ANNUAL REPORTSON THEPROGRESS OF CHEMISTRY.GENERAL AND PHYSICAL CHEMISTRY.1. MOLECULAR STRUCTURE.Electronic Spectra.-In the period under review a considerable number ofpapers has been published on molecular spectra in the visible and ultra-violetregions.The sharpening of absorption bands and the increase in intensity a t lowtemperatures have been discussed by Keilin and Hartree who have alsomade some interesting new observations on the absorption spectrum ofliquid oxygen in the visible region. Bands in the visible region similar tothose which were primarily observed with H20 vapour have been obtainedwith D,O vapour and assigned to OD.3 The absorption of fluorine has beenre-studied by Nathans who reports that there is no absorption in the 7000-9600-~. region.In the case of bromine monofluoride absorption is observedwith rotational fine structure from 18,500 to 21,100 cm.-l and a pre-dissociation is said to occur a t 21,800 cm.-l. If the molecule dissociates intounexcited atoms a t this frequency then Do = 59.6 -+ 0.2 kcals.has calculated the relative vibrational transition probabilities for the firstfour vibrational levels in the 2C+ - 211 system of OH.Very interesting emission spectra have recently been observed withfluorocarbon vapours under the action of electric discharges.' A system ofviolet-degraded double double-headed bands of which the most prominentis at 2240 A., and a system of red-degraded bands, with marked sequenceslying in the region 1970-2210 A., have been shown to be due to a diatomicmolecule.Vibrational analysis shows that these bands have a lower state incommon, which is probably the ground state, 21T, of CF. Venkateswarlu hasassigned the bands at 2240 A . to the transition 2C - 21T. Laird, Andrews, andBarrow have shown that the bands between 2340 and 500 A . are due to theSchulerKeilin and Hartree, Nature, 1950,165, 504.Idem, ibid., p. 543.Schuler and Reinebeck, 2. Naturforsch., 1949, 5a, 560.Nathans, J . Chem. Phys., 1950,18, 1122.Brodersen and Schumacher, Anal. Asoc. Q d m . Argentina, 1950,38,52.Schuler, J . Chem. Phys., 1950, 18, 1221.Venkateswarlu, Phys. Review, 1950, 77, 676.Laird, Andrews, and Barrow, Trans. Paraday SOC., 1950, 46, 8038 GENERAL AND PHYSICAL CHEMISTRY.CF, radical. They have studied these bands in absorption and preliminarymeasurements show that this radical has a life of about 1 sec.in electrodelessdischarge. Interesting new observations have been made on the absorptionspectrum of the CS,+ molecule-ion.Q Schmitz and Schumacher lo have foundthat ClP, shows absorption beginning a t 4700 A. and rising to a maximumabout 2200 A.Edse l1 has re-examined the absorption spectrum of hydrogen peroxidevapour and obtained results which agree well with those previously reportedby Holt, McLane, and Oldenberg.12 The absorption spectrum of iodine inacetone solution has been studied and the strong band a t 363 mp. is attributedto the tri-iodideCoriolis coupling between two fundamental vibrations of the formaldehydemolecule in its ground state has been shown l4 to perturb the rotationalstructure of the band in the ultra-violet spectrum a t 3600 A.Voden and Astoin l6 have described a new light source for use in thevacuum ultra-violet.It is claimed that with copper electrodes usefulemission is obtained between 100 and 2000 A. The absorption spectra ofmethane, carbon dioxide, water, and ethylene have been investigated in thevacuum ultra-violet .I6 One outstanding requirement of vacuum ultra-violetspectroscopy is an accurate method of determining extinction coefficients.Harrison, Gaddis, and Coffin l7 have developed a technique for determiningthe molar extinction coefficients of compounds in the vapour state and claima precision of 3--5%. The method has been applied to the spectrum ofdivinyl ether.Ethylene oxide has been studied by Liu and Duncan l8 whofind two Rydberg series, one beginning at 1935 and the other a t 1382 A.Both these converge to the same limit, giving an ionization potential of10.81 ev. Two non-Rydberg transitions are observed with origins a t 1713.4and 1572.4 A . These authors assume that the Rydberg transitions arise bythe excitation of an electron from a molecular bonding orbital very similar tothe one responsible for the Rydberg series in ethylene and related compounds.Intensity measurements on the vacuum ultra-violet spectrum of ethylene inthe gaseous state, and octenes, cyclohexane, octynes, and dihydropyran inhexane solution have been made by Platt, Klevens, and Price.lg Quinol,resorcinol, and catechol have been studied in the vapour state,,O and theorigin of main bands has been discussed and some vibrational assignmentshave been made.The hydrides of sulphur, selenium, and tellurium and theLaird and Barrow, Proc. Phy8. SOC., 1950,63, A , 412.10 Schmitz and Schumacher, Anal. Asoc. Quim. Argentina, 1950, 38, 61.l 1 Edse, J . Chm. Physics, 1950,18, 244.1* Holt, McLane, and Oldenberg, ibid., 1948, 16, 255, 638.1s Benesi and Hildebrand, J . Amer. Chm. SOC., 1950,72, 2273.l4 Brand, Trans. Faraday SOC., 1950,46, 805.l5 Voden and Astoin, Nature, 1950, 166, 1029.l6 Wilkinson and Johmon, J . Chem. Phys., 1950,18, 190.l7 Harrison, Gaddia, and Coffin, ibid., p. 221.Liu and Duncan, ibid., 1949,17, 241.l9 Platt, Klevens, and Price, ibid., p.466. ao Beck, ibid., 1950,18, 1135MCDOWELL : MOLEOULAR STRUCTURE. 9methyl derivatives of hydrogen sulphide have been studied in the far ultra-violet .zOaThe assignments of the singlet-triplet emission spectrum in benzene havebeen discussed by Craig, 21 who has also given an interesting theoreticaldiscussion 22 of the perturbation of the forbidden &,-Bern transition in ben-zene by the Egt vibrational frequencies. It is shown that the 606-cm.-lvibration is about 100 times more effective than the 1596-cm.-l vibration inproducing the forbidden transition. Experimental observations agree withthese theoretical results. Schull 23 has made a very complete study of thevibrational fine structure of the 3400-A. triplet-singlet emission band inbenzene as observed by the rigid glass technique of Lewis and K a ~ h a .~ *This 3400-~. band is identified as a 1A1g-3B8n transition. The bands between2000 and 4000 A. observed in absorption in liquid ethylene have been inter-preted by Reidz5 as also being due to a singlet-triplet transition. Con-siderable discussion has recently centred on the interpretation of the ultra-violet spectra of aromatic hydrocarbons, and much progress has beenmade.26* 27 Klevens 28 has indicated that there is a complete correspondencebetween the spectra of azulenes and their six-membered ring isomers withregard to intensity, vibrational structure, and sequence of the five bandsobserved. A simple theoretical treatment which is extremely successful inpredicting maximum possible extinction coefficients of polymeric andpolycyclic benzenoid hydrocarbons has been given by Bra~de.~g Platt 30has shown that the simple L.C.A.O.molecular-orbital treatment includingoverlap is quite successful in helping one to understand the spectra of complexconjugated molecules. In an interesting series of papers a similar theoreticaltreatment has been given 31 for monosubstituted benzenes, thiophenol, com-pounds of type (C6H5),X, and various substituted derivatives of benzene, andnew experimental results have been produced which are discussed in terms ofthis theory.Numerous publications have appeared on the ultra-violet absorption oforganic compounds in solution, and a selection of those of physico-chemicalinterest is given below.Bayliss32 has discussed the effect of electrostaticpolarization of the solvent on the electronic absorption spectra in solution.An interesting consideration of the technique of absorption spectrophoto-metry has shown that fluorescence is often a cause of deviations from the200 Price, Teegan, and Walsh, Proc. Roy. SOC., 1950, A , 201, 600.21 Craig, J. Chem. Phys., 1950, 18, 236.23 Schull, J. Chem. Phys., 1949,17, 295.24 Lewis and Kasha, J. Amer. Chem. SOC., 1944,66, 2100; 1945, 61,997.25 Reid, J . Chem. Phys., 1950,18, 1299.26 Platt and Klevens, Chem. Reviews, 1947,41, 301 ; J . Chem. Phys., 1948,16, 832 ;28 Klevens, J. Chm. Phys., 1950,18, 1063.Platt, J . Chem. Phys., 1950, 18, 1168.31 Matsen, J. Amer. Chem. SOC., 1950, 72, 5243; Robertson and Matsen, ibid., pp.8p Bayliss, J .Chem. Phye., 1950,18, 292.a2 Idem, J., 1950, 59.1949, 17, 470. 27 Platt, ibid., p. 454.2B Braude, J., 1950, 379.6248, 5250, 5252, 5256 ; Robertson, Music, and Matsen, ibid., p. 526010 GENERAL AND PHYSICAL CHEMISTRY.Beer-Lambert law.33 Amongst- the compounds which have been studied insolution have been substituted aromatic nitro-compounds,34 substitutedaniline~,~5 phenols,36 diphenylalkane~,~' various ketones,38 d i t o l y l ~ , ~ ~ cyclicdienes,4O phenolic compounds,P1 azlactones,42 a n t h r a ~ e n e , ~ ~ o-substitutedanilines,& arylmethylallyl alcohols,45 hydroxydiphenylmethane~,~~ hydroxy-naphthoic a ~ i d s , ~ 7 compounds containing the C-I and sulph0xides.4~Grubb and Kistiakowsky have attempted an explanation of thermo-chromism.Fluorescence spectra are often of great assistance in the interpretation ofultra-violet spectra.Bass and Spooner 51 have photographed the fluores-cence spectra of fluorobenzene and chlorobenzene and shown that theiranalysis yields results in agreement with their known ultra-violet spectra.These molecules belong to the symmetry group Cay and it is shown that thefirst electronic transition is A, --+ B,.Raman Spectra.-The experimental development of photoelectric Ramanspectrometers 62--54 has continued and new instruments have been de-scribed.65* 56 Lord and Nielsen 57 have described a simple apparatus whichenables Raman spectra to be obtained at temperatures down to -1150".Careful distillation has been shown to produce samples in which the back-ground scattering is considerably reduced.58 Bender and Lyons 69 havedescribed a modified version of the very satisfactory technique originated byReitz 6o for the determination of depolarization factors, while Rank andss Braude, Fawcett, and Timmons, J., 1950, 1019.34 Fielding and Le FBvre, J., 1950, 2812.36 Robertson and Matsen, J.Amer. Chem. SOC., 1950,72, 1543.36 Idem, ibid., p. 1539.38 Day, Robinson, Bellis, and Till, ibid., p. 1379.39 Pickett, Groth, Duckworth, and Cunliffe, ibid., p. 44.4~ Pullmann and Berthier, Bull. SOC. chim., 1950, 81.41 Coggeshall and Glessner, J. Amer. Chem. SOC., 1950, 72, 2275.4p Schueler and Wang, ibid., p. 2220.4s Gorinda Rau and Venkataraman, Current Sci., 1950,19, 9.44 Grammaticakis, Bull.SOC. chim., 1950, 158.4 5 Braude, Fawcett, and Newmann, J., 1950, 793.4 6 Hunter, Morton, and Carpenter, J., 1950, 441.4 7 Bergmann, Hirshberg, and Pinchas, J., 1950, 2351.Durie, Iredale, and Jarvie, J., 1950, 1181.I9 Koch, J., 1950,2892; Felmel and Carrnack, J. Amer. Chern. SOC., 1950, 72, 1292.Grubb and Kistiakowsky, J. Amer. Chem. SOC., 1950, 72,419.61 Bass and Spooner, J. Opt. SOC. America, 1950, 40, 389.6* Rank, Pfister, and Coleman, ibid., 1942, 32, 390;6s Chien and Bender, J. Chem. Phys., 1947.15, 376.54 Kinell and Traynard, Actu Chem. Scand., 1948,2, 193.5 6 Miller, Long, Woodward, and Thompson, Proc. Phys. SOC., 1949,62, A , 401.66 Sushchinsker, J. Exp. Thew. Phys., U.S.S.R., 1950, 20, 403.6 7 Lord and Nielsen, J.Opt. SOC. America, 1950, 40, 653.5 * Mallory, J. Chem. Phys., 1950,18, 898.69 Bender and Lyons, ibid., p. 348.6o Reitx, 2. physikal. Chem., 1936,3$, B, 368.37 Coggeshall, ibid., p. 2836.Rank and Wiegand, ibid.,1946,36, 325MCDOWELL : MOLECULAR STRUOTURE. 11Kagarise 61 have shown that though their new apparatus appears geometric-ally unsuitable for the accurate determination of p, nevertheless, the resultsare in good agreement with those obtained by more accurate methods whenconvergence corrections are applied.Interest has recently been shown in diatomic molecules and the Ramanspectra of bromine,62 chlorine monoflu~ride,~~ and fluorine have beenobtained. Cyanogen has been re-investigated.C6 Low-frequency Ramanspectra have been observed in single crystals of benzene,66 di-iodoben~enes,~~potassium hydrogen fluoride,68 and potassium chloride.69 Hydrogen bondinghas been shown to be present in crystals of KHC03.70The aluminium hydride ion has been shown to have a tetrahedral structureby Lippincott 7 1 who studied its Raman spectrum in ethereal solution.Inaqueous solution spectra have been reported for nitrates,72 S203-- ion,73di-, tri-, and tetra-thionate ions,74 and the AuC1,- Polarizationmeasurements in the case of S,O,-- have indicated that it is probable thatthis ion has symmetry C3v. The S , 0 6 - - ion would be expected to resembleethane and have either an eclipsed structure ( D 3 h ) or a staggered configura-tion with symmetry D S d .The Raman spectrum, taken in conjunctionwith earlier infra-red data,76 indicates that the structure probably is DSd.No definite decision has been reached concerning the structure of thetrithionate ion S306--. The AuC1,- ion has been shown to have a squarestructure.76Raman-spectrographic studies of nitric acid solutions have led tointeresting and important discoveries concerning the structure of thesesolutions and have shown the existence of a new ion, viz., NO2+, in nitrating78 This work has given a great stimulus to the study of the ion invarious concentrated acid solvents and has led to a most important develop-ment in inorganic chemistry, namely, the preparation and isolation of a hostof new compounds, the nitronium salts of the general formula (NO,+)(X-)61 Rank and Kagarise, J .Opt. SOC. America, 1950, 40, 89.63 Jones, Parkinson, and Burke, J . Chem. Phye., 1950,18,236.64 Andrychuk, ibid., p. 233.65 Langseth and Meller, Acta Chem. Scand., 1950,4, 725.6 7 Korshimov and Sel’ Kin, J . Exp. Theor. Phys., U.S.S.R., 1950, 20, 292.O 8 Mathieu and Conture-Mathieu, Compt. rend., 1950, 230, 1054.6D Stekhanow, J. Exp. Theor. Phys. U.S.S.R., 1950,20, 330.70 Couture-Mathieu, Compt. r e d . , 1949,229, 1215.71 Lippincott, J . Chm. Phys., 1949, 17, 1351.72 Mathieu and Lounsbury, Compt. rend., 1949, 229, 1315.73 Gerding and Eriks, Rec. Trau. chim., 1950, 89, 659.7 6 Goulden, Maccoll, and Millen, J., 1950, 1635.76 Duval and Lecomte, Compt. rend., 1943, 217, 42.7 7 ChBdin, ibid., 1936,292, 220; Ann.Chim., 1937, 8, 243.78 Goddard, Hughes, and Ingold, J . , 1950, 2589 ; Ingold, Millen, and Pooh, J., 1960,2576; Millen, J., 1950, 2589, 2600, 2606; Ingold and Millen, J . , 1950, 2612; Gouldenand Millen, J . , 1950, 2620.Stammereich, Phy8. Review, 1950, 78, 79.Fruhling, J. Chem. Phys., 1950, 18, 1119.l4 Idem, ibid., p. 72412 GENERAL AND PHYSICAL CHEMISTRY.where X- = ClO,-, HS207-, S20,--, FSO,-, S3OlO--, NO,-, e t ~ . ~ ~ This workhas now been described in Further studies 8o of the Raman spectrumof mixtures of nitric acid and acetic anhydride indicate that in this solutionN205 exists as [NO,+][NO,-]. The Raman spectrum of solutions of sulphurtrioxide in nitric acid indicate 81 the existence of the nitronium saltCompounds of which the Raman spectra have been studied include:aromatic nitro-compounds,82 buta-1 : 3-diene,s3 hexachl~rodisilane,~~ phenyl-b ~ t e n e s , ~ ~ amino-acids,86 GeC1Br3,87 SiHBr3,88 trichloromethane deri~atives,~galdehydes,g0 fluoromethane derivatives,gl difl~oroethylene,~2 deuterio-acetylacetone, deuterioacetoacetic ester,93 ethylchlorosilanes,94 methyl-chloro~ilanes,~~ trichlor~alkylsilanes,~~ linoleic monodeuteriatedtoluenes,98 and unsaturated alcohols.99 Binary and ternary mixtures ofacetone, methyl alcohol, and carbon disulphide have been investigated, looand also the effect of increasing acid character on the OH frequency in alcoholsand phenols.lo1Recently there has been a considerableadvance in the design of infra-red prism spectrometers, and many single-and double-beam instruments have been described.lo2 I n America a numberof double-beam instruments are now on the market.A very full account ofthe Perkin-Elmer double-beam spectrometer has recently been given,lo3 andcertain problems which arise in the design of double-beam instruments have“O2+1 CHS207 - 1Infra-red Spectra.-Technique.79 Millen, J., 1950, 2589.Ch6din and Feneant, Compt. rend., 1949, 229, 115.Cerding, Steeman, and Ravallier, Rec. Trau. chim., 1950,69, 944.Richards and J. R. Nielsen, J. Opt. SOC. America, 1950, 40, 438.82 Bolorich and Vol’kenstein, Doklady Akad. Nauk., S.S.S.R., 1950,71, 1045.84 Katayama, Simanonti, Inorino, and Mizushima, J. Chem. Phys., 1950,18,506.85 Golse, Compt.rend., 1950, 230, 1762.a6 Edsall, Otros, and Rich, J. Amer. Chem. SOC., 1950, 72, 474.88 Franqois and Buisset, ibid., p. 1946.90 Harrand, Compt. rend., 1949, 229, 1217.91 Rank, Shull, and Pace, J. Chem. Phya., 1950,18, 885.** Edge11 and Byrd, ibid., p. 892.93 Shirogin and Syrkin, Doklady Akad. Nauk., S.S.S.R., 1950, 70, 1033.95 Shimanouchi, Tsuchiya, and Mikawa, ibid., p. 1306.Delwaulle, Compt. rend., 1950, 230, 1945.Zietlow, Cleveland, and Meister, J. Chem. Phys., 1950,18, 1076.Murata, Okawara, and Watase, J. Chem. Phys., 1950, 18, 1308.Gonbeau and Siebert, 2. anorg. Chem., 1950, 261, 62.Pigulevskii and Naidenova, Doklady Akad. Nauk., S.S.S.R., 1950, 72, 717.98 Smith, Choppin, and Nance, J. Amer. Chem. SOC., 1950,72, 3260.99 Malyshev and Shishkina, J.Exp. Theor. Phys., U.S.S.R., 1950,20,297.loo Joerges and Nikuradse, 2. Naturforech., 1950, 5a, 25.101 Batuev, Merhcheryakov, and Matveeva, J. Exp. Thew. Phys., U.S.S.R., 1950, 20,318.102 Wright and IIerscher, J. Opt. Soc. America, 1047, 37, 211; Baird, O’Bryan,Ogden, and Lee, ibid., p. 754; Brownlie and Cumming, Nature, 1948, 164, 105; Elliott,Ambrose, and Temple, J. Sci. Inetr., 1950, 27, 21.103 ‘White and Liston, J. Opt. SOC. America, 1950, 40, 29, 36, 93MCDOWELL : MOLECULAR STRUCTURE. 13been discussed by Kivenson.lOQ Other technical developments which maybe noted are the suggested use of a carbon arc as an infra-red source,1o6 andthe continued development of photoconductive detectors. The variationof the long-wave limit of lead sulphide, telluride, and selenide photoconduc-tive cells has been investigated by MOSS,^^^ and Fellqett lo7 has discussed thetheory of the ultimate sensitivity of various radiation detectors.Daly andSutherland lo* have considered how the performance of a spectrometer islimited by the detector characteristics. Watts log has shown how a verysimple device increases the sensitivity of a photoconductive cell. A newgrating spectrometer including a thermopile as a photoconductive cell hasbeen described.Davies ll1 has observed that a synthetic silica prism in the Grubb-Parsons single- beam instrument gives a resolution which is comparable withthat observed with a grating spectrometer.Theoretical.-A rigorous treatment of the intensities of the vibrational-rotational absorption spectra of diatomic molecules has been given byCrawford and Dinsmore.l12 With regard to the theory of molecular vibra-tions a general solution for the secular equation has been given byTorkington 113 which is applicable to any molecule.The same authorhas made a normal co-ordinate analysis of the planar vibrations of varioussubstituted ethylenes 114 and also considered the problem of calculatingvalence force displacement co-ordinates for systems in which the anglesdeviate from the ideal,l16 the calculation of moments of inertia of moleculeswith internal rotation,l16 and the cubic secular equation for molecular~ i b r a t i 0 n s . l ~ ~W. J. Taylor 118 has considered the general form of the force constantmatrix for harmonic vibrations with interesting results.The vibrational-rotational energies of planar symmetrical X,Y,X2 molecules 119 and angularinteraction in pyramidal molecules such as phosphorus trifluoride andarsenic trifluoride have been considered.120 Pace 121 has computed the forceconstants of the fluoromethanes. The very extensive theoretical discussion104 Kivenson, J . Opt. SOC. Amrica, 1950, 40, 113.l o 5 Rupert and Strong, ibid., p. 455.lo6 Moss, Proc. Phys. Soc., 1949, 62, B, 741.'0' Fellqett, J . Opt. SOC. America, 1949, 39, 920.lo8 Daly and Sutherland, Proc. Phys. SOC., 1949, 62, A , 205.l U 9 Watts, ibid.. p. 486.l 1 0 Thompson and Miller, Proc. Roy. SOC., 1950, A , 200, 1.1 1 1 Davies, J . Chem. Phys., 1950,18,398.112 Crawford and Dinsmore, ibid., p.983.113 Torkington, ibid., 1949, 17, 1026; Trans. Faraday SOC., 1050, 46, 27.114 Ibid., p. 1279.l l 8 Ibid., p. 407.118 W. J. Taylor, ibid., p. 1301.l l B Herman and Schaffer, ibid., p. 1207.l * O Burnell and Duchesne, ibid., p. 1300; J . Phys. Radiol., 1950,121 Pace, J . Chem.. Phys., 1950,18, 881.115 Ibid., 1950,18, 93.11' Ibid., p. 773.119.1114 GENERAL AND PHYSICAL CHEMISTRY.of the vibrations of molecules recently given by Linnett and his collaborators 122has been extended by a consideration of the force constants of the non-metallic hydrides of Groups IV, V, VI, and VII.Attention has been drawn by Thomas lZ3 to inconsistencies in variousproposed structures for the carboxyl group, and a new assignment of fre-quencies has been submitted.Sheline has considered the problem of theeffective mass of the methyl group in Group-Iv tetramethyl derivatives.Experimental Observations.--Cole and Thompson 126 have measured theintensities of absorption bands due to the bending vibrations of substitutedbenzenes, and have calculated the dipole moment of the C-H bond.Extinction coefficients of near infra-red bands of C-H, N-H, C-C, (3x0, CzN,C-Cl bonds have been recorded.126 Francis lZ7 has given results ofmeasurements of the intensities of absorption bands for aliphatic hydro-carbons. Kaplan 128 has calculated the intensities of the 15-p. carbon dioxideband.Since infra-red absorption is only possible when there is a change in thedipole moment during the transition, homonuclear diatomic molecules shouldnot exhibit infra-red vibrational spectra.Quadrupole absorption has howeverbeen observed 1z9 for hydrogen, and a complete analysis of the quadrupolevibrational-rotational spectrum of this molecule has been given by Herz-berg.130 Using the long-path technique it has also been possible to obtainthe vibrational-rotational absorption spectrum of HD.131 Smith, Keller,and Johnson 132 have also observed infra-red absorption in liquid oxygensimilar to that previously reported by Crawford, Welsh, and Locke 133 fornitrogen, oxygen, and hydrogen at high pressures. They have also observedabsorption in liquid nitrogen. Similar observations in liquid oxygen andnitrogen have been made by Oxholm and Williams 134 and by Van Asseltand Williams.13sOzone has been re-studied by Gutowsky and Petersen 136 who state thatthe new observations do not allow a decision to be made between the acute-angled 13' and the obtuse-angled m0de1.l~~ Badger and Wilson 139 howeverlZ3 Heath and Linnett, Trans.Faraday SOC., 1948, 44, 556, 873, 878, 884; Linnettand Wheatley, ibid., 1949, 45, 33, 39; Heath, Linnett, and Wheatley, ibid., 1950, 46,137.lZ3 Thomas, J . Chern. Phys., 1950,18,76.lZ6 Cole a.nd Thompson, Trans. Faraday SOC., 1950, 46, 103.12a Suhrmann, Angew. Chem., 1950,62,507.lZ7 Francis, J . Chem. Phys., 1950, 18, 861.lZ9 Herzberg, Nature, 1949, 163, 170.Idem, Canad. J . Res., 1950, 28, A , 144.la2 Smith, Keller, and Johnson, Phys. Review, 1950,79, 728.133 Crawford, Welsh, and Locke, ibid., 1949, 75, 1607.lS4 Oxholm and Williams, ibid., 1949, 76, 151.136 Van Asselt and Williams, ibid., 1950, 79, 1016.13* Gutowsky and Petersen, J .Chem. Phys., 1950,18, 564.lS7 Adel and Dennison, ibid., 1946,14, 379.lBB M. K. Wilson and Badger, ibid., 1948,16, 741.1~ Badger and M. K. Wilson, ibid., 1950, 18, 998.ler Sheline, ibid., p. 602.128 Kaplan, ibid., p. 186.lS1 Idem, Naure, 1960, 166, 562MCDOWELL : MOLECULAR STRUCTURE. 16point out that the acute-angled model fails to account for the band observedat 1110 cm.-l. Interesting and important new observations on this problemhave been made by Wilson and Ogg 140 who studied pressure broadeningeffects in the infra-red spectrum. Using pure ozone they observed a Qbranch in the 1043-cm.-l band.An acute-angled model would require twoof the fundamental vibrations to have Q branches whereas the obtuse-angled model would only require one of the fundamentals to have a Q branch.It has previously been shown 137 that the 705-cm.-l vibration lacks a Q branch,so Wilson and Ogg conclude that their observations support the obtuae-angled model.Other simple inorganic molecules which have recently been investigatedare F20,141 S8,142 S2C12,142 and P4,142 16N14N0 and 14N14N0,143 H 2 ’ S 145sulphur monoxide,146 H 2 0 2, 14’* 148 and D202,147 COC1F,149 C1F,160 andCICN.lS1 Perhaps the most interesting observation has been the discoveryby Jones 146 that sulphur monoxide as prepared by Schank’s method cannot beSO but is probably S,02.Iodine pentafluoride has beenshown by Lord, Lynch,Schumb, and Slowinski 152 to have a tetragonal pyramidal structure (C4v)and iodine heptafluoride a pentagonal bipyramidal structure (Dbh). Shelineand Pitzer 153 have studied the infra-red spectra of Fe(CO), and Fe,(CO),.High-resolution measurements ls4 with diborane are consistent with theconclusion that this molecule has the bridge structure with symmetry V , inagreement with Price’s earlier work.156 In passing one may note that Webb,New, and Pitzer 166 have re-studied the Raman spectrum of diborane andalso investigated the infra-red spectrum of deuteriodiborane. Their resultsare in agreement with the bridge structure. Price 157 has studied the infra-redspectra of aluminium, lithium and sodium borohydrides and proposes abridge structure for aluminium borohydride and a tetrahedral structure forthe other two.High-resolution spectra obtained with polarized infra-red radiationsuggest that the urea molecule is planar in the crystalline state.158The vibrational-rotational bands of allene,15, vinyl chloride and fluoride,l 4 0 M. K.Wilson and Ogg, J . Ghem. Phys., 1950,18, 766.141 H. J. Bernstein and Powling, ibid., p. 685.143 Richardson and E. B. Wilson, jun., ibid., p. 694.144 Allen, Cross, and M. K. Wilson, ibid., p. 691.us Noble and Nielsen, ibid., p. 667.147 GiguBre, ibid., p. 88.149 E. A. Jones and Burke, ibid., p. 1308.l 6 0 E. A. Jones, Parkinson, and Burke, ibid., p. 235.151 Richardson and E.B. Wilson, jun., ibid., p. 155.lSa Lord, Lynch, Schumb, and Slowinski, J . Amer. Chern. Soc., 1950, 72, 522.163 Sheline and Pitzer, ibid., p. 1107.lS4 Anderson and Badger, J , Chem. Phys., 1950, 18, 698.156 ?rice, ibz’d., 1948,16, 894,lo@ Webb, New, and Pitzer, ibid,, 1949,17, 1007.157 Price, ibid., pt 1044.IL8 Waldron and Badger, ibid., 1950, 18, 566.lsg Miller and Thompson, Proc. Roy, SOC., 1949, A , 200, 1.142 Idem, ibid., p. 1018.146 A. V. Jones, ibid., p. 1263.148 R. C. Taylor, ibid., p. 89816 OEINERAL AND PHYSIOAL CHEMISTRY.vinylidene fluoride, glyoxal, 160 and methanethiol have been investigatedvery completely. In the case of allene, a type of Coriolis coupling originallypredicted by Nielsen 162 has been observed. cycEoPropane 163 has beenstudied by the long absorbing path technique.164Saksena and Narain 1e5have recently shown a preference for the D4d structure but it has beenpointed out 166 that this is certainly in conflict with much of the experi-mental evidence.167* lB8 Lippincott, Lord, and McDonald 166 believe thatthe D, structure is probably correct, i.e., the " crown " structure.Further studies 169* 170 have led to a new assignment of frequencies inethylene, but Torkington 171 has pointed out several difficulties which thisoccasions.etc. etcThermodynamically the osmotic pressure may be quite generallyexpressed in virial form (Zoc.cit.), and all reported measurements can befitted to this equation using not more than three terms. In many instances,34 Zirnm, J .Chem. Phys., 1948, 18, 1093; Blaker, Badger, and Gilmar, J . Phys.Colloid Chem., 1949, 53, 794; Hadow, Sheffer, and Hyde, Canad. J . Res., 1949, 27, B,791; Schulz and Harborth, Makromol. Chem., 1948, 2, 187; Jullander, Acta Chem.Scand., 1949, 3, 1359.Doty and Steiner34a Doty and Steiner, J . Chem. Phys., 1950, 18, 1211.35 Stuart, Makromol. Chem., 1949, 3, 176; Chem. Weekblad, 1949, 46, 293; Oster,Rec. Trav. chim., 1949,68, 1123; Rousset and Lochet, Rev. Sci., 1948,86,291; Hermans,Plastica, 1950, 3, 187, 222." Sands and Johnson, Ind. Eng. Chem. Anal., 1947, 19, 261 ; Sirianni, Wise, andMcIntosh, Canad. J . Res., 1947,25, B, 301 ; Browning and Ferry, J . Chem. Phys., 1949,17,1107 ; Bawn, Freeman, and Kamaliddin, Trans. Paraday SOC., 1950,47,862 ; Platzek,Chem.Weekblad, 1950,46, 193.37 Masson and Melville, J . Polymer Sci., 1949,4, 323, 337.38 Enoksson, ibid., 1948, 3, 31492 GENERAL AND PHYSICAL CHEMISTRY.linear relationships have been found between x/c, and c,,~~ and this impliesthat A , = 0 Unfortunately, contrary results have been reported by differentworkers for the same polymer in the same solvent, the X / C , against c, plot.being linear in some cases whilst in others it shows a very marked curvature.40Whether this is due to branching or to a different internal architecture of themolecule is not clear. Recent osmotic measurements include poly(viny1acetate)41 and polystyrene42 in a wide range of solvents and at differenttemperatures, polythene in xylene at 72.7" and 9 1 ~ 6 " , ~ ~ poly( isobutyl vinylether) in toluene,44 poly( butyl acrylate) in and poly(methy1methacrylate) in benzene.46The determination of the virial coefficient, A,, is important, since it is ameasure of the mutuaI interaction of the molecules in solution.When A ,is large the segments, through their excluded volumes, repel one another sothat the molecule is swollen, but when A , is made smaller the moleculeshrinks. When A , = 0 the average repulsive and attractive forces balanceand the extension of the molecule will be little affected by the ~ o l v e n t . ~ ~ * Contrary to previously held opinion recent theoretical and experimentalwork shows that A , depends strikingly on molecular ~ e i g h t . l l * ~ ~ * 48 Zimm l1observed a threefold variation of A , for polystyrene in dichloroethane overthe molecular weight range 1,78,3,000-23,700. A slight decrease in A , withmolecular weight was observed for poly(viny1 acetate) in ethyl methylketone.41 The dependence of A , and A , on molecular weight was measuredin toluene by Bawn et aZ.,36 who observe that the plot of A , against 1/Mwas linear although there was considerable scatter of the points at highmolecular weight, and the relationship may need modification when moreaccurate data is available.As would be expected the value of A , decreases inpoor solvents, and this is found in all the cases so far Forpolystyrene, A , has a negative temperature coefficient in good solvents anda positive coefficient in poor solvents.51 Both Zimm l1 and Iiunst 50observed a close correlation between the root mean square length, R, and A ,for single fractions of polystyrene in a range of solvents.Similar, althoughless extensive, results are observed with polyi~obutene.~l Zimm noted that38 Goldberg, Hohenstein, and Mark, J. Polyirier Sci., 1947, 2, 503; Melville and40 Cf. refs. 39 and 42.41 Browning and Ferry, see ref. 36.4 2 Bawn, Freeman, and Kamaliddin, see ref. 36; Breitenbach and Frank, Monatsh.,1949, 79, 445; 1950, 81, 455; Schick, Doty, and Zimm, J. Amer. Chem. SOC., 1950,72,630.Valentine, Trans. Faraday SOC., 1950,46, 210.43 Muthana and Mark, Rec. Trav. chim., 1949, 68, 758.44 Idem, ibid., p. 754.4 5 Bickel and Melville, Trans. Faraday SOC., 1949, 45. 10.19.4 6 Mackay and Melville, ibid., p.323.4 7 McMillan and Mayer, J. Chem. Php., 1945,13, 276.4 8 Benoit, Compt. rend., 1950, 230, 2024.49 Refs. 36, 41, 48.Kunst, Rec. Trav. chim., 1950, 69, 125; h'alure, 1949, 164, 535.51 See refs. 11 and 12BAU" : SOLUTIONS OF HIGH POLYMERS. 93the osmotic effect with polystyrene was able to change the mean extensionof the molecule by a factor of two.Viscosity.-The considerable interest in the experimental and theoreticalinvestigations of the viscosity of solutions of macromolecules arises from theinformation which may be derived about the size and weight of the molecules.The intensity with which this work is pursued may be judged from the factthat many of the new ideas referred to below arose simultaneously in differentplaces.It was early realised that the Einstein model of the compact spherewas not consistent with our ideas of chemical structure, and later calculationswith asymmetric molecules permitted the evaluation of the dimensions ofmolecules of simple shape, such as ellipsoids or rods, once the molecularweight was known. The extensions of these ideas to chain-type polymersshowed the intimate connection between viscosity and molecular weight , andfrom numerous measurements the familiar rule between the intrinsic viscosity,[ r ) ] , and molecular weight, M , viz. [q] = KM", where the exponent a assumesvalues between 0-60 and 1.00, was deduced. The magnitudes of K and ahave been determined for a large number of polymer-solvent systems, andnew work includes measurements of polyi~obutene,~~ p ~ l y s t y r e n e , ~ ~ poly-b~tadiene,~* and cellulose nitrate 55 in a wide range of pure and mixedsolvents.A single relationship has been found to exist between a and K forpolystyrene in pure and mixed s0lvents.~6Striking advances have recently been made in the theoretical under-standing of the relationship between viscosit,y and molecular weight, and ithas been shown that the modified Staudinger law may be derived for theflexible-coil model of a polymer chain. If the molecule is considered as anecklace in which the flow around each bead is not modified by the presenceof the other ones, then the intrinsic viscosity is proportional to R2, where Ris the distance between the chain ends.57 Combination of this result withthe theoretical relationship for R leads to Staudinger's equation.57The assumption that negligible interaction occurs between individualsegments is not true for a long-chain coiled molecule, and recent theories 57include the effects of hydrodynamic interaction. Brinkman and Deb ye andBueche 58 consider the polymer particle as a porous sphere in which thereare 2 beads distributed with uniform density, and evaluate the disturbanceof flow through the particle produced by its independent segments.These52 Baldwin, J . Anter. Cheni. (soc., 1950, 72, 1833; Fox and Flory, ref. 24.53 Bawn, Freeman, and Kamaliddin, Trans. Faraday SOC., 1950, 46, 11 07 ; Schulz,Makromol. Chem., 1949, 3, 146; Breitenbach, ref. 42; Vallet, Rec.Trac. chim., 1950, 69,325.54 Johnson and Wolfangel, Ind. Eng. Chem., 1949, 41, 1580.5 5 Zapf, Makromol. Chem., 1949, 3, 164.Bawn, Grimley, and Wajid, Trans. Faraday SOC., 1950, 46, 1112.Debye, J . Chem. Phys., 1946,14, 636; Hermans, Rec. Trav. chim., 1944, 63, 219;J . Polymer Sci., 1946, 1, 233 ; Kramers, J . Chem. Phys., 1946, 14, 415.5 8 Brinkman, A p p . Sci. Res., A , 1947, 1, 27; Physica, 1947, 13, 565; Debye andBueche, J . Chem. Phys., 1948,16,573 ; Kirkwood and Riseman, ibid., p. 565 ; Kuhn andKuhn, Helr. Chim. Acta, 1945, 28, 98, 1533; 1946, 29, 609. 830; J . Colloid Sci.. 1948,3, 1 1 : J . Polymer S c i . , 1950, 5, 51994 GENERAL AND PHYSICAL UHEMISTRY.segments interfere with one another hydrodynamically with the result thatthe inner parts of the polymer molecule are shielded from the solvent byoutlying portions.This shielding effect is greater for large molecules thanfor small ones. Kuhn and Kuhn and Kirkwood and Riseman s8 employed adifferent mathematical approach, using the statistics of the polymer chainwith hindered rotation rather than the hydrodynamically equivalent sphere.Both theories predict that the exponent o! should decrease from unity forlow-molecular-weight polymers to 0.5 for very high molecular weights andallow of the calculation of R from the measured [q] once the intrinsicviscosity/molecular weight relationship has been determined. The theoryhas also been worked out for rods made up of monomer molecules separatedby rigid bonds.59It has been repeatedly reported that chain polymers exhibit higherintrinsic viscosities in good solvents than in poor solvents, and these effectshave been interpreted in terms of the nature of the interaction betweensolvent and polymer molecules. The above theory may be used to determinethe influence of solvent on the molecular dimensions, and results have beenreported on polystyrene in pure and mixed solvents by Outer, Carr, andZimm 6o and by Bawn, Grimley, and Wajid.S6 The former authors, and alsoK u n ~ t , ~ ' show that the variation of R with solvent is similar to that deter-mined by the light-scattering method (Zoc. cit.).A close correlation between theory and experiment has also been noted byPeterlin 62 for polystyrene and amylose, but with the more rigid cellulosederivatives the theory is not so satisfactory.Both the rotary-diffusionconstant and the intrinsic viscosity depend on the same parameter describingthe shape of the particle and, by a combination of these measurements, offeran additional method of determining molecular weights.58* 61The numerous relationships which have been proposed to represent thedependence of viscosity on concentration 63 may be expressed in the expandedformthe coefficients a2 and a3 are empirically related by the equations a2 =k1[qI2 and a3 = k2[ql3. This equation is obeyed for many systems with k,in the range 0-3-0.6 for dilute solutions. By using certain reasonableapproximations and including hydrodynamic interaction and aggregation,Simha has shown that the above equations may be derived theoreticallyand that the calculated vaIues of k, (0.77 for coils) are of the right order ofmagnitude.Furthermore, the parameter k, is shown to depend in acharacteristic manner on the shape of the particle as determined by flexible?.P/C = [?I + a2c + a3c2;Kirkwood and Riseman, J . Chem. Phys., 1950,18, 512.Outer, Carr, and Zimm, ibid., 1950,18, 836.61 Newman, Riseman, and Eirich, Rec. Truv. chim., 1949,68,921.62 Peterlin, J. Polymer Sci., 1950, 5 , 473.6s Alfrey, " Mechanical behaviour of High Polymers," Interscience, 1948, pp. 459-64 Simha, J. Res. Not. Bur. Stand.. 1949, 42, 409; ,J. ColloidSci., 1950, 5, 386,461BAWN : SOLUTIONS OF HIGH POLYMERS. 95molecules by the solvent environment and the molecular weight.Similarcalculations by Saito 65 lead to the same relationship, with k, = 0.43 forchain molecules, in very close agreement with experiment. At present thetheories do not specifically include polymer-solvent interaction.Fractions of polystyrene prepared a t different temperatures obey theseequations with k, = 0.38 (toluene),66 but when the polymer is made frommonomer containing small quantities of divinylbenzene the resulting k,values are changed t o 0-50--0-84. This is ascribed to cross linkage orbranching. I n ethyl methyl ketone k, has been found to increase from 0.42to 0.63 as the molecular weight increases, and this may be due to progressivecoiling in the bad solvent.Polgelectrowes.-The ionisation of macroelectrolytes containing largenumbers of ionisable groups of different types has attracted much attentionboth experimentally and theoretically.The difference of the titrationbehaviour of a polymeric acid from that of monobasic acid is due to the workexpended in removal of the hydrogen ion from the field of the ionised groups,as well as the work performed by electrostatic forces on stretching the polymermolecule. Several calculations of the field effect have been made and usedto develop theories of titration and dissociation of polymeric acids.67Overbeek, by treating the polymer molecule as a sphere containing a con-tinuous charge, expresses the change of dissociation constant as a functionof the radius of the molecular coil and the charge on the particle. Measure-ments 6* with solutions of poly(methacry1ic acid) agree with theory for lowdegrees of dissociation but for higher extents of dissociation the treatmentof the polymer coil as a sphere is no longer satisfactory.Expressions havebeen developed for the electrical free energy of a statistically coiled electricallycharged chain molecule.67* 69 I n these calculations the end-to-end length Rof the polymer molecule is used as a characteristic parameter €or the shapeand the expansion of the molecule caused by electric charges. The effect ofconcentration and added salts thereon may be expressed as a function of R.The configuration of ionised polymer molecules has also been discussed byGuinard et aL70In a macromolecule of the poly(viny1pyridine) type, in the presence of,for example, an alkyl halide, every carbon atom of the long chain carries apositive pyridinium ion; the charges on the polycation cannot diffuse apartany farther than that corresponding to the maximum extension of the coil.Therefore, in a very dilute solution where the molecules are separate, thereexists a small volume of high positive-charge density localised in the poly-6 5 Saito, J .Phy. SOC., Japan, 1950, 5, 4.6 6 Walker and Winkler, Canad. J . Res., 1950, 28, ;B, 298.6 7 Katchalsky and Gillis, Rec. Trav. chim., 1949, 68, 879; Kuhn, Kunzle, andKatchalsky, Helv. Chim. Acta, 1948, 31, 1994; J. Polymer Sci., 1950, 5 , 283; Overbeekand Hermans, Rec. Trav. chim., 1948, 67, 761 ; Overbeek, Bull. SOC. chim. Belg., 1948,57, 252.Arnold and Overbeek, Rec.TTUV. chitn., 1950,69, 192.Guinard. Boyer, Kawenoki. Dobry, and Tonnelat. Compt. rend., 1949, 229, 143.** Kunzle, ibid., 1949, 68, 69996 GENERAL AND PHYSICAL CHEMISTRY.cation, and we should expect a large number of anions to accompany thepolymeric ion and exist within the molecular coil. Outside, in the body ofthe solution, only anions will be found. When the coils overlap the solutionwill resemble a concentrated solution of a one-one electrolyte. This view-point has been confirmed by Edelson and Fuoss 71 from measurements of theconductivities and viscosities of poly( N-n- butyl-4-vinylpyridinium bromide)and poly(sodium acrylate) at concentrations up to 0 . 3 ~ . These authors alsoshow that, although aqueous solutions of these two polymers at the sameequivalent concentration have approximately the same conductivity, therelative viscosity of the polyacrylate was much greater than that of thepolybromide.The difference is ascribed to cross linking 73 through hydrogenbonds with water molecules and not to differences in molecular weight. Theconsiderable evidence that in aqueous mixtures of poly(acry1ic acid) andsodium hydroxide the sodium and polyacrylate ions are associated to animportant extent has been established by transference measurements,radio-sodium being used as a tracer.74from theoretical calculations of the dependence of pHand equivalent conductivities as a function of concentration, conclude thatthe conductivity of a polyelectrolyte is due almost entirely to hydrogen ions.Theoretical expressions for the intrinsic viscosity of polyelectrolytes havebeen derived by Kuhn 67*75 and Hermans and O ~ e r b e e k , ~ ~ and these havebeen used to evaluate the dimension of the polymer coil in a mannersimilar to that described for un-ionisable polymer .The viscosity behaviourof strong polyelectrolytes has been the subject of research by Puoss and hiscollaborator^.^^ Empirical functions have been deduced for the viscosity ofsolutions which besides containing a factor dependent on molecular weight,as with ordinary polymers, includes terms determined by electrostaticinteractions.The effect of salts on the viscosity of poly(acry1ic acid) solutions has beenshown to be in agreement with Smoluchowski’s electrical theory of viscosityat low concentrations and that of the folding chain at higher concentrationsof added salt.77 The theory of light scattering from solutions and chargedmolecules has been worked out by Hermans 78 and by Doty and Steiner.79The latter authors report measurements on various polyelectrolytes, and havefound that there is a deficiency of light in the forward direction instead of anexcess.The investigation of the properties and behaviour of polyelectrolytes as aWall andi 1 Edelson and FUOSS, J . Amer. Chem. SOC., 1950, 72, 306.72 Wall and Butts, J. Chem. Phys., 1949,17, 1330.73 Edelson and FUOSS, J . Amer. Chem. SOC., 1950,72, 1838.74 Huizenga, Grieger, and Wall, ibid., p. 263.7 5 Kuhn, Kunzle, and Katchalsky, Bull. SOC. chim. Belg., 1948, 57, 421.i0 Fuoss, J . Polymer Sci., 1948, 3, 603; 1949, 4, 47; Fuoss and Strauss, Ann. N . Y .7 7 Markoritz and Kimball, J . Colloid Sci., 1950, 5, 115.7 8 Hermans, Rec. Trav. chim., 1949,68, 859.Acad. Sci., 1949, 51, 836.Doty and Steiner, J . Chem. Phys., 1949, 17, 743BAWN : SOLUTIONS OF HIGH POLYMERS. 97model substance for understanding the more complicated behaviour ofbiological systems has been stressed recently. Methacrylic acid polymerscontaining o.07y0 dipinylbenzene as a cross-linking agent absorb up to 326times their weight 80 of water, giving jellies which are weak although stable inform. Jellies swollen 81 -fold by water shrink to 6-fold in 0-Oh-hydrochloricacid, whereas in water containing 7 yo equivalent weight of potassium hydrox-ide a 202-fold swelling occurred.The molecular expansion and contraction of ionisable polymers may bemade evident on a macroscopic scale as shown by Kuhn et u Z . , ~ ~ who foundthat filaments of poly(acry1ic acid), bearing a load, expanded and contractedreversibly in the presence of alkali or acid, respectively. The degree ofexpansion or contraction was determined by the extent of cross linking ofthe polymer.