- nating frequencies n even < n odd CH rocking modes; nearly constant Symmetrical skeletal deformation modes ; decrease regularly as n in- crease s n> 6 * Characteristics frequencies found also in spectra of the trans-isomers are in italics.are valuable for recognising the n-paraffins or the presence of an extended unbra,nched carbon chain are limited in number the most reliable being the group of Raman lines between 900 and 800 cm.-l and the infra-red absorption bands in the region 730-700 cm.-l. 4. The tert.-Butyl Group (a) neoPenta,ne.-neoPentane ( 2 2-dimethylpropane) may be considered as the first member of the series of p a r a f i s having the general formula CMe,R where R is an alkyl radical. This symmetrical molecule has apart from the CH torsional modes seven distinct vibrations which should occur in the region below 1350 cm.-l two of which are inactive in both the Raman and the infra-red spectra.The substance has been extensively investi- gated 1 2 ) 25) 349 48 4 9 7 and attempts to assign the frequencies have been A.P.I. Infra-red Spectrograms Nos. 104 442 514 784. 91 Young Koehler and McKinney J . Amer. Chem. SOC. 1947 69 1410. 9 2 Rank and Bordner J . Chem. Phys. 1935 3 248. 93 Kohlrausch 2. phys. Chem. 1936 23 28 340. C 34 QTTARTERLY REVIEWS made by many authors but as yet complete agreement has not been achieved. The following are those assignments (Y in cm.-l) about which there is no controversy (see also Table 3) i z l Symmetrical skeletal stretching vibration 733 335 41 5 n2 Doubly degenerate skeletal deformation vibration n4 Triply degenerate skeletal deformation vibration Ambiguity arises over the interpretation of the 1249 and 921 cm.-l fre- quencies both of which appear as depolarised Raman lines and intense infra-red bands.The assignment proposed originally by Kohlrausch and Koppl 49 has since been favoured by a number of authors.34* 4 8 9 91 9 2 p 94 This puts the 1249 cni.-l frequency as the triply degenerate CH rocking mode and the 921 cm.-l frequency as the triply degenerate skeletal stretch- ing vibration n,; the reverse interpretation has been preferred by other investigator^.^! 109 46 95* O6 Most of these authors have attempted some TABLE 3. Relationships between the skeletal modes of neopeiztune and those of parafins with one tert.-butyl group ?a1 . . 1 a .. 2 n3 . . 3 Q1 . . 3 Paraffins with one CH,-C-K. krt.-butyl group I CH P ia dP ia dP a \ dP a \ Selection rules. Numbering Degeneracy I I Numbering v1 . . vp . . -vg . . .v* . . -vg . . .vg . . Selection ru!es Degeneracy Raman mfra-red I P dP dP dP P P p = polarised ; dp = depolarised ; a = active ; ia = inactive kind of theoretical treatment. Pitzer and Kilpatrick 46 have suggested schernat,ically that the two inactive CH rocking modes are at about 1150 cm.-l for the doubly and 950 cm.-l for the triply degenerate vibrations while Rasmussen 34 prefers 1200 and 900 cm.-l respectively. Unfortunately it does not seem possible to come t o any final decision on the data a t present available for neopentane and investigations of the deuterated derivative (CH,),C*CH,D have not been conclusive.48 (See the discussion between Rank 97 and Sheppard Simpson and Sutherland.98) For the purpose of considering the characteristic frequencies of the tert. - butyl group this is not a serious difficulty. 9 4 Wagner 2. pJbys. Cham. 1939 B 45 341. 95 Wall and Eddy J. Chem. Phys. 1938 6 107. 9.5 Silver ibid. 1940 8 919. 97 Discuss. Faradccy SOC. 1960 9 219. s8 Ibid. p. 216. TABLE 4. Characteristic frequencies (in cm.-l) of parafins with tert.-butyl groups in the liquid state * ,4. Paraffins with one tert.-butyl group. NI 364 N I 319 ~~ Paraffiii vc or CH roching voc d, 1 Refs. API Nos. 13 14 13 14 2 2-Dimethylbutane . . 2 2-Dimethylpontane . . 2 2-Dimethylhexane . . 2 2-Dimethylheptane . . 1223 1206 1212 1205 1210 2 2 3-Trimethylbutane 2 2 3-Trimethylpent,ane 2 2 3-Trimethylhexane 2 2 4-Trimethylpentane 2 2 4-Trimethylhexane 2 2 4-Trimethylheptane 2 2 4-Trimethyloctane 2 2 5-Trimethylhexane 2 2 6-Trimethylheptane 2 2-Dimethyl-3- ethylpentane ethylhexane .. . . . . . . . 2 2-Dimethyl-4- 924 720 1 310 928 745 NI 931 749 301 926 744 NI 927 744 305 IR R IR R IR TC IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R - - (2) 1252 1258 1250 1254 1250 1251 1250 1251 - 1252 1255 1244 1245 c- 1239 1247 1254 1247 1250 t- 1252 t- 1252 1248 1257 1255 1249 1236 1236 t- 1245 (3) 1217 1221 1209 1210 1202 1204 1203 1202 929 929 926 92 8 932 928 932 930 71 1 714 74 1 746 - 746 - 750 NI 361 NI 340 NI 303 NI 283 686 686 716 717 Ins - _ f - 1205 1206 1204 1201 1218 1226 920 929 920 923 919 925 13 655 670 14 13 571 588 14 25 14 2 14 - 746 - 745 729 727 NI 302 NI NI 308 - To observations -> 1205 926 739 307 15 13 14 13 14 - 15 - 15 13 14 25 14 13 14 - 15 573 587 576 590 577 589 667 676 579 592 683 685 36 QUARTERLY REVIEWS B.Paraffins with two tert.-butyl groups. Paraffin 2 2 4 4Tetramethyl- 2 2 5 5-Tetframethyl- 2 2 3 4 4-Penta- peiitaiie hexane nicthylpentane met hgl hexane methylheptane 2 2 3 5 5-Penta- 2 2 4 6 6-Penta- Infra-red intensity . . . Raman intensity . . . . vc-c or CHa rocking vCc 6c-c Refs API Xos. IR 1244 R 1249 IR 1247 R t R t - IR f- IR 1238 R 1255 IR 1244 R 1248 st - 916 730 13 583 591 - 920 731 11 - 910 755 13 690 691 No observations + - 1214 930 748 13 693 694 No obsorvat'ions + - No observations ___+ - 1211 924 753 15 5 1209 924 758 14 1208 927 ~ - 1 m m i s t l m *For the meaning of symbols and abbreviations see the foot of Table 1.( b ) Paraffis with one tert.-Butyl Group.-Before discussing the fre- quencies characteristic of the tert.-butyl group it is helpful to consider briefly how the skeletal vibrations of ~eopentane are likely to alter when the symmetry of the molecule is reduced by the substitution of an alkyl radical R for one CH group. Table 3 shows the relationships between the modes of the parent and the derived molecules together with the degener- acies and the selection rules; these relationships are well In order t o get some idea of the magnitude of the splitting of the frequencies of n3 and n4 and the variation with the nature of R of the frequencies v1 to v6 for the molecules CMe,R it is necessary to assume a suitable model and to make calculations using force constants chosen from the values available in the literature.Two such investigations have been made. Kohlrausch and his co-workers loo were interested primarily in those mole- cules in which R is a halogen atom ; Simpson and Sutherland * were concerned with the 2 2-dimethyl paraffins. The conclusions reached in both treatments are similar. The vibrations vl v4 and vj shift steadily to lower frequencies as the mass of R increases ; Y after an initial decrease changes more slowly and appears to approach a limiting value while the modes v2 and v6 remain effectively invariant. Further it seems probable that the maximum split- ting of n for the 2 2-dimethyl paraffins is unlikely to be greater than 50 cm.-l though that of n4 is probably considerably larger.* 99 Kohlrausch op.cit. 1943 p. 168. 100 Idem ibid. p. 188. SHEPPARD AND SIMPSON SPECTRA O F HYDROCARBONS. PART I1 37 Tables 4A and 4B summarise the characteristic frequencies in the Kaman and infra-red spectra of the liquids for most of the paraffins having one or two tert.-butyl groups hitherto investigated (some very complicated molecules are not included). Figs. 4A and 4B show typical spectra for a paraffin of this class. The tert.-butyl group provides a particularly favour- able configuration for the discussion of the characteristic frequencies of such a unit because it forms a rigid group which cannot because of its threefold symmetry axis itself give rise to rotational isomerism i . e . the existence of rotational isomers in any molecule in which it occurs must be dependent on changes in the radical R.It seems plausible to suppose that such changes are unlikely to influence t'he vibrat'ions of the tert.-butyl group though direct experimental evidence on this point is not available. The various characteristic frequencies of molecules having a single tert . - butyl group will now be discussed in turn. As already noted (p. 22) there is a Splitting of the band due to the CH synimetrical deformation modes to give a doublet a t 1397 and 1370 cm.-l. The other series as given in Table 4 can all be related to frequencies in neopentane by using the theoretical conclusions outlined above. Col. 6 lists the frequencies of the symmetrical stretching vibration vl of the C unit associated with the tcrt.- butyl group (750-7 10 em. -I). These frequencies are easily recognised in the Raman spectrum on account of their strength and high degree of polarisation ; 10 the corresponding infra-red bands are not prominent and are in many cases overlapped by strong absorption due to the CH rocking modes.This series is related to the Raman line a t 733 cm.-l in the spectrum of neopentane. From the figures given it is apparent that the actual fre- quency of this mode is dependent on the nature of the radical R ; thus if R is branched next to the tert.-butyl group (as in 2 2 3-trimethylbutane) a marked lowering occurs. The second general trend is that as the inass of R increases for a given type of R the frequency of this vibration also increases. This is contrary to the theoretical predictions but as they were based on drastically simplified models it is perhaps not surprising.Col. 7 gives the frequencies of the symmetrical skeletal deformatmion vibration v5 as observed in the Raman spectra below 400 cm.-l. This mode can be picked out with fair confidence on account of its polarisation but is by no means easy to identify in the more complex molecules. The actual frequencies decrease with increasing mass of R and seem to be dependent to some extent on its structure. This vibration on account of its considerable variation is less valuable as a means of identifying the tert.-butyl group. The skeletal deformation modes v4 and v6 which should occur as depolarised Raman lines below 500 cm.-l cannot be recognised with certainty though tentative series have been proposed for these vibrations in the simple 2 2-dimethyl paraffin^.^ The two frequencies near 1250 and 1210 cm.-l (cols.3 and 4) are of the greatest practical importance in the identification of the tert.-butyl group as they almost invariably appear as fairly prominent Raman lines 8 10 and as very strong infra-red b a n d ~ . ~ j They undoubtedly arise by the splitting of the 1249 crn.-l frequency in neopentane. According to the FM. 4A FIG. 4B A. IrLfm-retl spedtu bctwceiz 1500 t o i t ? 300 ctti.-I o j some / y p i c i l 0tmtclted pwujirts i t i the liqzcid stalp. (A,flcr -4 mericcrrr J’cft CIP14)il I/,stitzite I,zfra-red Spcclropums Nos. 57 1. 588 ; 552. 554 ; 372 58s 002. ti1 1 .) ( VtJrticn/ I i n p s indicnle (*lrnracfcristic ahsorptior& frequemic.9 w~e,itiotr~tl it1 llrc tc rt.) 33. Ramatb spectra betwee,% 1500 atid 300 c m - 1 of the sume paraffh i t z the liquid state.(After B’enske et al.14) (Characteristic frequencies metrtioned in the text are indicated by *.) I 2 2-Di?neti~ylpe,ita}ze. III 3 3-Dimethy1penta;ne. 11 2-Methylhexune. IV 3-Methylhexane. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 39 interpretation given for the symmetrical molecule (p. 33) they may be assigned either as CH rocking modes of the tert.-butyl group or as the two skeletal vibrations v2 and v3. As may be seen from col. 3 the 1250 cm.-l frequency remains remarkably constant whatever the nature of the radical Rt. The other series near 1210 cm.-l (col. 4) shows more variation ; there is a general lowering as the mass of R increases though the lowering is less pronounced if R is branched near the tert.-butyl group.The behaviour of the 1250 and 1210 cm.-l series is in general agreement with the theoretical predictions for v2 and Y, though this is not conclusive evidence in favour of these assignments. The relative intensities of the infra-red. bands near 1250 and 1210 cm.-l are interesting; in general the former are stronger unless R is branched near the tert.-butyl group in which case the lower frequency becomes more prominent. Col. 5 gives figures for the remaining frequency (near 930 cm.-l) that can be definitely associated with the tert.-butyl group. This appears as a prominent Raman line and a recognisable but not very strong infra-red band. Like the 1250 cm.-l frequency it remains nearly invariant whatever the nature of R but is less valuable for the identification of the tert.-butyl group because many other paraffins have frequencies in this region.The 930 cm.-l series in molecules with a tert.-butyl group is undoubtedly associ- ated with the 921 cm.-l frequency in neopentane and so may be assigned either to the invariant skeletal modes v2 or to CH rocking vibrations of the tert.-butyl group. It should be noted however that if the former interpretation is preferred it is not possible to pick out any series which could be attributed to the other skeletal vibrations Y,. Of the substances listed in Table 4A only 2 2-dimethylbutane has been the subject of detailed consideration probably because rotational isomerism is absent in this molecule. The conclusions of Kilpatrick and Pitzer lol and Sheppard lo2 are in fair agreement. Table 4A does not include the data for some ten more complicated paraffins with one tert.-butyl group. All the characteristic frequencies noted above appear consistently except that the 1210 cm.-l frequency is missing for the two molecules with a 2 2 3 3-tetramethyl grouping. It is also interesting to find that these same frequencies may readily be recognised in the spectra of tert.- butylethylene and tert.-butylacetylene (3 3-dimethylbut-1 -ene and -1 -yne). l o 2 Two somewhat weaker frequencies near 1025 and 880 cm.-l which appear to be associated with the tert.-butyl group in these molecules cannot be recognised with certainty in the spectra of the relevant paraffins. ( c ) Paraffis with two tert.-Butyl Groups.-The molecule 2 2 3 3- tetramethylbutane might be considered as the first member of the series of paraffins with two tert.-butyl groups.However both the electron- diffraction data loa and the spectroscopic evidence lo* suggest that it has a staggered configuration analogous t o that of et'hane. Because of the I o 1 Kilpatrick and Pitzer J. A ~ 2 c v . P h p i i ? . Sot. 1918 68 1066. lo? J . (:hem. Pltys. 1949 17 466. lU3 Bauer and Beach J . Amer. Chem. Soc. 1942 64 1142. lo* Cleveland Lamport and Mitchell J. Chew. Phys. 1950 18 1320. 40 QUARTERLY REVIEWS resulting centre of symmetry the skeletal vibrations of such a molecule cannot be completely related to those of other substances with a tert.-butyl group. The spectra of this paraffin have been investigated by a number of authors 2 5 9 27b 4 7 9 l o 4 9 lo5 and an assignment of its frequencies which seems satisfactory has recently been given.lo6 The data on other molecules with two tert.-butyl groups are somewhat scanty; they are collected in Table 4B. Most of the frequencies charac- teristic of a single tert.-butyl group (except the skeletal deformation mode) may be recognised without difficulty in the spectra of these more complex paraffins. 5. The isoPropy1 Group ( a ) isoButane.-isoButane may be considered as the first member of the series of paraffins having the general formula CHMe2R where R is an alkyl group ; it has nine distinct vibrations (other than the CH torsional modes) that should occur in the region below 1350 cm.-l all but one of which are active in both Raman and infra-red spectra; the remaining mode is forbidden. This molecule has been the subject of many investi- gations l 7 3 2 4 9 25 34 1°79 lo* but no definite assignment of the frequencies is generally accepted.Apart from the doubly degenerate CH deformation mode observed as a Raman line near 1335 cm.-l the only vibration about which there is no controversy is the non-degenerate skeletal stretching mode n at 795 cm.-l (see also Table 5). The two skeletal deformation frequencies occur a t 435 and 370 cin.-l. Most authors allot these respectively to the symmetrical non-degenerate (n2) and unsymmetrical doubly degenerate (n4) vibrations on the basis of Ananthakrishnan's polarisation measurements. lo8 Simpson and Sutherland have given reasons for preferring the opposite interpretation but further work will be needed before a final conclusion can be reached. There has been considerable discussion about the assignment of the unsymmetrical doubly degenerate skeletal stretching vibration n3 which has been variously located at 925,46 965 34 94 and 1170 ~ m .- l . ~ lo The infra-red spectrum of the molecule (CH,),*CD has been published but not analysed.log It is clear t'hat some of its modes of vibration are coupled together (compare discussion on propane I->. 28) ; the observations do not however allow the identification of' n3. The inactive CH rocking mode has been schematically put at 1200 cm.-l by Pitzer and K i l p a t r i ~ k ~ ~ and at 900 cm.-l by Rasn~ussen.~~ The active singly degenerate CH rocking vibration probably occurs at 1095 The two active doubly degen- erate CH rocking modes correspond to the two frequencies of those listed above which are not attributed to the unsymmetrical skeletal stretching vibration.105 A.P.I. Infra-red Spectrogram No. 444. Io6 Scott Douslin Gross Oliver and Huffmann J. Arner. C'hern. Soc. 1952 74 883 Io7 A.P.I. Infra-red Spectrograms Nos. 62 374 439 1142. Io8 Proc. Indian Acad. Sci. 1936 A 3 527. log Condon McMurry and Thornton J. Chem. Phys. 1951 19 1010. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART 11 41 It is unfortunate that the assignment of frequencies of this key molecule is so uncertain ; however it is possible to discuss the frequencies character- istic of the isopropyl group without difficulty although their exact inter- pretation is ambiguous. ( b ) ParafEns with one isoPropyl Group.-It is clear from the foregoing discussion that it is profitable to consider the relationships between the skeletal vibrations of the most symmetrical and the other members of a related series.Thus in the present instance it is desirable to know how t'he skeletal modes of isobutane are related to those of other molecules of the type CHMe,R. The general arguments are analogous to those already given for the tert.-butyl group and the relevant calculations have been carried out on similar lines by the same author^.^ loo Their conclusions are summarised in Table 5. TABLE 5. Relutionships between the skeletal modes of isobutanw uncl those of parafins with one isopropyl group H I XsoButanc C€13-C -CII Numbering Selection rules Degeneracy Raman H I wopropyl group I CH3 Paraffins with one CJI,-C-R Selection rules - Numbering Raman '::$- 1 P 1 2 dp a -v3 . A V g . /v4 2 dp a / -v6 . . Effect on frequency of Increase of mam of R Decreases fairly slowly Decrease.steadily Effectively constant Decreases fairly slowly Effectively constant Decreases steadily Tables 6A and 6B list the characteristic frequencies in the Raman and infra-red spectra for a selection of pmaffins having one or two isopropyl groups respectively ; all substances were investigated in the liquid state. Figs. 4A and 4B show typical spectra for a paraffin of this class. Unlike the tert.-butyl group the isopropyl unit is not symmetrical so that even the simplest molecules (such as 2-methylbutane) 61 can exist as mixtures of rotational isomers in the liquid state. Moreover in all the molecules listed (except 2 2 3-trimethylbutane where such isomerism is absent) the isopropyl group itself is actively concerned in the variations in configuration.Since it is probable from the work on molecules where one form has been isolated that certain frequencies of the various isomers of a given molecule may differ considerably it follows that those given in Table 6 may in fact correspond to different forms of the molecules. However it is possible to make quite reasonable assignments on the assump- tion that they all correspond to one particular type of isomer although 42 QUARTERLY REVIEWS TABLE 6. Characteristic frequencies (in crn.-l) of parafins with isopropyl groups in the liquid state * A. Paraffins with one isopropyl group. Paraffins %-R IR R 1343 I R R 1344 I R R 1345 IR R 1344 IR R 1345 IR f i G C 463 459 N I 442 N I 430 N I 434 N I KI - - BPI Sos 550 55i 55 1 55. 552 55 353 55 605 614 608 614 - 1149 1154 1147 1154 1145 1148 1144 1149 1143 1148 1142 - - 953 959 964 - 957 961 968 95 1 961 957 - ~ 920 920 918 928 919 - I - I 919 - 918 - 794 794 816 815 822 822 813 814 824 826 833 - 1176 1175 1172 1176 1171 1179 1171 1177 1167 1175 1167 2-Methyl- butane 2 -Methyl- pentane 2-Methyl- hexane 2-&hhyl- heptane 2 -Methyl - octane 2-Methyl- nonane 4-isoPropyl- heptane N I 368 XI 325 N I 311 N I 289 NI 256 XI 13 14 13 14 13 14 13 14 13 14 13 - 927 920 928 931 920 917 920 923 - 832 830 826 829 826 828 835 833 - 13 14 13 14 13 14 25 14 IR 1161 - 957 R 1331 - - - I R 1168 - 956 R - 1176 - 957 IR 1170 1121 950 R 1345 - 1125 958 IR 1170 1125 949 R - - 1125 - N I 449 N I 423 N I 407 N I - 573 587 577 589 579 592 2 2 3-Tri- met h ylbutane met hylpent an6 me thylhexane methylheptanc 2 2 4-Tri- 2 2 5-Tri- 2 2 6-Tri- B.Parfins with two isopropyl groups. 2 3-Dimethyl- 2 4-Dimethyl 2 5-Dimethyl- 2 &Dimethyl- butane pentane hexane heptane - I R R 1348 1165 I R 1170 R 1348 1164 I R 1171 R 1341’1175 I R 1170 R 1341 1171 1153 - - - - 1154 1150 1149 ~ 921 919 924 920 911 920 - - N I N I 414 N I 440 N I 423 - N I - N I 306 N I 263 N I 255 13 14 13 14 13 14 2 15 656 6iC 659 672 663 674 - 939 956 956 949 963 950 953 ‘ G 809 810 839 840 835 840 * For the meaning of symbols and abbreviations bee foot of Table 1. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 43 the resulting conclusions must be considered as tentative until confirmatory experimental evidence becomes available. The various characteristic fre- quencies of molecules having a single isopropyl group will now be discussed in turn.As already noted (p. 22) there is a splitting of the band due to the CH symmetrical deformation modes of the group to give a doublet a t 1385 and 1370 cm.-l. The CH deformation vibrations give rise to a series of Raman lines near 1345 crn.-l,lO which are shown in Table 6A col. 3. Col. 8 gives the series (835-749 cm.-l) that can certainly be attributed to the symmetrical stretching vibrations vl of the isopropyl group. This is easily recognised in the Raman spectrum on account of the strength and polarisation of the lines ; the corresponding infra-red bands are usually weak.10 These frequencies correspond to the vibration observed a t 795 cm.-l in the spectra of isobutane. Prom the figures given it is apparent that the actual frequency of this mode is not strongly dependent on the nature of R though a slight shift to higher frequencies may be noted as the mass of R increases.This variation is not in agreement with the predictions of the simple theory. Col. 9 gives the series of polarised Raman lines (465-430 cm.-l) that may be assigned to the constant skeletal deformation vibrations v4. These may be recognised without much difficulty lo and could be used to detect the presence of an isopropyl group although in such paraffins there are usually a number of lines in this spectral region. Col. 10 lists another series of polarised Raman lines below 400 cm.-l. These lines may be assigned to the skeletal deformation modes Y, and diminish in frequency as the mass of R increases in accordance with the theoretical predictions.They may be identified with fair certainty in the simpler molecules but would be of little value for recognising the presence of the isopropyl group. The series of depolarised lines below 500 cm.-l corresponding to the skeletal deformation modes Y cannot be definitely identified ; a tentative inter- pretation has been proposed for the 2-methyl paraffin^.^ Smith has noted a series of weak infra-red bands between 550 and 530 cm.-l appearing regularly in the spectra of paraffins with an isopropyl group which can possibly be assigned as combination bands. The spectra in this region are however rather complex so this series is not very helpful for structural analysis. The frequencies near 1170 cin.-l listed in col. 4 are those which are of the greatest importance in the identification of the isopropyl group ; this series is always prominent as strong bands in the infra-red spectrum 49 5 that usually have corresponding Raman lo The frequencies are very insensitive to changes in the nature of R and are undoubtedly associated with the 1170 cm.-l frequency in isobutane.Col. 5 lists a fa'irly constant frequency near 1150 cm.-l which appears regularly in the spectra of the 2-methyl pa,raflfins though not generally i n the more complex molecules. This series also appears t'o be associated with the 1170 cm.-l frequency of isobutaiie. These two series may be interpreted either as CH rocking modes of the isopropyl group or as the skeletal stretching vibrations v3 and v5. 44 QUARTERLY REVIEWS The series listed in col. 6 appears as prominent Raman lines and as infra-red bands of variable intensity near 955 cm.-l; the frequencies remain almost unaffected by alteration in the structure of the radical R and may be correlated with the vibration appearing a t 965 cm.-l in the Raman spectrum of isobutane.Col. 7 shows a series near 920 cm.-l which appears as prominent infia-red bands that sometimes have corresponding Raman lines. These frequencies show very little variation; they would seem to be correlated with the 925 cm.-l absorption in isobutane. Paraffins with a tert.-butyl group also have frequencies in this region (Section 4). It should be noted that there is no real evidence for any consistent split,ting of either the 955 or the 920 cm.-l series in isopropyl compounds i.e. if either of these is assigned to the skeletal modes v3 there are no characteristic frequencies recognisable in this region that could correspond to skeletal vibrations v,.Data are available for some 25 more highly branched paraffins with isopropyl groups. Of the characteristic frequencies listed in Table 6A the Raman lines near 1345 cm.-l are usually present. The 1170 and 955 cm.-l frequencies are nearly always found in both infra-red and Raman spectra and those near 920 cm.-l as infia-red bands. A frequency corresponding to the symmetrical stretching mode v1 and another that may be attributed to the consta,nt skeletal deformation vibration v4 may be identified in the Raman spectra of about half the molecules. The behaviour of the 540 em. -l infra-red series is somewhat erratic ; the 1150 cm.-l frequency is usually absent and the deformation vibration Y cannot be distinguished with certainty.(c) Paxaffis with two isoPropyl Groups.-The data on the simpler paraffins with two isopropyl groups are summarised in Table 6B which shows that most of the characteristic frequencies of the isopropyl group may be easily recognised in these more complex molecules. The frequencies of 2 3-dimethylbutane seem a little anomalous. In the liquid this molecule is known to consist of a mixture of two rotational isomers one of which has a symmetrical structure with a centre of symmetry.61 This may possibly account for the discrepancies observed. Further the two isopropyl groups must be closely coupled together so that deviations due to their interaction may be expected. 6. Paraffins with an Internal Quaternary Carbon Atom C I C The molecules of this class possess a C structural unit C-A-C in common with those with the tert.- butyl group previously discussed (Section 4). Consequently analogies may be sought in the specha of these two classes of paraffin. Further it might be expected that the assignments made for neopentane would prove helpful in the interpretation of the spectra ; this method of approach has already been used by both Sheppard lo and Smith.5 A number of paraffins with an internal quaternary carbon atom have been SHEPPARD AND SIMPSON SPECTRA O F HYDROCARBONS PART I1 45 1216 1210 1211 1230 1229 TABLE 7. Characteristic frequencies (in cnz. -I) of pamfiIis with an tnternal qua'tern-crry carbon atom in the liquid state * - 1193 1195 1188 - Paraffin 1017 * * 1010 1018 1016 - 1000 999 1012 1015 3 3-Dmethyl- pentane hexane 3 3-Dimethyl 3 3-Dmethyl- heptane 2 3 3-Trimethyl pentane 2 3 3-Trimethyl hexane 2 3 3 4-Tetra- me thylpentane 3 3 4-Trimethyl hexane 3 3 5-Trimethyl heptane 3 3 4 4-Tetra- methylhexane 4 4-Dimethyl- heptane 2 4 4-Trmethyl hexane 658 - - NI 753 493 723 481 721 478 686 468 681 475 N I N I 706 485 - 455 684 - 3 -Ethyl- 3 - met h y lpen t ane met h ylhexane 3-Ethyl-3- 1232 - 1235 1229 3 3-Diethyl- 3 3-Diethyl- pentane hexane 1179 1192 1183 1191 Infra-red intensity .. Raman intensity - IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR It IR R IR R IR R IR R - - - - (2) I c-c 01 CH roiking 1217 1217 1212 1208 1215 121 1 1218 1206 1210 1221 1227 1205 121 1 1211 I 1192 1200 1189 - 1187 - 1188 - 1186 1191 1183 1186 1182 1192 1186 +- No ( I 1225 I 1190 - - 1000 1007 1006 1010 1005 - - - * * * * 1006 1011 - 695 695 725 722 727 724 672 676 689 69 1 672 67 1 701 704 721 482 482 482 492 NI 478 472 470 490 483 47 1 467 482 484 - 1220 1175 1015 733 N I No observations - Hefb 13 14 13 14 2 14 13 14 13 14 13 14 13 14 13 - - 14 2 14 13 14 13 14 2 14 13 15 2 - - A H Nos 572 588 574 589 578 590 668 676 585 693 681 682 686 688 580 592 581 591 * For the meanmg of symbols and abbreviations see foot of Table 1.46 QUARTERLY REVIE\VS investigated ; Table 7 includes the relevant experimental data on the Raman and infra-red spectra of the liquids for all these except for the four which also possess a tert.-butyl group. Fig. 4A and 4B show typical spectra for a paraffin of this class. As noted on p. 22 there is a characteristic splitting of the band due to the CH deformation modes of the 3 3-dimethyl paraffins to give a doublet a t 1384 and 1367 cm.-l.Table 7 shows that corresponding to the 1250 1210 cm.-l frequencies of the tert.-butyl group two series are found consistently in the infra-red spectra of paraffins with an internal quaternary carbon atom ; the corresponding Raman lines are not always present. One series near 1190 cm.-l remains nearly constant in frequency (col. 4) and in the infra-red spectra is more intense than the other near 1210 cm.-l (col. 3) which is more variable in its behaviour. The former provides the most reliable method of distinguishing the presence of an internal quaternary carbon atom. The region 1100-1000 cm.-l which is not very helpful for structural analysis gives rise to a number of characteristic frequencies in molecules with an internal quaternary carbon atom that seem to have no counter- parts in paraffins with a tert.-butyl group.The bands in this range are somewhat complex with much splitting but regions of prominent absorption are apparent near 1100 1040 and in most cases near 1010 cm.-l and many molecules also show corresponding Raman lines. The frequencies of the first two series behave in a somewhat erratic fashion there being no obvious relationship to the structures. Tentatively they may be assigned to CH rocking modes not involving the C structural unit since this is the only type of vibration in the appropriate range common to all the molecules listed and similar frequencies are found in some other classes of paraffins (compare Table 10).The absorption bands near 1010 cm.-l (col. 5) are always prominent in paraffins with an internal quaternary carbon atom if they also have at least one ethyl group ; the Raman counterparts are usually absent. It seems plausible to suggest that these bands are to be assigned to the carbon-carbon stretching vibrations within the ethyl group analogous to the mode that appears as a Raman line a t 993 cm.-l in the spectrum of ethane. Such molecules also have intense bands near 770 cm.-l which were noticed by Smith ; these are attributable to the CH rocking modes of the ethyl groups. Characteristic frequencies analogous to those near 930 cm. -l in the spectra of the tert.-butyl compounds might be expected to appear for paraffins with an internal quaternary atom in the range 950-900 cm.-l.The infra-red bands in this region are usually complicated and on the whole not very intense ; Raman lines are found for some molecules near 930 cm.-l and in the range 920-910 cm.-l. It is difficult to distinguish any plausible series that is consistently observed for all these molecules and could be used to characterise them especially as the other groups present also have frequencies in this region. As may be seen from col. 6 all paraffins with an internal quaternary carbon atom possess a strong polarised Raman line that may be assigned SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 47 to the symmetrical stretching mode of the C unit (750-685 cm.-l) ; the corresponding infra-red bands if they can be picked out from among those of the CH rocking vibrations are weak.These frequencies are analogous to those already described for the tert.-butyl compounds and correspond to the 733 cm.-l Raman line in neopentane. Moreover their variation is similar; in particular substitution near the C unit produces a marked lowering in frequency. This variation reduces the usefulness of these lines as a means of identifying the presence of an internal quaternary carbon atom. Below 500 cm.-l two series of Raman lines can be picked out that may be assigned t o skeletal deformation modes of the C unit though the exact relationship t o the corresponding vibrations of the tert.-butyl group is not very clear. One series (col. 7) has a fairly constant frequency; the other between 350 and 305 cm.-l (not listed) is polarised and the frequencies decrease as the molecule becomes larger.Neither would be very valuable as a method of identification as paraffins with an internal quaternary carbon atom possess a number of Raman lines in this region. The above discussion has not specifically taken into account the prob- ability that paraffins with an internal quaternary carbon atom exist in the liquid state as mixtures of rotational isomers. It is significant that the symmetrical stretching vibration of the C structural unit does not appear t o show any tendency to split. This suggests that the vibrations of this unit are little affected by the configuration of the rest of the molecule (just as in the analogous tert.-butyl paraffins) and provides support for the assignment of the pair of frequencies near 1200 cm.-l and the Raman lines near 485 cm.-l t o such modes.7. Paraffins with other Structural Units (a) Para5u with One Internal Tertiary Carbon Atom.-The molecules H I I of this class possess a C structural unit C-C-C in common with those c with the isopropyl group previously discussed (Section 5). The most hopeful method of approach already used by Sheppard lo and Smithy6 is to seek analogies between the spectra of these two classes of paraffin. As will be apparent from the following discussion the interpretation of the spectra of molecules with an internal tertiary carbon atom is not easy. Serious difficulties are encountered probably because the different rotational isomers have differing frequencies even for those vibrations that Can be associated with the C unit. Experimental data for a large number of molecules are available.Table 8 lists the relevant Raman and infia-red frequencies of the liquids for those paraffins which have a single C unit and no other branched grouping. Figs. 4A and 4B show typical spectra for a paraf5.n of this class, 48 QUARTERLY REVIEWS 1043 1016 1046 - 1032 1002 1036 - - 1011 1049 1012 1041 - - 1012 1044 - TABLE 8. Churactetistic frequencies (in cm.-l) of parafins with one internal tertiary curban atorn in the liquid state * - 443 819 446 817 NI 816 436 819 NI - NI - 422 819 NI 820 - Paratan 1166 1166 1166 1166 1167 3-Methylpentane . 3-Methylhexme . 3-Methylheptane 3-Methyloctane . 3-Methylnonane . 3-Methylundecene . 1047 1 1042 1 1 1044 834 NI 1044 * 830 414 1046 * 1 842 1 NI 4-Methylheptsne . 4-uethyloctane . 4Methylnonane . - 3-Ethylpentane . I-Ethylhexme .. 3-Ethylheptane . 1167 1164 1166 1166 1166 1161 Infra-red intensity. Ramanintensity . 1042 1004 832 467 1039 1007 832 447 1043 1011 821 NI 1042 - 823 434 1060 1006 836 NI 1047 - 836 467 IH R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R IR R -_ - __- 1166 1162 1165 1168 1166 1166 1166 1166 1166 t 1166 4- No observations ,-+ I I I I I I I 1 . 1146 1046 1006 860 NI 1164 1 1047 1 - 1 848 I 433 Refs. 13 14 13 14 13 14 13 14 13 6 - - 13 14 13 14 13 - 13 14 13 14 2 14 2 14 API Nos. 6011 611 602 611 603 612 606 612 609 616 604 613 607 613 610 616 667 671 660 671 For the meaning of symbols and abbreviations see foot of Table 1. In the Raman spectra of p a r a f f i with an internal tertiary carbon atom lines are observed for most molecules near 1360 cm.-l (not listed) ; these can be assigned to the CH deformation modes.Table 8 shows that a nearly constant frequency (1156 cm.-l) appears consistently in the Raman and infra-red spectra of these molecules (col. 3). This series may be readily SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART 11 49 identified and forms the most useful means of recognising the presence of an internal tertiary carbon atom. The infra-red bands are prominent (though they show some evidence of splitting) and the corresponding Raman lines are the only ones that occur in this region. There is little doubt that this series is analogous to that occurring near 1170 cm.-l in the isopropyl paraffins. The assignment is thus dependent on the final interpretation of the 1170 cm.-l frequency in the isobutane spectra. It should be noted however that the persistence of this series even when no CH groups are directly attached to the tertiary carbon atom does seem to suggest that the relevant vibration must be predominantly skeletal in character.The frequencies shown in col. 4 near 1040 cm.-l are found in the Raman and infra-red spectra of most of the molecules listed and are only slightly affected by changes in structure. This series seems analogous to that already discussed for the paraffins having an internal quaternary carbon atom (Section 6) and presumably may also be attributed to CH rocking modes not involving the C structural unit. The infia-red bands near 1010 cm.-l (col. 5) are found in all molecules having an ethyl group in agreement with the interpretation previously given. Examination of the 850-700 cm.-l region of the Raman spectra shows that paraffins with a single internal tertiary carbon atom have at least one nearly constant frequency (col.6) that can be assigned to the sym- metrical stretching vibration of the C structural unit analogous to those already described for the isopropyl compounds. The corresponding infra- red bands are insignificant. There is evidence especially in the 3-ethyl paraffins of a second Raman series at a somewhat lower frequency (770-750 cm.-l) which can be attributed to the same vibration in a second type of isomer. Analysis in the range 770-740 cm.-l is however complicated by the presence of the CH rocking modes of the ethyl and n-propyl groups which are also present. Col. 7 shows that a series (450415 em.-l) attributable to skeletal deformation modes can be dis- tinguished in the Raman spectra ; it is presumably analogous to the series already noted in the isopropyl paraffins near 435 cm.-l.The actual fre- quencies depend on whether the molecules are 3-methyl 4-methyl or 3-ethyl paraffins but are fairly constant for a particular class of molecules. Neither of these Raman series is very helpful for identifying the presence of an internal tertiary carbon atom because in both cases a number of other lines are present in the same frequency region. Besides the series shown in Table 8 which are common to all these molecules other regularities have been separately noted in the spectra of the 3-methyl 4-methyl and 3-ethyl paraffins ; these are given in Table 10. It must however be emphasised that correlations based on a limited number of molecules are sometimes misleading and are found to be invalid when more experimental data become available.The frequencies near 1130 cm.-l of the 3-ethyl paraffins are remarkably constant. The 1080 cm.-l infra-red frequencies of the 4-methyl paraffins are presumably analogous to those already discussed for molecules with an internal quaternary carbon atom D 50 QUARTERLY REVIEWS (Section 6). The various series observed in the region 970-870 cm.-l can be correlated with those found near 955 cm.-l in the Raman spectra and 920 cm.-l in the infra-red spectra of paraffins with an isopropyl group. The weak bands near 550 cm.-l in the spectra of the 3- and 4-methyl paraffins can probably be correlated with similar series found at slightly lower frequencies for paraffins with an isopropyl group.Data are also available for some twelve molecules having an internal tertiary carbon atom together with a tert.-butyl group or an internal quaternary carbon atom. The prominent features which can be attributed to the internal tertiary carbon atom can also be recognised in the spectra of these more complex molecules. ( b ) Paraffins with Two Adjacent Tertiary Carbon Atoms.-Data are avail- able for a fair number of paraffins with two adjacent tertiary carbon atoms Le. a C structural unit C-C-C-C; this may be terminal (as in the 2 3-dimethyl parafhs) or in a more limited number of cases internaL5 It is desirable to discuss whether any of the characteristic frequencies already noted for molectiles with a single tertiary carbon atom persist in these more complex paraffins.Accordingly Table 9 lists t,he relevant data on the liquid spectra of some molecules with this C structural unit but does not include any of the more complicated examples. With regard to the interpretation of the frequencies dealt with in the following discussion it should be emphasised that in these rather complex paraffins significant disturbing effects due to the presence of mixtures of rotational isomers are to be expected as in the case of 2 3-dimethylbutane already noted.61 Col. 3 gives EL series of Raman lines near 1160 cm.-l which usually have no infra-red counterparts; in paraffins with an isopropyl group the cor- responding frequencies always appear in both types of spectra. Col. 4 on the other hand shows a series of prominent infra-red bands of nearly constant frequency (near 1125 crn.-l) ; no corresponding Raman lines can be recognised ; these bands are absent in other paraffins having a tertiary carbon atom.These two series taken together provide the most reliable guide for the identification of this C structural unit. The spectra of paraffins with this C structural unit are difficult to analyse in the range 1100-900 cm.-l. Three or four Raman lines occur sporadically and the infra-red spectra show a large number of small bands. On the whole absorption is weaker if the C unit is terminal. Most mole- cules show a band near 1070 cm.-l (col. 5) ; but this feature is not usually prominent (compare Table 10). Some molecules also have somewhat vari- able frequencies near 1040 cm.-l in both Raman and infra-red spectra. These series are evidently analogous to those discussed for other paraffins with a tertiary carbon atom and can probably also be assigned to CH rocking modes.The absorption band near 1010 cm.-l shows up regularly for molecules with an ethyl group. Below this frequency there are indica- tions of a variable series near 960 cm.-l and tt more constant series of H H I I I I c c SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 51 infra-red bands near 916 cm.-l (col. 6). Possibly they are analogous to those found near 955 and 920 cm.-l for paraffins with an isopropyl group. None of the series in the range 1100-900 cm.-l is very helpful in identifying this C structural unit. Col. 7 shows that it is possible to distinguish prominent polarised Raman lines (770-740 cm.-l) which by analogy with the paraffins previously discussed may be assigned to symmetrical skeletal vibrations.The fre- TABLE 9. Characteristic frequencies (in cm.-l) of parafins with two adjacent tertiary carbon atoms in the liquid state * Paraffin vcc or CH rocking 2 3-Dimethylbutane IR R 2 3-Dimethylpentam IR R 2 3-Dimethylhexane IR R 2 3-Dimethylhept'ane ' IR R 2 3-Dimethyloctane IR R 3 4-Dimethylhexane IR R 3 4-Dimethylheptane IR R 4 5-Dimethyloctane I IR R Infra-red intensity . Raman intensity. . - 1129 1066 921 1165 - - - 1166 1122 1075 918 1166 - - 915 - 1127 1075 919 1163 - - - - 1126 1080 918 1159 - - 926 - 1128 1080 915 4- No observations 1160 1166 - 1166 - t - W m (3) 1122 1071 - l i j l i 3 920 - - 1124 1077 910 - o observations st m m abs. abs. usually abs. vC-C - 754 740 747 756 765 7 60 760 770 736 738 765 751 - 480 477 464 459 457 NI 469 N I -+ - 467 47 1 NI NI - I - Refs API Nos ' 13 14 13 14 13 14 2 16 2 - 13 14 2 14 2 - 656 670 658 672 661 673 664 673 *For t,he meaning of symbols and abbreviations see foot of Table 1.quencies are moderately constant ; the corresponding infia-red bands are weak. This series gives a valuable indication of the presence of the C structural unit. It is also possible to pick out a fairly reliable series of Raman lines (col. 8) which lie in the region of the spectrum in which the skeletal deformation modes are found (480460 cm.-l). Smith ti has found evidence for a series of weak bands near 550 cm.-l similar to those mentioned for other molecules with a single tertiary carbon atom (pp. 41,47). Some eight other parafis with adjacent tertiary carbon atoms have been examined.It is satisfactory to find that all the frequencies listed 52 QUARTERLY REVIEWS in Table 9 may be found in their spectra ; those near 1125 and 755 remain prominent and so seem to provide a reliable methocl of identifying this C structural unit. (c) The Ethyl Group.-Many of the paraffins discussed in the preceding sections possess one or more ethyl groups and two characteristic frequencies have already been noted that can definitely be attributed to this radical &. an infra-red band near 1010 cm.-l which may possibly be assigned to a carbon-carbon stretching vibration and one near 770 em. -l attributable to a CH rocking mode. In this connection it is noteworthy tha,t Kohl- rausch,llO working some years ago on the Raman spectra of other types of molecule was able to recognise a Raman line near 1000 cm.-l as char- acteristic of the ethyl group.No definite evidence is available to suggest that the other CH deformation modes appear at frequencies differing from those discussed on p. 23 ; nor can any series (other than the 1010 cm.-l bands) be found which might be assigned to the two CH rocking modes of the ethyl group. Presumably they occur in the complicated range 1150-850 cm.-l; either they are not sufficiently prominent to be recog- nised or they coincide with the frequencies of other radicals. The deforma- tion skeletal mode of the CH,*CH,*C group must occur below 500 cm.-l but cannot be identified among the many Raman lines that most paraffins show in this region. 8. Conclusion We conclude with some general remarks on the preceding sections.First it must be emphasised that the larger the number of molecules of a particular class the more reliable are the characteristic frequencies. Secondly where there is evidence for disturbing effects due to the presence of mixtures of rotational isomers the interpretation of the characteristic frequencies must be considered tentative. In this connection there is no doubt that investigations of the solid-state spectra would prove as illuminat- ing for other molecules as they have been for the n-paraffins. Thirdly examination of the spectra of paraffins which have two or more structural units of the types described in the earlier sections suggests that in many cases each unit retains at least some of its characteristic frequencies. This is particularly apparent for the tert.-butyl group but in other cases some exceptions have been noted,5 and accidental coincidences in frequencies become more probable the more complex the molecule.Fourthly it must be admitted that the use of characteristic frequencies to recognise certain groupings is to some extent a matter of experience. This is particularly true of the region below 500 cm.-l in the R,aman spectra and for infra-red bands generally. The absorption spectra of the larger paraffins between 1350 and 800 cm.-l tend to degenerate into a somewhat puzzling area of bands of moderate intensity with complicated splitting patterns and only after comparison with those of smaller molecules of the same type do the regularities become apparent. Finally it is desirable to indicate those regions of the spectra which 110 Op.cit. 1943 p. 228. SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS. PART I1 53 TABLE 10. Characteristic frequencies (in cm.-l) of parafins in the liquid state (The frequencies of less well establishsd series are shown in parentheses. Fre- quencies marked with an asterisk are particularly prominent.) I Frequency range. crn.-l * 1470-1440 . * 1397 1370 . * 1385 1370 . * 1384 1365 . * 1380 . . . *1360-1330 . 1370 . . . 1355 . . . 1340 . . . * 1310-13UO . * 1260-1250 . * 1220-1200 . * 1215-1210 . * 1200-1190 . * 1175-1165 . * 1165-1150 . * 1160-1150 . * 1150-1140 . "-1140 . . * 1130-1120 . (1130) . . 1120-965 . (1100-1080) 1080-1065 . (- 1080). . * - 1070 . . (-1040). . * 1020-1000 . (- 960) . . 1045-1035 . * 960-950. . (950) . . . Infra-red intensity strong strong strong strong strong o bscurod medium medium medium medium strong strong strong strong strong weak m.strong strong medium strong active mediiim active medium active not promin. weak active strong active medium active Raman intensity strong absent absent absent' absent prominent absent absent absent prominent medium medium} E%-} I I medium medium medium medium prominent absent absent wea.k active absent absent prominent medium active medium active medium absent Classes of paraffin and comments CH asymm. 6 CH bending tert. -butyl { CH isolated CH symm. 6 $ ~ ~ ~ ~ ~ e t h y l CH deformabion n-paraffins hydrogen deformation CH wagging or twisting tert.-butyl internal quaternary ca.rho~~ atom isoprop yl adjacent tertiary carbon at>oms internal tertiary carbon atom isopropyl n -p araffins adjacent tertiary carbon atoms 3-ethyl ri -parafins internal quaternary carbon atom adjacent tertiary carbon atoms &methyl n-paraffins internal tertiary carbon atom int$errial quaternary and adja- cent tertiary carbon atoms :thy1 3 -methyl and adjacent tertiary kopropyl t-methyl carbon atoms Table 2 2 '1 Y 4 4 7 7 6 9 8 6 2 I) 3 54 QUARTERLY REVIEWS TABLE 10 (cont.) Frequency range cm.- (-930) . . * 930-925 . 920-915. . "-920 . (920-910) . (900-890) . (- 890) . . (- 880) . . * 870-840 . * 840-810 I * N 890 850-815. . * 835-795. . (770-750) . 770-740. *"770 . . * 750-710. . * 750-685. * -740 . . *-720 . . (- 550) . . (660-530) - 490-480. . 480-460. . 465-430. . 450-415. . *425-150. . 370-255. . 360-280. (350-305) . Infra-red intensity active medium m.strong medium active ao tive active active weak weak medium obscured weak s t,rong weak weak strong strong weak weak weak weak NI NI NI NI NI NI Raman intensity active medium medium absent active active active active prominent polarised medium prominent prominent medium absent prominent polarised prominent polarised absent absent weak weak medium medium prom ineri t polarised medium polarised medium polarised weak polarised Classes of paraffin and comments internal quaternary carbon atom tert. -butyl isopropyl adjacent tertiary carbon atoms internal quaternary carbon atom 4-methyl 3 -ethyl 3-methyl n-paraffins v c ~ internal tertiary carbon atom isopropyl vGG adjacent tertiary carbon atoms V C ~ CH rocking ethyl tert . - butyl vC4 internal quaternary carbon CH rocking n-propyl CH rocking 5 carbon chain w-c 3-ethy1 V C ~ atom vcx 3-methy1 4-methyl adjacent isoprop yl tertiary carbon atoms internal quaternary carbon adjacent tertiary carbon atom SC+ atoms sc-c ~opropyl 6c-c internal tertiary carbon atom n-paraffins Sc-c isopropyl Sc-c tert.-butyl &-( internal quaternary carbon atom 6~ Table 4 6 9 SHEPPARD AND SIMPSON SPECTRA OF HYDROCARBONS.PART I1 55 are most profitable for the purposes of structural identification. Table 10 gathers together all the characteristic frequencies for the liquid paraffins which have been listed or discussed in the preceding sections and includes also the hydrogen deformation frequencies common to all types of molecule. Examination of this table suggests that the infra-red bands in the range 1260-1100 cm.-l and the polarised Raman lines between 900 and 650 cm.-l are the features likely to be most valuable for this purpose since they are prominent and on the whole show little variation in frequency. In the intermediate region the series are less satisfactory since they often lack one or both of these desirable characteristics and some are common to several classes of paraffin. These series and other weaker frequencies are not in themselves sufficient to characterise the various structural groupings.

 

点击下载:  PDF (2931KB)



返 回