-AAA._:‘A A A- “kn-NA .‘AuuAA .A “H. “A ..,.. 7.90“, Y.‘ IJ-‘A y... r . J... f.,..;...A - 'U’. -A'A..A/--¢.i A A, t . , A ;_ I A . s , , I 5 -.‘ AA “:5; 25-3.. 5.5 ., w, ,.; .>1-A¢¢-.II' I, II 4 ." . I'M”: A ‘:;.'.'A III? A I’A A-IA.IAAIA.,,,A,,, AHA... .... ‘i‘. H AIVA‘ ..',.. An” I] ..,A_,-. AA. A I, ”in IAA.’;A1:. ( I A A y f‘r-“I‘g-wb. Inmib/éi-r any?» .,,, 15-4 I; ”AWL. ., . ' I. ‘ .1; .IA VA. A' A; "-1”: "V’YII‘IW :2“ 5; ILII:J>)'—’ . ‘IAIA A 3.. ...;.’.’!‘..‘ l I .A I N 1:; .I-AA..,..,,,,:;‘H‘ 5') “—uA... A .A.A.. .. >1» .. A . 5.1.”, 35., I». w I L t I: I J ILA”- ..,. I... .‘vwn .' ‘3'] .AV LIA I.-.aA ..,, ..-.‘I—'.....,',A . ”-1.1;- a’: A “>31. . ‘1“ A..-.»:.£..’L" " .IML £2»: 35 A 'I - I. "I! " ' ; DJ 1 I. . . . .. . , . .4“ h v r‘" ‘L/LR‘I‘JI 4.1.55,» In...» a - u A. w. W ‘ . A. AA AA», '"I-A-L»;21A-.; A1. . ‘wm'ym. “31,6: '1, I I- A A’mfi‘b If»: I In.“ H h '4; :;I my "l‘lfi'ilfilw‘ .r .. .... .1" A» . is“ -;rA.IA;A WA; .3:..‘..‘,. AA ‘-/ f“: “5-1,”... 'I_ wu- {W ’F»"‘" w'.’. A“. II-MIIIn-Iiuuu- ‘ .. Ll 'rrr 3.103;»: ; 31.1w ,, . 1,2 . a; «4- Hair?“ «an "I; .— ’ r! ”’I ’,, «633” “m m“ ‘- 159*.1'; "I‘;( ’ ‘3'».332‘, iii? U};" :“L’: W “:3. ’nm»I;:ui»’L-3£¢~).J‘ Wyafi—a no :34 .A ""51: ‘,‘L ' “+251”. 8.333123? '2. vy‘rryvrw Ar ~I a... :5.“- ~.L .‘.n. A-mluunnh-wm “.5... 1"?” H AA . . WI» Ullvi‘l wA >1 . 4 u a IAILI~AAI«_ m ) f w o’ I ":39? 'A' . :2. ””2"" ' v ‘ ’ ‘ yd? ""1“ A. . 7:) RAVI}? AI; .. .j . I d ILL ’1..th . I. .7 .. pm | n... ’W. A... I-,I. “1‘ A. ..A 4' {4, )‘v Ami-Alia, -A ,A I‘l’fii. , .31 . , 'r‘““ A’.A. , ,...A .. . K . 1... A .'-‘"’H {In ""1‘ ”lid-2;, I“ I I’~’.‘ 313‘) I .A‘ . )rlrfi :v-nr' 1’! “A I- .5; w M 123m mr ‘2' ,i’, 1":‘r 3" ’ :IIIJLée'EEL’K' ... a: ’ -A "Wu," I ....A.A..- 'Qiirjrr: 557.73; ,. “1*;22 ’13;- ImM’J‘II-prh" I’L‘é). um- . A... 2-,". . I; _ y"... I...I""'A‘ . l - ' '511‘1'31}. y “Y “" w .A ”.5 7mm 72:3: - a, . pm”. M" I, “Yr" ”$4.5 (in! .A AA... u. Imrg: M A, .“P ;.‘l y J., .Irnf’ .- A. o Mk : ;"‘~ man-£23., “ ”WW- y": m- at,” ga- «$5.. .. . , t A . . . .A #W'Jfi’rp‘rr’lrn 44:7,»; ' {- ”3;; g . In“ "}~ I mu > (3/1!“ ”Aft. .at'b'of: I5» I 932:...“ 4.1.6."; Muzak, ......L'1 mm. ”"4“... 1’ 71“" 'u- ”A. H 'h . M'f*.->~’ row a»; ’5'? W53.1¢ ”, A Inga-«gr 0’ ".‘J " ‘ A PM, if? h";- :55? ’figfii‘? - .2. :.-§;M”A.« fit: _..,r In “up...” II”... A. A!“ ‘1‘. A ‘ yL’._»‘«'aAu' IIA . (.1 I.“ u- ILA: ’1' - '17,. "f’lfithK-M, Jan- :2 Igmtw . I . r» 5- .. I- W. I? .I-I... " " 'wrM‘ ”I .... IEJFIE§NP,.~1..;,.:, _>_~r “A v0 by,“ I ,7. no: 1”. 21.5? a; .. fwfinv... Lia; 33.35552? .. ... "E. #22:?!“ III-2:75" w: r "*3":h‘;J-L‘:'L" » IwI- Yb 1"” H’ ”213541“ n M..- .4. Inuit.» ..,.,.,. «m, «Iq- AW‘N- :21. II own. .,'.'.:_.., ._ ngfifi‘uqrjzr H véhnz-TW ”'* *' '1. 5+ II->1__.A—I.II_.I:I. HMI ‘2' we; mp4 I”- "~31" .75.; ”' w; 9::- 1-1 M wh‘flh‘Ih-f‘ZU‘ £32.. :5: ~54» $13"'w3'"f~ ”-3" .. ' “an ~I¢Nr "fax“: ‘9'” A4,...‘ ~m- :vn-m ”'3‘ Fr!)- Mr!” ma 1-: I... ”#9 )2 “int-rm '12."; 31...”: "ragga ..A “'22“; ' POM-Ir “0'31 ' - “kc-”24:" 1: Arm-p» ”5": an: .. .1." H A , Aw} ’1»..- ‘ I W4 ”4......" Y'f‘Mww "A. .W... ’ 9595.2; «mm, Fun I F)‘:. r u,“ «r >I3,500 A, produces HCCN and N2. Not much is known re- garding the vibrational spectroscopy of diazo compounds in general. The only members of this class which have been studied in some de- tail to date are CHZNZZ, HCECCHN2,3 H3C8CHN2,4 and :_>CN2.5 There- fore, the study of the vibrational spectrum of DAN is of interest to the vibrational spectroscopist since apart from obtaining infor- mation regarding the chemical bonding in this Species, the transfer- ability of force constants among these molecules can also be tested. Furthermore, characterization of the precursor will help in the un- derstanding of the vibrational Spectrum of HCCN due to the presence of similar characteristic groups of atoms in these molecules. The primary goal of this work is to obtain information regar- ding the geometry and nature of chemical bonding of the species HCCN through the study of its vibrational spectrum. A parallel interpre- tation of the vibrational spectrum of the precursor is also given. Comparison of its vibrational potential function with the ones of similar species is critically discussed and the conclusions arrived at from the overall vibrational study are compared with the results of the microwave observations. CHAPTER I THE FREE RADICAL HCCN BaCkgrOund‘InfOrmation on HCCN The only spectroscopic study of HCCN which has been published to date is with regard to its ESR Spectrum, which was reported by Skell, Bernheim et al. at Penn State in 1964.6 HCCN was produced via photol- ysis of its precursor, diazoacetonitrile, isolated in a polychlorotri- fluoroethylene (PCTFE) matrix at 77 K. This report was followed by two 7 and the other in l970.8 others by the same researchers, one in 1965 Their results can be summarized as follows: HCCN possesses a triplet ground state and the g-factors of its electrons are considerably dif- ferent from the g-factor of a free electron, indicating a significant amount of spin-orbit coupling. Linearity' of the free radical was as- sumed because of the zero value of the E zero-field splitting para- meter. In l97O it was realized that free rotation of a bent free radical along its long axis in a matrix results in a low E value, thus giving the false impression that the radical is linear. Since different ma- trices pose different barriers with regard to the rotation of the guest molecules, it was felt that taking ESR spectra of the same free radical in different matrices was necessary before any statements re- garding its geometry could be made. 9 A study of this kind was made on HCCN by Nasserman et al., and a brief note stating the results of that study was included in a pa- per they published on CH2;9 "The attempts we have made to examine HCCN in a variety of matrices at 4 K have only given Spectra corresponding to E=O, thus supporting the linear form." An attempt to study its UV spectrum in the gas phase via flash 10 How- photolysis experiments was made by Merer and Travis in l966. ever, their observations were complicated by the simultaneous produc— tion of other free radicals such as CNC and CCN, thus preventing a complete study of HCCN. In 1968 an LCAO MO calculation of the extended HUckel type carried out on HCCN among other free radicals appeared in the literature.11 The result was stated as follows: "There is no doubt that the ground state of these molecules will be a linear triplet.” The molecules im- plied were HCCN and NCCCN in their carbene configurations. Finally, in a paper published in 1973 by Skell et al.],2 it was shown, based on the stereospecificity of the reaction between ciS-2- butene and the photolysis product of diazoacetonitrile, that HCCN is produced in its singlet state. The triplet state population increased at the expense of the singlet with time, indicating that the triplet is the ground state of this molecule. At this point it should be mentioned that despite the available information in the literature, the linearity and/or electronic charge distribution of HCCN is far from being conclusively proven either ex- perimentally or theoretically. Experimentally, the argument is the lack of hyperfine structural data in the ESR spectra. Theoretically, recent ab-initio calculations on C3H213 andC3N214 free radicals (which were also proclaimed as linear in their ground states based on LCAO MO calculationsl1) suggest that both of these molecules have bent structures. The results of preliminary ab-initio calculations on HCCN done in this department15 make this even more clear. Two minima were found in the potential energy surface of HCCN, one corresponding to a carbene bent structure and the other to a nitrene linear one. The linear form Was found to lie lower than the bent one by 'V4 Kcal/mole. This energy difference however is within the uncertainty inherent in the calculations, leaving the question of linearity and electronic charge distribution still unanswered. ApproaCh to the Study of HCCN In the gas phase HCCN is believed to have a half life of MO”6 seconds (deduced from flash photolysis UV experiments).10 Contrary to their extensive use in the UV region of the spectrum, flash photolysis experiments in the IR are not possible. This is due to the absence of fast responce IR detectors. Therefore the only alternative left for obtaining the IR spectrum of HCCN was to use the matrix isolation tech- inique. Matrix isolation spectroscopy, as it is known today, was developed by Pimentel and his co-workers at Berkeley. Since their first report],6 a multitude of papers have been published on the IR, UV and recently Raman observation of unstable species employing this technique. There are many excellent review articles and books which treat the subject. Some of these, the most informative in the author's opinion, are included in the list of references.17 The method can be briefly described as follows: the molecule of interest or the precursor of the molecule of interest is mixed with at least a hundred fold excess of an inert gas (usually noble gases or N2), and the mixture is rapidly condensed on to the cold substrate of a cryogenic cooler at a temperature sufficiently low that diffusion of the solute species is prevented. The matrix sample so obtained can then be subjected to spectroscopic investigation as it is or, if the Species of interest is reactive, after photolysis of its precursor in the matrix. This is of course a highly simplified presentation of the tech- nique. Many variations have been incorporated over the years, estab- lishing the matrix isolation method as a valuable tool for the spec- troscopic examination of a variety of chemical species.17j One aspect of the technique that should be stressed specifically is the condition of the guest molecules in the matrix. The condition is very similar to the gas phase, something that is exemplified by the fact that absorp- tion bands of matrix-isolated species are generally observed within 5 or l0 cm'1 of the gas phase band origins. This permits the theory of 18 developed for molecules in the gas phase to be molecular vibrations applied equally successfully to matrix-isolated Species. The absorption bands of the guest molecules are very sharp for several reasons. First because low temperatures in general result in small band widths, second because intermolecular interactions present in other condensed phases are almost completely absent in the matrix, and third because rotation is quenched for all but very small molecules. This sharpening of the bands allows near—degenerate bands, which over- lap completely even in the vapor phase or in dilute solution at room temperature, to be resolved in the matrix spectra. However, informa- tion provided by the band envelopes in the gas phase concerning the symmetry of the vibration is lost along with the possibility of deriv- Iing structural data from rotational analysis. Therefore, the prepara-‘ tion and observation of isotopically labelled Species is even more im- portant to the identification and band assignments of free radicals in matrices than it is in gas phase studies. Finally, it Should be mentioned that on occasion matrix spectra are complicated by the Splitting of the absorption bands of the guest molecules due to their entrapment in different sites in the solid lat- tice. This problem however can be easily overcome by annealing the sample at a temperature where the matrix cage loosens enough to allow some movement of the molecules. All the peaks decrease somewhat in intensity, but the ones due to alternate sites decrease much more ra— pidly. The only method by which HCCN has been positively identified is the ESR technique. Therefore, before the IR experiments were begun, a repeat of the ESR experiment done by Skell et al.6 was thought ne- cessary to ensure that the right track was followed. The photolysis light source available to the author was different than the one used by Skell et al., thus making this experiment even more necessary. The matrix sample was prepared the same way as in Reference 6, the only difference being the solvent which was dioctyl phthalate in- stead of methylene chloride. The matrix sample was photolyzed for one hour with a l50 Watt Xenon lamp. A pyrex filter and a water filter were also used. The ESR spectrum obtained is Shown below in Figure l. 6156 G Figure l. ESR Spectrum of HCCN in PCTFE (77 K) (v = 9.207 KMC/SeC) The observed band (whose shape is very similar to the one obtained by Skell et al. )6 disappeared upon annealing of the matrix, indicating that its appearance was due to a reactive species. The field at which this band is observed (6,l56 G ) is slightly different than the one 10 reported by Skell et al.6 (6,256 G). However, it was obtained at a slightly different frequency, and is considered to be well within ex- perimental error. It is quite possible that upon photolysis of a compound in a ma- trix several species are produced. Some of them may be stable, some unstable. It is up to the experimentalist to first minimize the pos- sibility of simultaneous production of species other than the one of interest and second to identify the bands due to the latter. In the case of HCCN, restricting the photolyzing radiation to wavelengths lon- ger than 3,500 A ensured that no C-H bond breakage took place, there- fore overcoming the problem that Merer and Travis had in the gas phase UV experiments. For band identification and eventual Species identification, the 171 was employed. following method, generally accepted in the literature, The rate of growth of new bands as a function of photolysis time and the rate of disappearance of these bands upon controlled diffusion were recorded. One expects that bands due to the same species will have the same rate of growth, while those due to different species will differ appreciably. However, this is not adequate for proving that the new bands are due to reactive species. The proof comes with the results of the controlled diffusion technique. The latter consists of slowly raising the temperature of the matrix up to a certain point where the matrix is known to relax its rigidity (40 K for Ar and 35 K for N2). At this point, some of the free radicals have the chance to escape from their cage and react with one another or other species, thus causing a decrease in intensity of their IR bands. This obviously does ll not affect the bands of stable_species. This way the groups of bands due to reactive species in the matrix are distinguished from those of stable species. Again, the rate of dissapearance of bands as a func- tion of time spent at the controlled diffusion temperature is expected to be the same for bands due to the same species and different for bands of different species. Finally, isotopic substitution and observation of the correspon- ding frequency shifts coupled with normal coordinate analysis allows verification of the band assignments and positive identification of the species responsible for them. In the next section the methods followed to obtain the matrix sam- ples and for the synthesis of all the isotopic precursors of HCCN are described, along with the instrumentation employed by the author in or- der to obtain the spectroscopic data. Method of Obtaining the Matrix Sample The precursor of HCCN, diazoacetonitrile, is explosive,19 has a low vapor pressure ( asooix hv X>3500A hv X>3500A hv X>3500A hv X>3500A Hc‘3CN HCCN DCCN HCCN HCC'SN Figure 4. Scheme of Isotope Synthesis The general synthetic route employed for the preparation of diazo- acetonitrile and its isotopically labelled counterparts is shown sche- matically in Figure 4 above. Modifications of the procedure described in Reference 2l included a different starting material, HZNCHZCN-HZSO4, instead of HZNCHZCN-HCl (the cause of the presence of chloroacetonitrile 17 in the final product) and adeStment of the pH of the aqueous solution of HZNCHZCN-HZSO4 from pH l to pH 4 prior to addition of NaNOZ. It was found that very acidic solutions_prevented diazotization from taking place. Furthermore, the time allowed for reaction prior to each extrac— tion had to be considerably longer since the reaction rate is much slower for HZNCHZCN-HZSO4 than for'HZNCHZCN-HCl.23 The temperature was kept between 10°C and 12°C since lower temperatures resulted in very slow reaction rates. A variety of solvents were used for the extraction of DAN from the aqueous solution; these included dipentyl ether, dichloromethane, di- butyl phthalate and dioctyl phthalate. A useful solvent had to meet the following requirements: immiscibility with H20, low freezing point ( <51C), and vapor pressure different enough from that of diazoaceto- nitrile to facilitate eventual retrieval of the latter in pure form through vacuum distillation. Dioctyl phthalate was the only solvent that met the requirements which allowed complete separation of diazo- acetonitrile from the solvent. Complete removal of water from the final solution of diazoaceto- nitrile in dioctyl phthalate turned out to be a difficult task, and a variety of drying procedures were tested. Anhydrous NaZSO4 and mole- cular sieves (type 4A) seemed to give the best results, although not without considerable loss of diazoacetonitrile due to its adsorbance by the drying agents. P205, CaClZ, CaSO4, and Mg(ClO4)2 did not prove to be successful drying agents. Vacuum distillation employing a variety of cold-temperature baths, attempted after drying the solution with Na $04 or molecular sieves, also did not lower the water content any 2 18 further. As a result, small quantities of water were always present in the spectra.. However, that turned out to be an advantage rather than a disadvantage Since, depending on the number of peaks observed for each vibration of H20, it was possible to assess the effectiveness of the isolation of the trapped species. When only one peak for each normal mode of H20 was observed (water monomer) the quality of the spectra was the best for the species of interest. NNCDCN was prepared by an exchange process. Eighteen mls of NNCHCN in dioctyl phthalate were cooled to 5°C using a cyclohexane- 24 liquid N slush bath and 12 mls of a .05% solution of NaOD in D 0 2 2 were introduced dropwise over a period of one hour. The mixture was vigorously stirred throughout the entire reaction. After all the NaOD solution was added, the mixture was stirred one half hour longer. NaCl was used to facilitate the extraction of diazoacetonitrile from the aqueous layer. Separation of the two layers was more rapidly accom- plished by the USe of a centrifuge (as opposed to a separatory funnel) followed by removal of the upper layer with a hypodermic syringe. Speeding up the separation of the two layers is necessary since NNCDCN is converted back to NNCHCN upon standing in contact with the aqueous layer at room temperature. 60% - 80% deuteration could be achieved this way. Deuteration higher than 60% was achieved only when the reac- tion mixture was centrifuged immediately after the reaction was over. The NaOD solution was added dropwise since the reaction is highly ex- othermic. If the temperature rises above 15°C thermal decomposition of diazoacetonitrile Occurs rapidly. Lowering the temperature below 5°C resulted in the solidification of 020 ( freezing point 3.82°C), so 19 a temperature of 5°C seemed to be the best compromise. After completion of the reaction, the ratio of the amount of deu- terated diazoacetonitrile to the amount of unlabelled diazoacetonitrile in the mixture was found to be independent of the c0ncentration of NaOD in D20. At the same time, the higher this concentration the greater the decomposition of diazoacetonitrile (both labelled and unlabelled), and therefore the lower was the yield of the product. 15NNCHCN (55% in 15N) was prepared in the same way as unlabelled diazoacetonitrile by the use of Na15NO2 (55 atom % 15N) instead of un- labelled NaNOZ. NNCHCN (60% in 13C) and NNCHCN (60% in 15N) were pre- 25,26,21 pared through a three-step process Kl3 as depicted in Figure 4. CN (60 atom % 13C) and KC15N (50 atom % 15N) were used as starting materials. The labelled potassium cyanide starting material was pur- chased from Merck, Sharp and Dohme of Canada, Ltd. Before concluding this section, it Should be mentioned that the synthesis, although seemingly trivial when written in terms of reac- tions, required a considerable amount of time and extra precautions. This was due to the explosive nature of diazoacetonitrile, the purity necessary in the final product, the absence in the literature of any established technique regarding the deuteration procedure, and the 13 15 availability of only small amounts of C and N labelled starting materials. InStrumentation An Air Products model CS-202 Displex cryogenic helium refrigera- tion system was used for all the solid state experiments. Typical 20 temperatures were between 12 K and 21 K measured by a gold (.07% iron) vs. chromel thermocouple imbedded in the cold substrate. Thennal con- tact between the cold substrate and the thermocouple junction was achieved with Air Products Cry-Con grease or woods metal. The volume around the cold substrate was continuously eVacuated by a vacuum line having an oil diffusion pump backed by a rotary mechanical pump. Liquid nitrogen-cooled traps were used between the cooler and the diffusion pump and between the diffusion pump and the mechanical pump. The pressure in the area around the cold substrate prior to deposition was measured by an ion gauge attached directly to the Shroud of the cryo- genic cooler (see Figures 2 and 3). It was normally better than 10"7 Torr. For controlled diffusion experiments, a Cryodial Model ML 1400 automatic temperature regulator was used, with the heater (40 Watt Zener diode) and temperature sensor (calibrated platinum resistor) mounted on a copper block next to the cold substrate. Temperature con- trol capability was i.02 K. For IR experiments, CsI was used for the outer windows and the cold substrate. For UV-Vis experiments, quartz (UV—grade) was used for the outer windows and sapphire for the substrate. Quartz was used for the outer windows and copper for the substrate in the Raman work. Thin sheets of indium were placed between any two parts required to be in good thermal contact. The following spectrometers were employed. Perkin-Elmer Model 225 IR grating spectrophotometer (4000-200 cm']) with resolution better than 1 cm'1 above 500 cm"1 and between 1 and 2 cm"1 below 500 cmu]. Reported frequencies should be 21 accurate to :l‘cm'1. Cary Model 17 UV;Vis spectrophotometer Raman spectrometer comprised of a Jarrell-Ash 25-100 double Czerny- Turner monochromator coupled with a thermoelectrical1y-cooled RCA C31034 photomultiplier tube and a spectra-Physics Model 165 Kr ion laser as exciting source. Baird-Atomic spike filters were used for the 6471 A and 5682 A laser lines to eliminate plasma lines. Varian E-4 EPR spectrometer system with 8.8-9.6 GHz operating frequency range, 100 KHz field modulation frequency and 250-10,000 G magne- tic field strength. For the photolysis of the precursor, a 150 Watt Xenon lamp (Bausch and Lomb) was employed. A 0-52 Corning filter was used to cut off ra— diation of X<3,500 A and a water filter was inserted between the light source and the cryostat to prevent heating of the matrix. A quartz lens was used to concentrate the photolyzing radiation over the area of the cold substrate. Research grade Ar and N2 (Matheson) were used for matrix gases without further purification. The majority of the experimental work was done with Ar. At this point the author would like to justify the use of Ar as a matrix gas as opposed to other inert gases such as Ne, Kr, Xe, and N2 which are extensively used in matrix studies. Xe and Kr are inferior to Ar since they, in general, result in more lightscat» tering due to the larger atomic Size. Ne, which by the same token would have been preferred over Ar, could not be used in these 22 experiments because the low temperature limit of our cryogenic cooler is higher than the temperature at which Ne is known to relax its rigid- ity. Finally N2, which was considered for some time the ideal matrix gas for isolating molecules of the size of NNCHCN and HCCN, was used in a few experiments but was finally discarded for two reasons. First, an extensive Site splitting was observed in almost all bands in the spec- trum which made the band assignment task extremely difficult. Second, there are some recent cases in the literature where complexes with N2 and the species of interest are reported to exist.27 Vibrational Spectra of HCCN and Its Isotopes 28 Application of elementary group theory methods Shows that all vibrations for HCCN (5 for the linear structure, C symmetry or 6 for 00v the non-linear, CS symmetry) are both IR and Raman active. Therefore, theoretically either kind of spectroscopy employed should be adequate by itself in terms of furnishing the needed data (i.e. vibrational fre- quencies). It is well known however that the intensities of some of the vibrational frequencies change drastically in going from one tech- nique to the other. Therefore peaks that are not observed in one case might be intense enough to be observed in the other. In addition, it is is always a way of double checking the validity of experimental frequen— cies, which although not necessary still remains desirable. However, repeated attempts to obtain the Raman spectrum of matrix- isolated HCCN failed. This was of no surprise since no one has been able to obtain conventional Raman spectra of Similar Species to date. This is due to the fact that Raman spectroscopy in general is not as 23 sensitive as IR spectroscopy and also to the inherent weak Raman scat- tering properties of species of this nature. Two other techniques how- ever have been successfully applied to the observation of vibrational frequencies of reactive species, laser induced fluorescence spectrosco- py 29 and resonance Raman spectroscopy.30 Either technique requires the use of a laser line with frequency near a strong absorption band of the species under investigation. Since the electronic absorption spectrum of HCCN was not known, the next step was to obtain this infor- mation in an Ar matrix. This was accomplished and is discussed in de— tail in a later section of this dissertation. The only absorption band which could be attributed to HCCN was a band in the UV part of the spec- trum between 2400 A and 3400 A. Since no laser line is currently avail— able in this frequency range in our laboratory, neither method could be employed. Therefore infrared spectroscopy remained the only source by which information on vibrational frequencies of HCCN could be obtained. Fig- ure 5 shows the IR spectrum of HCCN in an Ar matrix (B) and the same frequency ranges before photolysis (A) and after extensive annealing (C). Figures 6, 7, and 8 show the IR spectra of DCCN, HC13 HCCISN respectively in Ar matrices. CN, and °/oT 24 3250 3200 I750 1720 1190 1170 470 450 1 ' I 1 1 n 1 1 B “V” W“ ”V %T 1735.0chT' 1178.5 oni‘ 458 cm" 3229cm" U (crrT') Figure 5. IR Spectrum of HCCN in an Ar Matrix 2440 2400 1740 1720 1140 11210 420 390 330 310 I I 1 1 IM WW ‘ 1 1127 cm" 405 0m" I ' 3175 crfi \ -1 2424 CNN 1729.5 cm c——\7 (chl—I) Figure 6. IR Spectrum of DCCN in an Ar Matrix 25 32.50 3210 171.0 1690 |i90 "'70 470 440 1176\5 " .cm 45806' %T 1698 cm" 3229 cm" +————fi(an) Figure 7. IR Spectrum of HC13CN in an Ar Matrix 3240 3220 1730 1710 1180 1160 500 480 1 I 1 1 IM' 1 I ”THUNJTJ usecd' I7N3cnfl 458 _| o _ cm /°T 3229 cml 4—V1crfi') Figure 8. IR Spectrum of HCC15N in an Ar Matrix 26 Each observed fundamental frequency corresponds to one normal mode of vibration. The latter is defined as a vibration during which all atoms in the molecule move with the same frequency in such a way that the Cartesian components of the displacements change according to sine curves.31 The normal modes are completely independent of one another, whereas each normal mode may be a mixture of various internal modes such as bond stretching or angle deformations. Although it is easy to measure the normal frequencies, what is of interest in chemistry are the properties of individual chemical bonds. To relate these quanti— ties, one must perform a normal coordinate analysis. Normal coordinate analysis involves computation of theoretical spectra from assumed struc- tures and force constants; these are then brought into coincidence with the observed bands by adjustment of the assumed structure or force con- stants. Unfortunately, the set of force constants obtained this way is not unique. The reason for this is because there are always more force constants in a molecule than normal modes. Therefore, additional data are required to “constrain" the solution so that it fits the observed Spectrum and the additional information. One important source of ad— ditional data is provided by the Spectra of isotopically substituted molecules. Since the bonding Should be essentially unaffected by iso- topic substitution, the normal coordinate analysis can be repeated with only atomic masses changed; the computation should then reproduce the isotopic Spectrum. The most serious defect of the method lies in the assumption of purely harmonic motion. AS anharmoniCity effects are often of the order of 1% of the observed frequencies, it is pointless 27 to try to reproduce the spectrum more precisely than this unless enough information on anharmonicity can be obtained to enable corrections to be made on the observed frequencies. In this work normal coordinate analysis was performed by using the Shimanouchi computer programs.32 Two articles written also by Shimanouchi et a1. were found pertinent 33,34 for the understanding and more efficient use of the programs. The calculations were carried out on a CDC 6500 computer located in the Computer Center at Michigan State University. A detailed description of normal coordinate analysis will not be presented here since it is not essential to the understanding of the text. Excellent accounts on 35 A brief normal coordinate analysis are abundant in the literature. discussion of force constants and their significance will however be given because of their importance in understanding other sections later in the thesis. The potential energy of a molecule (V) as a function of its 3N-6 internal displacement coordinates, Ri (which Specify completely the in- ternal configuration of the molecule, where i=1 to 3N-6) may always be expanded in a Taylor series about the equilibrium configuration:36 2 _ 3V 3 V (1) V ‘Ve + 12[§R_.]e (Ri) + 1/2§}[6Ri—21R371e(Ri-11Rj1+°“ In the harmonic approximation terms of cubic order and higher are dis- carded. The first term, Ve’ simply defines the arbitrary zero of the energy scale while the second term is zero since the derivatives are taken in the equilibrium configuration, in which by definition V is a 28 minimum with respect to all the R1. So, equation (1) becomes: .azv v =1/2 ZZLRiaR.]e(Ri)°(R1) l J J The terms in brackets represent the force constants, fij’ of the mole- cule under consideration. These are the parameters a vibrational spec- troscopist hopes to get from the observed vibrational frequencies. fij is the restoring force in coordinate i caused by a small unit displace- ment of coordinate j, keeping the other coordinates fixed.37 If i=j= a bond length, then fii is a measure of this bond's strength (although not completely comparable to a diatomic molecular force constant since it includes effects arising from changes in other internal coordinates) which is determined by the electronic distribution when the nuclei are at their equilibrium positions. Caution should be exercised in the use of the term "bond strength” Since the term used here is different than the energy required to break a bond. Rather the force constant is a mea- sure of the bond strength at the minimum in the potential energy curve and is therefore directly related to the equilibrium configuration. The dissociation energy (DE) is a different measure of strength since when a bond stretches enough to dissociate, its electronic struc- ture (and apparent force constant) changes considerably. Therefore, the dissociation energy is not closely related to the binding forces at the equilibrium configuration. The usefulness of force constants in general relies on: 1) the correlation of their values with bond nature, electron delocalization, and interatomic interaction. 29 2) the use of their values to calculate and estimate vibrational frequencies and the further use of the results for band assign- ments. Perhaps the best way to close this brief presentation on force constants is by giving a definition of the latter from a different per- spective: Near the equilibrium point the force constant associated with the repulsive energy is dominant over that associated with the attractive energy. When the binding force between two atoms forming a chemical bond.is large, the distance be- tween the two atoms becomes small, the repulsion between the two nuclei including surrounding inner-shell electrons is large, and, accordingly the force constants are also large. In this way the force constant is a measure of the binding force.33 Since for reasons stated earlier the linearity of HCCN could not be taken for granted, it seemed necessary to perform two parallel nor- mal coordinate analyses assuming two different structures, one bent and one linear. Table I shows the molecular parameters used for the normal coordinate analyses of the two structures along with the molecular par- ameters corresponding to an allene—type structure transferred from the 38 (allene) and NCN.3g molecules C3H4 These molecular parameters were used for a third normal coordinate calculation to test the sensitivity of the frequency fitting and values of force constants as a function of the bond lengths of the molecule. 30 aowz zoo oowz zoo oowz zoo comp oo: . oomz oo: oomz oo: m mm.— zuo < ¢¢.F zio < om_.z zWo N no.— 11o < mo._ Ito < wo.z Ito m om.— ouo < mp.~ omo < mwm.z oio .mpmcozpm_:o_mo m—mcozpozzozmo mm.wmzoz .m:m__< ”mmozzom owpozin< zzmcwEWszo "moezom ozow:zin< zxocwEPcho ”mozzom zozwz-oeo__:o zoo: zozz_-oeoaoozo zoo: zo:w_-oeaoaaoo zoo: zoo: mo mzmz_mc< mpmczvzooo zmszoz mzp 20% com: mzmmemcmo gmzzomzoz .H wznmz 31 Figure 9. Internal Coordinates for the Free Radical HCCN (A- Linear, B- Bent) Figure 9 and Table II give the internal and symmetry coordinates of HCCN for the bent structure and the linear. Symmetry coordinates are derived from internal coordinates (which are the bond lengths and an- gles of the molecule) by use of elementary Group Theory methods.28 The reason behind their use is due to the great Simplification of nor- mal coordinate analysis which, in this way, can be performed on one symmetry species (of the symmetry point group the molecule belongs to) at a time. This is due to the simplification of the corresponding di- agonalization of the kinetic energy and force constant matrices which are directly involved in the calculation of vibrational frequencies. 32 Table II. Symmetry Coordinates for the Free Radical HCCN Linear Form Bent FOrm _Cav.5¥mmetry Point Group ..C,;$¥mmetr¥:991nt Group, S1 7 Ar23 S1 = Ar23 S2 = A"14 S2 = A"T4 S3 = Ar12 A' S3 Ar12 S4 = A0 S4 = Am] 35 = Am 55 — sz A”. S6 = A6 The initial force constants were transfered from the following 40 41 42,43 free radicals: C3H2, CCO and NCN. The results of the force constant calculations are shown in Tables III, IV, V, VI and VII. Since only five frequencies were observed at most, two normal coordi- nate analyses had to be performed on the bent structure depending on 1 the assignment of the frequency observed at 405.0 cm' as an in—plane or or out-of—plane fundamental. Note that on the following Tables (III, IV, V, VI and VII ) the nota- tion in the column under PED is as follows: AB= AB stretch ABC= ABC bend 33 Table III. Normal Coordinate Analysis of HCCN: Linear Form*; Nitrene-like Molecular Parameters Frequencies (cm'1) Primary Contributors ' 0 "Obs.' ' ca1. 'Ap "'PED***" ‘ HCCN \fi ‘ 3229.0 3228.9 (-0.1) CH(96) 2+ 92 1735.0 1737.3 (+2.3) CN(103) CC(28) 03 1178.5 1178.7 (+0.2) CC(80) up 458.0 458.5 (+0.5) CCH(108) CCN(17) H vs .<359:5)** 372+0,1 w <+2-5>1, .cc~<92>, DCCN 9] 2424.0 2424.9 (+0.9) CD(90) + 2 1729.5 1730.8 (+1.3) CN(104) CC(24) E 93 1127.0 1128.8 (+1.8) CC(79) CD(8) H v4 405.0 405.0 (0.0) CCN(85) CCD(56) 05 317.5 317.5 . (0.0) . CCD(53)CCN(24) 13 “C C” 9] 3229.0 3228.6 (-0.4) CH(96) + 02 1698.0 1695.7 (-2.3) CN(104) CC(26) 2 93 1176.5 1175.1 (-1.4) CC(82) 04 458.0 457.1 (—0.9) CCH(109) CCN(14) H 95 (364.5)** 363.0 (-1.5) CCN(95) 15 “CC N 9, 3229.0 3228.9 (-0.1) CH(96) + 92 1718.0 1716.5 (-1.5) CN(102) CC(31) 2 03 1168.0 1167.3 (-0.7) CC(77) 04 458.0 458.4 (+0.4) CCH(108) CCN(17) H,. 05 ._-.--,. . 370.3. -_--. . .CCN(92).... *C v symmetry point group 00 .**Uncertain band, not used in f0rce constant calculations ***Potential Energy Distribution 34 Table IV. Normal Coordinate Analysis of HCCN: Bent.Form* (405cm'1 band as in-plane) Frequencies (cm—1) Primary Contributors 0 Obs. Cal. Av ‘PED*** ' ”CC” 9, ‘ 3229 0 ‘ 3236.8 (+7.8) CH(99) 02 1735.0 1736.4 (+1.4) CN(86) CC(38) A' v3 1178.5 1170.6 (-7.9) CC(65) CN(18) V4 458.0 458.6 (+0.6) CCH(99) CCN(16) V5 (369.5)** 379.2 (+9.7) CCN(87) A" 06 ----- 313.7 ---- CCN(100) DCCN 9] 2424.0 2391.4 (-32.6) CD(97) 02 1729.5 1723.3 (-6.2) CN(87) CC(36) A' V3 1127.0 1155.3 (+28.3) CC(65) CN(17) V4 405.0 404.8 (-0.2) CCN(86) CCD(35) 05 317.5 317.7 (+0.2) CCD(68) CCN(17) A" 06 ----- 313.7 ---- CCN(100) HC13CN V1 3229.0 3236.8 (+7.8) CH(99) 02 1698.0 1691.4 (-6.6) CN(87) CC(37) A' 03 1176.5 1169.5 (-7.0) CC(66) CN(17) 04 458.0 456.6 (-1.4) CCH(100) CCN(12) 05 (364.5)** 370.4 (+5.9) CCN(91) A" 06 ----- 305.3 ---- CCN(99) HCC15N 01 3229.0 3236.8 (+7.8) CH(99) 02 1718.0 1720.7 (+2.7) CN(84) CC(41) A' 03 1168.0 1156.2 (-11.8) CC(62) CN(20) 04 458.0 458.4 (+0.4) CCH(99) CCN(16) 05 ----- 376.8 ---- CCN(87) A" 06 ------ 311.9 --—- CCN(99) *C symmetry point group, **uncertain band, ***P0tential Energy Distrib. S , 35 Table V. Normal Coordinate Analysis of HCCN: Bent Form*; (405 cm'] band as out-of—plane) Frequencies'(cm-]) Primary Contributors v Obs. Cal.i Av PED*** ”CC” 1 3229.0 ‘ 3236.8 (+7.8) CH(99) 92 '1735.0 1736.4 (+1 4) CN(86) CC(38) A‘ 03 1178.5 1170.6 (-7.9) CC(65) CN(18) 04 458.0 458.4 (+0 4) CCH(102) CCN(10) 95 ----- 367.9 ---- CCN(93) A" V6 (369.5)** 403.4 (+33 9) CCN(100) DCCN v, 2424 2391.4 (-32.6) CD(97) 92 1729. 1732.3 (+2.8) CN(87) CC(36) A' 93 1127. 1155.3 (+28.3) CC(65) CN(17) v4 ..... 392.7 ---— CCN(81) CCD(43) 95 317 5 317.6 (+0.1) CCD(60) CCN(23) A" 96 405.0 403 4 (-l.6) CCN(100) 13 “C C” v, 3229 0 3236.8 (+7.8) CH(99) 02 1698.0 1691.4 (-6.6) CN(87) CC(37) A' 03 1176.5 1169.5 (-7.0) CC(66) CN(17) 04 458.0 457.0 (-1.0) CCH(103) 05 ————— 358.9 -—-- CCN(95) A" 96 (364 5)** 392.6 (+28.1) CCN(99) 15 “CC N 0] 3229.0 3236.8 (+7.8) CH(99) 02 1718.0 1720.7 (+2.7) CN(84) CC(41) A' 03 1168.0 1156.2 ‘(-1l.8) CC(62) CN(20) 04 458.0 458.2 (+0.2) CCH(102) CCN(10) 05 ----- 365.5 ---- CCN(93) A" 06 ----- 401.1 ---- CCN(100) *CS symmetry point group .**Uncertain band, not used in force constant calculations .,***Potential Energy Distribution 36 Table VI. Normal Coordinate Analysis of HCCN Linear F0rm* Allene-like Molecular Parameters Frequencies (cm-1) Primary Contributors 0 Obs. Cal. . Au, PED*** HCC” 0] 3229.0 3228.9 (-0.1) CH(96) + 02 1735.0 1737.3 (+2.3) CN(103) CC(28) Z 03 1178.5 1178.7 (+0.2) CC(80) 04 458.0 458.5 (+0 5) CCH(103) CCN(17) H 05 (369.5)** 371.0 (+1 5) CCN(89) DCCN 0, 2424.0 2424.9 (+0.9) CD(90) 2+ 02 1729.5 1730.8 (+1 3) CN(104) CC(24) 03 1127.0 1128.8 (+1 8) W( 9) D(8 ) v4 405 0 405.0 (0 0) CCN(81) CCD(49) H 05 317.5 317.7 (+0.2) CCD(56) CCN(24) 13 “C C” 01 3229.0 3228.6 (-0.4) CH(96) 2+ 02 1698.0 1695.7 (-2.3) CN(104) CC(26) 63 1176.5 1175 1 (-1.4) CC(82) 04 458.0 456.9 (-1.1) CCH(104) CCN(14) H 05 (364.5)** 362.2 (-2.3) CCN(92) 15 ”CC N v] 3229 0 3228.9 (-0.1) CH(96) 2+ 02 1718.0 1716.5 (-1.5) CN(102) CC(31) 03 1168.0 1167.3 (-0.7) M( 7) H 94 458.0 458.5 _ (+0.5) CCH(103) CCN(16) 05 ----- 368.8 ---- CCN(89) *me symmetry point group **uncertain band, not used in force constant calculations ***Potential Energy Distribution 37 Table VII. Force Constants for HCCN Bent Form . Bent FOrm 405.0 cm‘1 band as CCN ip 405 0 cm’1 band as CCN op KCC= 7.480 (1.220) KCCT 7.480 (1.287) KCH= 5.716 (0.030) KCH= 5.716 (0.032) KCN= 10.922 (2.236) KCN= 10.921 (2.358) HCCH= 0.122 (0.007) HCCH= 0.124 (0.065) HCCN_ 0.315 (0.038) HCCN: 0.294 (0.791) HCCN: 0.192 HCCN: 0.317 (0.025) FCC,CN= 2.003 (0.899) FCC,CN= 2.002 (0.949) FCCH,CCN= 0.040(0.015) FCCH,CCN= 0.040 (0.079) Linear Form Linear Form Nitrene-like Bond Lengths Allene-like Bond Lengths KCC= 7.533 (0.049) KCC= 7.533 (0.050) KCH= 5.566 (0.008) KCH= 5.567 (0.008) KCN= 12.241 (0.168) KCN= 12.241 (0.170) HCCH= 0.116 (0.001) HCCH= 0.113 (0.001) HCCN: 0.335 (0.004) HCCN: 0.307 (0.004) FCC,CN= 2.842 (0.111) FCC,CN= 2.842 (0.112) FCH,CC= -0.456 (0.032) FCH,CC= -0.456 (0.033) FCCH,CCN= 0.060 (0.001) FCCH,CCN= 0.046 (0.001) Numbers in parentheses represent uncertainties; see Appendix A. K- stretching force constant (mdyn/A) H- in-plane bending force constant (mdyn°A) H'- out-of—plane bending force constant (mdyn-A) F- stretching-stretching interaction f0rce constant (mdyn/A) F"- bending-bending interaction force constant (mdyn-A) _ 38 Assignment of the Observed Fundamentals* of HCCN Since the normal coordinate analysis supports the linear geometry, the following discussion is based on the results shown in Tables III and IV. V1 The band observed at 3,229 cm'1 in HCCN is shifted down to 2,424 cm'1 in DCCN. No detectable frequency shift in the other two isotopes could be observed and no resolvable Shift is predicted by the normal coordinate analysis. The frequency range within which this band ap- pears, the large shift upon deuteration and the normal coordinate analy- sis results all point towards assignment of this fundamental as the (v1) C-H stretch. 02, v3 The two bands at 1,735 cm'] and at 1,178 cm'1 will be examined together due to their behavior upon isotopic substitution which shows that a strong coupling exists between them. Their isotopic pattern is very similar to the one observed for v3 (anti-symmetric stretch) and v] (symmetric stretch) of NCN42’43 (which is isoelectronic to HCCN) as is shown on the following page. *The fundamentals are numbered in decreasing order of frequency by convention.31 (A) CD I HCCN : NCN I I I v? 05 1 .V] ~93 HCCN 1,735 1,178.5 : 1,197 1,475 NCN HC‘3CN 1,698 1,176.5 1 1,195 1,435 N‘3CN I HCC‘SN 1,718 1,168 5 1,178 1,468.5 Nc‘5N (frequencies are given in cm-]) 1 1 Accordingly, the bands at 1,735 cm' and 1,178.5 cm" are assigned as the (02) anti-symmetric and (03) symmetric stretch of the CCN group respectively. The normal coordinate analysis results support this as- signment. v4 The band observed at 488 cm-1 in HCCN was shifted down to 317.5 cm'] upon deuteration while it remained unshifted in the other two iso- topes. This large isotopic shift alone implies that this band is due to a hydrogen motion of some kind. Since the CCH bending motion is the only other hydrogen motion left, it seems appropriate to assign this band to the (v4) CCH bend. Again this assignment is fully supported by the normal coordinate analysis results. It is interesting to note that the ratio of the two isotopic frequencies (458/317.5) is 1.44 which is higher than the theoretically predicted one (1738): This is really sur- prising since the ratios of experimentally observed isotopic frequencies of similar motions in other molecules are found to be lower than 1.38 40 (e.g. cyanoacetylene,44 H-CEC-CEN, for which the ratio is 663/552=l.27). Anharmonicity effects usually account for ratios smaller.than the theo- retically predicted ones. Therefore, this large shift is probably due to the fact that a rather drastic difference exists in terms of the pri- mary contributors to this normal mode between HCCN and DCCN . This ar- ~gument seems to be supported by the n0rmal coordinate analysis results. (See Table III) V5 A band observed at 405 cm'] in the deuterated species is assigned as the deuterated analog of the (05) CCN bend. This assignment is based solely on the results of the normal coordinate analysis, Since this band's counterparts in HCCN and the other two isotopes have not been positively identified. Two weak bands, one at 369.5 cm'] and one 1 13 at 364.5 cm" were observed in HCCN and HC CN respectively. These bands grew upon photolysis and disappeared upon annealing but their ra§e_of growth and disappearance was hard to establish due to their weak intensities. Therefore they are tentatively assigned as the 405 cm'1 band counterparts in HCCN and HC13CN. The normal coordinate analy- sis results are not incompatible with this assignment. An interesting feature of 05 is its shift towards higher frequency upon deuteration. (This situation although rarely encountered is reported in the litera- ture for CH N22 and HCECCHN2.3) This again can be explained by in- 2 voking the difference in terms of potential energy contributors for 05 between HCCN and DCCN. 1 Finally, if HCCN is not linear, the band at 405 Cm' would have two possible assignments, one as a CCN in-plane bend (05) and another 41 as a CCN out-of—plane bend (06). Results of the corresponding normal ' coordinate analysis treatments are shown in Tables IV and V. These possibilities will be discussed in the following section. Other Bands Upon photolysis of diazoacetonitrile several bands grow in addi- tion to the ones assigned to fundamentals of HCCN. None of these how- ever showed behavior attributable to a reactive species. These bands are obviously due to stable molecules which are formed in the matrix through reactions between two or more HCCN molecules, diazoacetonitrile and impurities due to the synthesis methods employed. Their appearance in the matrices was not reproducible, which indicates that impurities are largely responsible for these bands since impurities are expected to differ in the different synthetic routes employed. The Vibrational Potential Function of HCCN The force constants obtained from the normal coordinate analysis performed on the linear configuration will be used here in this discus- sion, the reason being the considerably better frequency fit for this structure as compared to the bent. These f0rce constants are shown in Table VII on page 37. The following force constants45 may be used as a guideline for an easier understanding of the discussion centered around the force con- stants of HCCN. Their values are considered typical for the bonds writ- ten next to them. 42 C-H Sp 4.8 C-C sp3-sp3 4.5 C-N ~4.8 C-H sp 5.3 C=C spZ-spz 9.7 C=N (spZ—C) 10.5 C-H sp 5.9 CEC sp-sp 15.6 CEN (Sp-C) 17.73 The stretching force constant for the CH bond is calculated at 5.567 mdyn/A which lies between the one for cyanoacetylene44 (5.86 mdyn/ A) and that of methyl cyanide46 (5.0 mdyn/A) and is similar to that of ketene47 ( 5.439 mdyn/A, taken as an average between the two force con- stants corresponding to the symmetric and anti—symmetric stretch of the CH2 group). Therefore the value for HCCN seems to be compatible with a structure of the CCH group more like H-C=C (which is the case in ketene) rather than H-CEC or H—C-C which are the cases in cyanoacetylene and methyl cyanide respectively. The CC stretching force constant is calculated at 7.533 mdyn/A and is much higher than that of methyl cyanide (5.161 mdyn/A), somewhat lower than that of ketene (8.387 mdyn/A) but very close to one of the two CC stretches of cyanoacetylene (7.83 mdyn/A, corresponding to the CC bond adjacent to the CN bond), a molecule in which resonance is quite extensive. From the above comparison, it seems that the CC stretching force constant is c0mpatible with a CC bond lying somewhere between a single and a double bond but closer to the latter. The CN stretching force constant is calculated at 12.241 mdyn/A which is lower than those of methyl cyanide (17,982 mdyn/A) and cyano- acetylene (15.7 mdyn/A) but higher than that of diazomethane2 43 (8.34 mdyn/A) therefore implying a CN bond lying somewhere between a double and a triple bond, but closer to the former. Now that the CC and CN bonds have been discussed individually, it seems appropriate in view of their strong coupling to examine them as a .group and compare the latter with those of similar isoelectronic free radicals. The Table below makes such a comparison much easier. KCC or KCN KCO or KCC or KCN (mdyn/A) (mdyn/A) CCO 7.97 14.06 HCCN 7.53 12.24 HCCCH4O 12.05 12.05 NCN 8.60 8.60 An interesting pattern emerges. If the values of the force constants can be taken as rough measures of the electron densities between these atoms (which is not unreasonable, since they are measures of the bin- ding forces between these atoms and these f0rces are presumably created by a strong electron overlap between the atoms) then the variation of the force constant values can be explained by simply invoking the dif- ference in electronegativities between the atoms C, N, and 0. Using HCCCH as a starting point, one would expect that in going from this mol— ecule to NCN a considerable decrease in electron density around the car- bon atom would take place with.a Simultaneous increase around the ni- trogen atoms; This would cause some weakening of the CN bonds and therefore lowering of their force constants; 0n the other hand going from HCCCH to HCCN one would expect the electrons to shift towards the 44 nitrogen end of the molecule which would tend to break the equal strength balance of the two CC bonds in favor of the CN bond; this effect is expected to be even more pronoUnced as one moves to the C00 molecule. Indeed the argument, Simple as it is, seems to qualitatively explain the force constant pattern in these simple isoelectronic free radicals. The force constant of the CCH bend is calculated at 0.124 mdyn°A. This value is lower than that of cyanoacetylene (0.150 mdyn-A) and those of propargyl halides (0.14 mdyn-A).48 This can be explained by the fact that the unpaired spin density on the carbon atom next to the hydrogen would tend to more easily allow the rehybridization which accompanies the CCH bending and thus lead to a lower bending force con- stant.40 The force constant of the CCN bend is calculated at 0.307 mdyn-A for the allene-type linear configuration, but at 0.335 mdyn'A for the nitrene-type linear configuration. This is actually the only force con- stant where a significant value difference appears between these two configurations. Both values are considerably higher than that of cyano- acetylene (0.210 mdyn-A), but very close to those of aliphatic nitriles (30.310 mdyn°A).49 Finally three interaction force constants were found to be impor- tant for a good_agreement between the observed and calculated frequen~ ' cies, namely those of the interaction between CCH and CCN bends (+0.046 mdyn3A), between CH and CC stretches (-o.456 mdyn/A), and especially be- tween CC and CN stretches (2:842 mdyn/A): The high value of the latter is typical of Similar free radicals Such as NCN (3.22 mdyn/A), 45 C00 (2.37 mdyn/A) and HCCCH (0.87 mdyn/A). The high value of the CC,CN stretching interaction force constant is.the reason why it is inappro— priate to assign v2 and 03 as "primarily CN and CC stretches, respec~ tively." Rather, they are designated as the anti-Symmetric and Symme— tric stretches of the CCN group. ConclUsion Although only five frequencies were observed for HCCN, this does not prove that the molecule is linear since one can always argue that the sixth frequency was too weak to be observed. This is indeed the 50 (based on elementary case for HOCN which is expected to be non-linear considerations of the molecular orbitals involved in the bonding between an 0-H group and a CEN group) and for which only five frequencies were observed in the IR spectrum.50 Therefore the observation of only five frequencies can be considered as evidence towards the linearity of the molecule, but certainly not proof. However, a mere comparison of the normal coordinate analysis results for the two structures shows a much better frequency fir for the linear rather than the bent structure. (The frequency fit for the latter does not Show any improvement when the 405 cm"1 band is assigned to Y6 rather than to 95 as a comparison be- tween Tables IV and V reveals.) This is considered strong evidence that HCCN is linear. Another fact that points towards the linearity of the molecule is the very low frequency of the HCC bending motion (458 cm'1). CCH bending frequencies in non—linear molecules usually lie above 750 cm'1 . 51 Although more than 8 force constants were initially used for both 46 structures, it was found that the fit was equally good with only 8 force constants, indicating that the rest of the interaction force constants were of little importance to the vibrational potential fonction. It is interesting to note that for the linear configuration the total number of force constants comprising the general valence f0rce field is 9 and therefore calculation of the 9th f0rce constant results in the Com- plete determination of the f0rce constant matrix. The value of this force constant which, is a measure of the interaction between the CH and CN bonds, was found to be 0.022 mdyn/A with an Uncertainty of 0.175. Unfortunately the large uncertainty associated with this force constant makes its value meaningless. Finally attention is called to the large interaction force constant between the CC and CN bonds. According to a discussion of the physical significance of interaction force constants by Linnett and Hoare,43’52 the relatively large positive value of the interaction constant for HCCN implies the presence of delocalized electrons in the groud state of this species. This is supported by the isotopic frequency shift pat- tern which implies that the CC and CN bonds are of similar strength. The above may be interpreted to suggest that the structure of HCCN lies somewhere between a carbene and a nitrene form, i.e. H-C=C=N, which is an allene type structure. This structure is also favored by the arguments used in the discussion of the vibrational potential function of HCCN in the previous section. 47 1 i x i X 1 " ' . . \ ' 1 \. 1 T‘\- \ 1 \ ~ :— . L 4 l J l 1 L l 1 1 1 l 1 1 L 1 1 1 1 2000 A 2500 A 3000 A 3500 A Figure 10. UV Spectrum of Diazoacetonitrile Before Photolysis (dotted line) and After Photolysis (solid line) The Electronic Absorption Spectrum of the Free Radical HCCN The electronic absorption spectrum of diaancetonitrile isolated in an Ar matrix, before and after photolysis, was scanned between 185 nm and 800 nm. Only the UV part of the spectrum however is shown in Figure 10, since no bands were observed in the wavelength region between 350 nm and 800 nm. 48 The multiplet structure whichpappeared.between.2400 A and 3400 A after the precursor was photolyzed showed Similar behavior upon annealing as the bands of HCCN in the IR and therefore is attributed to the free ra- dical HCCN. The multiplet structure is rather complicated but at least one progression can be singled out(members of the progression are marked with an x) with a spacing of 1052 wavenumbers which may be asso- ciated with the upper symmetric stretch of the CCN group (the ground state symmetric stretch appears at 1178.5 cm’l). It is interesting to compare the observed spectrum with spectra of 42( other free radicals such as NCN which is isoelectronic with HCCN) and which has one more electron than NCN and HCCN). NCN shows a progression between 3500 A and 2400 A involving excitation of the upper state symmetric stretching fundamental, while NCO shows a progression between 3200 A and 2650 A which also involves excitation of the upper state symmetric stretching frequency. These two progressions have been assigned to n-+ n transitions, more specifically: B(32u') - X(329') for NCN and B2(H) - X(2H) for NCO. However both of these free radicals Show another intense absorption band (0 + n transition) without any vibrational structure, A(3Hu) - X13Zg') for NCN and A(2)u+) - x(zn) for NCO, at longer wavelengths which is missing in the HCCN spectrum. It is con- ceivable that the analogous band in HCCN is also observed but is not as intense, since weak peaks are observed in the spectrum of HCCN around the wavelength range where the band is observed for NCN (3290 A). If this is the case, though, it is not clear why the relative intensities of the two bands Should be so different in the two molecules. The most 49 obvious explanation is that.the band is missing because the electron in the molecular orbital responsible for this transition in NCN is not available in HCCN due to its participation in the CH bond, since this is the part of the molecule that differs from NCN. Regardless of what the reason for this apparent discrepancy is, it does not seem unreasonableé- based on the strong Similarities of the two molecules shown by their vibrational Spectra, the apparent similarity of the progressions in the electronic Spectrum and their being isoelec- tronic-- to tentatively assign the progression observed in the UV spec- trum of HCCN to a 3) - 32 transition involving excitation of an elec- tron from a n orbital of lower energy to a n orbital of higher energy, in view of a similar assignment made for the analogous progression in NCN. The complicated structure of this band as opposed to the simple 55 since HCCN is not a progression observed for NCN is not unexpected simple triatomic molecule with a center of symmetry like NCN. The electronic absorption spectrum of a (50%-50%) mixture of HCCN and DCCN was also taken, but no change in the vibrational spacing was observed despite the large shift that the symmetric stretch was found to 1 to 1127 cm'1). experience in the IR spectra upon deuteration (1178.5 cm- This is probably due to the resolution under which the UV spectra could best be taken (1 A, which corresponds to I~40 cm'1 in this part of the spectrum). 50 Discussion of theCompatibility of the ESRgpatafwith the Proposed Allene-Type Structure of HCCN Only the problem of the electron distribution will be dealt with here, since the linearity of the molecule, strongly indiCated by the vibrational spectrum, is in agreement with the ESR data available to date (see Section on background information on HCCN). Therefore, the linearity of HCCN will be assumed throughout this section. Neglecting hyperfine interactions, the spin Hamiltonian for a linear triplet mol- ecule is written a556 H = BH:9°5 + D(SZ2 - 2/3) where H= the spin Hamiltonian, B= Bohr magneton, H= the magnetic field, g= g-factor, D= zero-field splitting parameter, and S= the total spin angular momentum (S2 is the 2 component of S). The value of D is roughly inversely proportional to the cube of the separation of the two unpaired spins and is a measure of electron delocalization. Therefore, the discussion will be based on the D values of HCCN and other free ra- dicals available to date. The 0 value for "fixed” (i.e. non-rotating in the matrix) methylene 1 1 57 ranges from 0.74 cm' to 0.93 cm' depending on the matrix, while that -1 58 of NH is 1.86 cm Having established the 0 value for the two ex- treme cases, i.e. methylene (both unpaired electrons localized at the carbon atom) and nitrene (both unpaired electrons localized at the nitro- gen atom) we proceed to examine what happens when the opportunity for delocalization is offered to the unpaired electrons. HCCCH and NCN pro- vide good examples of such a case. Both of theseITEe radicals are 13,40,42,59 known to have largely delocalized electrons. The 0 value of 51 HCCCH is 0.628 cm'1.which is significantly lower than.that of CH The * . 2 3 two unpaired electrons in this molecule are, due to delocalization, fur- ther apart than they were in CH2, thus resulting in a lower 0 value. A similar situation exists between NCN and NH: The 0 value of NCN is 1.544 cm'1 60 which is much lower than that of NH. If HCCN had a carbene-type configuration, its 0 value would be very similar to that of CH2 although somewhat larger because of the ad- ditional muclear charge due to the presence of the nitrogen atom in the molecule.6 If HCCN had a nitrene configuration, its 0 value would be very similar to that of NH. The 0 value of HCCN has been found to be 0.836 cm'] 7 which is sig- nificantly larger than that of CH2. The difference between the two values is too large to be explained as being due only to the presence of the nitrogen atom in HCCN. Thus, at first glance it seems that the nitrene-type configuration contributes significantly (in terms of valence bond theory) to the overall 0 value, something that has already been speculated in the literature.7 However a closer look at the fac- tors which contribute to the D valueshows otherwise. Along with the spin dipole-dipole interaction that provides a first order contribution to the overall 0 value, there is a second order con- tribution which is due to the spin-orbit coupling. This contribution 57 ( 10%) and HCCCH61' although small for CH2 ( 16%) is expected to be much larger for small free radicals containing nitrogen atoms. There- fore, for our purposes a more meaningful comparison between the D values of these free radicals can be made only after the spin-orbit contribu- . tions are taken out of the D values. Unfortunately the spin—orbit *Assume D= 0.74. 52 contribution to the D value_of HCCN is.not.known.._However, one can get some idea of the magnitude of this contribution by citing the results obtained for a very similar free radical, namely isoelectronic NCN. ‘According to ab+initio calculations employing minimal STO}4G and extended 4-310 basis sets,62 the spin-orbit contribution to the total 0 value for NCN is 60%!!! Even if the spin-orbit contribution for HCCN is only 20% (which is not unreasonable in view of the 16% contribution in HCCCN), its 0 value comes down to .690 cm'1. This is lower than ' that of CH2, which becomes .722 cm'] after the spin-orbit contribution to the 0 value of the latter is taken out. Electron delocalization must now be assumed to account for the difference between the 0 value of HCCN and that of CH2. It should be noted that the 19we§t_experimen- tal 0 value for CH2 (.74 cm']) has been used throughout this discussion. Therefore, the value .722 cm"1 represents roughly the lower limit of the 0 value of CH2.. If the 0 value of CH2 is taken to be larger than .74 cm"1 and the spin-orbit contribution in HCCN larger than 20%, the difference between the two values will increase further, thus favoring even more the allene-type configuration where delocalization of the unpaired electrons is quite extensive. In view of the above it can be concluded that the ESR data avail- able to date are compatible with the proposed allenejtype structure for 'HCCN. However positive proof of this and the linearity may come only with hyperfine structure data obtained from the isotopic modifications of HCCN. CHAPTER II A STUDY OF THE VIBRATIONAL SPECTRUM OF DIAZOACETONITRILE Background Information on Diazoacetonitrile Diazoacetonitrile (DAN) was first prepared in 1898 by Curtius63 who found it to be an orange-yellow oily liquid boiling at 46.5°C (15mm Hg). He also found that when isolated it is liable to explode violently. Some IR and UV data on diazoacetonitrile have been reported in the lit- erature. Specifically three UV bands (2050 A, 2480 A, and 3280 A) and 1 1 three IR bands (3110 cm- , 2220 cm' , 2100 cm']) are mentioned. Both the UV work and IR work have been done in solution (EtOH and chloroform respectively). From a spectroscopist's point of view these results are rather meager. This is probably due to the different perspective asso- 64 Diazoacetonitrile was also found to exhibit a 64 ciated with that work. strong absorption in NMR at 5.50 T (CDCl3 was used as a solvent). The most complete spectroscopic work to date has been done in the MW by 65 Their observations were consistent with C.C. Costain and J. Yarwood. a planar structure and the molecular parameters they obtained for dia- zoacetonitrile are shown in Table VIII on page 62. The dipole moment of the molecule was found to be 3.45 10.07 D with the vector having a direction almost parallel to the CCN chain (see Figure 18). 53 54 Approach to the Study_of Diazoacetonitrile Due to diazoacet0nitrile's low vapor pressure and high instability (halleife —'5 minutes when in the IR beam), it Was impossible to obtain its IR spectrum in the gas phaseI' (The thought of using a long path cell was discarded since diaancetonitrile leaves brown stains, on everything it contacts, which are hard to remove.) This was a serious drawback since the information regarding the symmetry of the vibrations which could be obtained fron the band envelopes in the gas phase spectra was lost. Consequently, the spectra of diazoacetonitrile in its matrix-isolated form were the main soUrce of information available for the band assignment task. A helpful supplement regarding the latter were the spectra in the pure solid form. Spectra of diazoacetonitrile in solution (solvents employed were dichloromethane and dioctyl_phthal- ate) were useful for quickly checking the success of the synthesis before engaging in the long and frustrating matrix isolation experiments. Bands due to diazoacetonitrile were easily identified in the spec- tra by the decrease in their intensity upon phtolysis. Peaks due to impurities in general did not Show any appeciable change in their inten- sities. Finally, it Should be mentioned that careful handling of diazo- acetonitrile coupled with the available information in the literature made it possible to avoid violent eXplosions, although a couple of minor ones did take place unexpectedly. Vibrational spectra of Diazoacetonitrile and Its Isotopes Application of elementary Group Theory methods28 shows that all 12 vibrations of diazoacetonitrile (CS symmetry) are IR and Raman active. 55 Of those 12 vibrations, 9 belong to the A' Symmetry species representing the in-plane vibrations and 3 belong to the A1 symmetry species repre- senting the out-oféplane vibrations; IR survey spectra of diaancetonitrile in N2 (A) and Ar (B) are shown in Figure 11. A general comparison of the two shows clearly the extensive amount of site splitting caused by the N2 matrix. This is particularly obvious in the region between 300 and 400 cm'1. The single peak at 362 cm"1 in Ar is due to the CH wag out-of—plane vibration. The same peak is multiply split in the N2 matrix. (The shifting towards higher frequency and the broadening with consequent loss of intensity of this peak is due to hydrogen bonding between diazoacetonitrile and the N2 matrix. This is not unusual for compounds of this nature when )50 Peaks marked with an x are due to chloro- isolated in N2 matrices. acetonitrile (see page 16), those with a check are due to H20 and those with a circle are due to C02. Figure 12 Shows the IR survey spectrum of diazoacetonitrile in N2 before and after photolysis. Peaks marked with an arrow are due to HCCN. Peaks due to diazoacetonitrile are easily identified due to the decrease in their intensities upon photoly- sis. Figures 13, 14, 15, and 16 Show IR survey spectra in Ar of NNCHCN (50% 0), NNCHCN (60% 13C), NNCHCN (50% 15N), and NNCHCN (55% 15N) re- spectively. The peaks which were marked with an x in Figure 11 are missing in Figures 14 and 15. This is due to the modification of the synthesis method for diazoacetonitrile as mentioned on page 16, which overcame the problem of Simultaneous production of chloroacetonitrile. 56 monocumz usmcH cw mzwspwzoumomonwo oo mspomom mo .zz mszmwo OON __ a _ o _ o _ _ _ A o _ _ _ . o _..._____z_z_0_|_lj 00m 009 009 OOON comm 000m . _ _ z “759» ”111' ‘_.__—._..z__z——o .. . 57 mzmz_0pozo smpw< cam weowmm xmspmz Nz m cw w—zspzcoomooOwao oo Ezspomqm mo .Nz mszowo A758 9 1“ 00m 000 z z z z # A z z _ _ _ _ z _ _ 4-174141%...4111411-fi 41 olflfilolalw 000_ 009 000m 000m 000m 000m _ 88:26:“: 8:4 F 2.6.2220 828m __ re : , 58 xzsoa: a: :6 em zo :om-zo:ozzo ozzsozzooaoaoaazo ea Ezsoooam :H .m_ 62:82: .4. , OON 000 000. . 00m. 000w 4 4 — 1 a z u — 1 J 3 q _ 000m 000m 000m hfl 59 Xmaoaz a: :8 ea zoo _ soo-zo:ozzo ozooozeooooeoNazo :6 Ezooooom :H an .a_ pesos: 00m #1 q d d 14 - d d 1 q q fl q q 4 q 1 d d H d 000. OOON OOmN 000m 000m :5... 60 stpms L< cm a? Azm — :om-zo:ozzo az_eooeooaoaONaao :6 Ezsooaam z: .m_ oozes: CON ‘1 u d 1 d 000. q d 1 — I 000m I 4 d — iq 1‘ 000m d d — 000m .hx 61 xteoaz a: :8 at zzoz emm-zo:ozmo o_ozozeooooaonazo :6 Ezzoomom zo .o_ deemed CON 1 00m :MCOV P I'll . — 000m 1 d 1 q — 000m u q 41d #14 000m 4 1n . _ 000m H3 62 2220.7 1333.9 1156.4 993.4 1 1019.5 2101.5 2087.6 193.6 623.3 A M5 M 3M94£ frequency 1cm") -—3'ZOa#|+- Figure 17. Raman Spectrum of Diazoacetonitrile (Pure Solid) Finally in Figure 17 the Raman spectrum of diazoacetonitrile in pure Solid fOrm is shown. The observed frequencies from the IR spectra and the Raman spectrum used in the normal coordinate analysis are listed in Tables IX, X, XI, XII, and XIII where they are compared with the results of the n0rmal coordinate analyses. 63 Table VIII. 'Molecular Parameters and Symmetry Coordinates for Diaancetonitrile Bond Lengths (A) . Bond Angles r12 = 1.424 01 = 117° r23 = 1.082 02 = 119.534 r14 = 1.165 83 = 123.466 r26 = 1.280 r56 = 1.132 —A—I All S1 = Ar12 S11 7 Acb1 S2 = Ar23 512 = AC12 S3 = Ar14 313 = A1 S4 = Ar26 $5 7 Ar56 S6* = (l/V3)(AO]+AOZ+AO3) S7 = (.1/V6)(21Ie2 - A01 -AO3) $8 = (1/V’2")(Ae1 - 003) $9 = Aw S10 = Aw *redundant symmetry coordinate Normal Cobrdinate Analysis of Diazoacetonitrile The molecular parameters and symmetry coordinates used for the force constant calculations are given in Table VIII. The internal co- ordinates, in-plane and out-of—plane,are shown in Figure 18. 64 Y X Z All 2 [pi (it y "xi/1:113 X «N, \\S§?//‘ (5’ 0!;5 xy J;\*—" 5 N Figure 18. Internal Coordinates of Diazoacetonitrile (A' in-plane and A" out—of—plane) Initial force constants were transferred from the molecules CH2N22 and ,HCSCCHN2.3 The results of the calculations are shown in Tables IX, X, XI, XII, XIII and XIVI Fifteen f0rce constants were used to fit 35 experimentally observed frequencies. 65 Table IX. Normal Coordinate Analysis of Diazoacetonitrile (NNCHCN) Frequency (cm-1) Primary Contributors 0 Obs. Cal. A0 PED*‘ AI 0] 3118.0 3114.9 (-3.1) CH(99) 02 2228.0 2229.2 (+1.2) C3N(86) CC(12) 03 2102.0 2102.2 (+0.2) NN(94) C=N(15) CN,NN(-10) 04 1349.0 1349.0 (0.0) CHwag(64) C=N(26) CN,CHwag(-18) 05 1162.0 1162.0 (0.0) C=N(43) CHwag(25) C=N,CHwag(l4) NN(6) v6 ----- 907.9 ---- CC(41) CNN(15) CC=N(15) CHwag(l3) 07 620.0 615.4 (-4.6) CNN(60) CC(22) v8 ----- 433.1 ---— CCEN(68) CC=N(15) CNN(12) 09 164.0** 163.8 (-0.2) CC=N(60) CCEN(27) CNN(9) All V10 ----- 719.1 ---- CNN(79) CHwag(16) V1] 488.0 488.4 (+0.4) CCEN(58) CHwag(22) CNN(17) 012 362.0 362.3. (+0.3) CHwag(60) CCEN(35) *Potential Energy Distribution **Raman frequency, N2 matrix Note that on this and the following Tables the notation in the column under PED is as follows: AB= AB stretch ABC= ABC bend . ‘ AB,BC= interaction between AB and BC. 66 Table X. Normal Coordinate Analysis of Diazoacetonitrile (NNCDCN) Frequency'(cm71) Primary Contributors v Obs. Cal. (fl Av PED* 0] 2295.5 2314.1 (+18.6) CD(89) CEN(5) v2 2224.0 2223.7 (-0.3) CEN(81) CC(lO) CD(6) 03 2100.5 2097.4 (-3.1) NN(93) C=N(13) C=N,NN(-9) 94 1286.0 1285.4 (-0.6) C=N(51) CC(26) CDwag(21) CN,CDwag(-l4) 05 ----- 989.2 ---- C=N(26) CDwag(22) V6 ----- 793.3 ---- CDwag(57) CC(25) v7 ————— 615.0 ---- CNN(60) CC(21) o8 ..... 426.9 ---- CCEN(66) CC=N(14) CNN(12) o9 ----- 162.7 ---- CC=N(61) CC5N(27) All 010 ----- 713.2 ---- CNN(83) CDwag(12) 0]] 477.0 476.8 (-0.2) CC3N(75) CNN(12) CDwag(11) V12 292.0 291.6 (-0.4) CDwag(76) CCEN(18) *Potential Energy Distribution 67 Table XI. ,Normal Coordinate Anaysis of Diazoacetonitrile (NNCH13CN) Frequency (cm- Primary Contributors 0 Obs. Cal. Av PED* A' 1 3118.0 3114.9 (-3.1) CH(99) 2 2178.0 2176.0 (-2.0) C5N(86) CC(12) 3 2097.0 2102.0 (+5.0) NN(93) C=N(14) CN,NN(~10) 4 1349.0 1348.7 (-0.3) CHwag(64) C=N(26) CHwag,CN(-l8) 5 1162.0 1161.9 (-0.1) C=N(43) CHwag(25) CHwag,CN(l4) 6 ----- 905.1 ---- CC(41) CNN(15) 7 612.0 613.2 (+1.2) CNN(61) CC(22) 8 ----- 421.4 ---- CCEN(68) CC=N(15) CNN(10) 9 ----- 163.4 ---- CC=N(60) CC5N(28) CNN(10) An 10 ----- 718.4 ---- CNN(79) CHwag(16) 11 479.0 479.7 (+0.7) CCEN(55) CHwag(26) CNN(17) 12 359.0 359.0 (0.0) CHwag(57) CC§N(39) *Potential Energy Distribution 68 Table XII. Normal Coordinate Analysis of Diazdacetonitrile‘(NNCHC15N) Frequency (cm-1) Primary Contributors 0 Obs. Cal. Av PED* A' 01 3118.0 3114.9 (-3.1) CH(99) V2 2204.0 2202.7 (-1 3) C3N(85) CC(13) V3 2099.0 2102.1 (+3.0) NN(94) C=N(14) CN,NN(-10) V4 1347.0 1346.7 (-0.3) CHwag(65) C=N(26) CN,CHwag(-18) V5 1162.0 1161.7 (-0.3) C=N(43) CHwag(25) CN,CHwag(14) V6 ----- 903.6 ---- CC(41) CNN(15) CC=N(15) V7 ----- 612.7 ---- CNN(60) CC(22) V8 ----- 431.3 ---- CCEN(67) CC=N(15) CNN(12) V9 ----- 161.7 ---- CC=N(60) CCEN(28) CNN(10) Au V10 ----- 719.0 -—-- CNN(79) CHwag(16) V1] 488.0 486.9 (-1.1) CCEN(58) CHwag(23) CNN(17) V12 362.0 361.0 (-1.0) CHwag(59) CC;N(36) , *Potential Energy Distribution 69 Table XIII. Normal Coordinate Analysis of Diazoacetonitrile'(15NNCHCN) Frequencies (cm-1) Primary C0ntributors v ..... Obs. .,..... Cal. Av PED* A' 01 3118.0 3114.9 (-3 1) CH(99) 02 2228.0 2229.2 (+1.2) CEN(86) CC(12) 03 2079.0 2073.6 (—5.4) NN(93) C=N(16) CN,NN(—10) 04 1347.0 1346.8 (-0.2) CHwag(65) C=N(25) CC(18) CN,CHwag(-l7) 95 ----- 1154.9 ---- C=N(42) CHwag(24) CN,CHwag(l4) V6 . ----- 906.8 --—— CC(41) CNN(15) CC=N(15) v7 ----- 612.0 ---- CNN(60) CC(22) v8 ————— 431.9 ---- CCEN(68) CC=N(14) CNN(12) 09 ----- 162.0 ---- CC=N(60) CC3N(27) CNN(10) A11 '010 ------ 716.1 ---- CNN(78) CHwag(l7) 011 488.0 487.3 (-O.7) CC5N(59) CHwag(22) CNN(18) 012 362.0 362.0 (0.0) CHwag(60) CC3N(34) , *Potential Energy Distribution 70 .< xwucmaa< mom .mmwucwmosmucz mom mommzpcmsmq cm memnE:z A<.c>ueo pcmumcoo mosoo cowpomsmpcw mcvvcwnimcwocmn ieo Acmeso “zoomeoo ooooe eooooesooee ozoeeoo-ozw:oooeoa -_: A<\czvsv pcwumcou mosey cowpomgmocm mcwzopmsuwimcwzopmspm no A<.c>nev pampmcoo wusoo mcwccmn mcmzoieoipzo 1.: A<.czuso pcmpmcoo mosoo mewucma mcmzaicw 1: a A<\czcev pcmpmcou mosoo mazupmspm iz a a zooo.oo ooo.o n az:o zzw m n zooo.oo zmz.o n as:w zNo.oo oma.o u ziow zoo—.oo Noo.o u zzw .: . n mm: . u zeNo.oo Nom._ 1 zz.z o: Amoo.oo oam.o- u :o z o zmzo.oo omo.o " zzo: zom_.oo zem.o u zwoo: zooo.oo woe.o . znoo: o , . Amoo.oo ooe.o n as:o: zzmo.oo Neo.e_ u zzz zeoo.oo oom.: n z oz Ammo.oo ooo.o_ u zmoz zeoo.oo _om.m u :o: z_o_.oo om_.o n ooz _: mzzspwcopmomonwo Lem mwcmpmcoo mosoo .>Hx ozone 71 Band Assignment of.the.0bserved Fundamentals of Diaancetonitrile 1 1 1 2280 cm— 1 3100 cm"1 2310 cm' 3130 cm I . A' Class “1 1 in NNCHCN Shifts to 2295.5 cm”1 in The band observed at 3118 cm- NNCDCN. (Both of these bands are shown above.) No detectable shift could be observed in the other isotopes and indeed, none is predicted . by the normal coordinate analysis. The frequency range within which this band is observed, coupled with the large deuterium shift and the results of the normal coordinate analysis leave no doubt that this band is the (0]) CH stretching fundamental. This frequency is very close —1)3 \ to that of HCECCHN2 (3100 cm and that of CH2N2(3132cm-1, taken as an average of the Symmetric and anti-symmetric stretches of the‘CH2 group).2 The large difference between the predicted and observed fre- quency in NNCDCN is not unusual for molecules of this~nature.2’3 72 v] is very weak in.the Raman.where it is observed at 3100 cm.1 (pure solid): 2210 cm‘1 2240 cm'] ' v2 V2 1 The band observed at 2228 cm' in NNCHCN, based on its isotopic frequency pattern and the normal coordiante analysis results, is assigned to the (92) CEN stretching fundamental. Group frequency consideration55] support this assignment. The band and its deuterated analog are shown above. 02 is observed at 2220.7 cm.-1 1 in the Raman spectrum of the pure solid and at 2230 cm' in the Raman spectrum of diazoacetonitrile in a N2 matrix. Contrary to the IR (where it is of medium intensity) it is the strongest band in Raman!! 73 mzwsuw:OpoumONozo mo Am>v zmpcmsmuczo mcwzoumzpm 22 .m_ mszmzo .a .\.mn.zozozuo Soaomizooozzo n3 msom LE N08 7:; «EN fixonzuuzozzo EBNQN mo. $08-2wxozzo onmmON .mahmom :Eomofi ONE ONE 0N_N 000m 000m 000m 000m “20:022. n5 New 000m 0N_N 74 V3 The band observed at 2102 cm‘1 in NNCHCN, for the same reasons stated above, is assigned to the (v3)-N5N stretching fundamental. This value is close to that of HCECCHN 2069 cm-1) and almost identical to 2 ( that of CHZN2 (2102 cm'1). Perhaps the most interesting aspect of this band is the fact that it is split in the spectra of diaancetonitrile 13 15 with 60% C and 55% N in the CEN group. The splitting may be inter- preted as the result of a rather strong interaction between the CEN and 55 If this NN bonds, something that is implied by the MW eXperiment. was the case one would expect that the band due to the CEN stretch would appear Split in the spectrum of diazoacetonitrile with 55% 15N in the NN group. No such splitting was observed however! This band and its counterparts are Shown in Figure 19. 03 although the strongest peak in the IR is very weak in the Raman. 1 It is observed at 2109.5 cm- in the Raman spectrum of the pure solid. 1360 cm'] 1340 cm'1 1 1 V4 The band observed at 1349 cm-1 in NNCHCN is assigned to the (v4) 75 CH in plane wag vibration. This assinnment is based mainly on the normal coordinate analysis resu1ts, which also Show that it is very strongly coupled with the C=N stretching fundamental. It is interesting to note that this band shifts by an equal amount (2 cm'1) in both NNCHC15 ‘5 N and NNCHCN isotopes, something that may be explained by the symmetry of the molecule. (The two terminal nitrogen atoms are almost equally far away from the hydrogen atom.) This assignment is also sup- ported by the assignment of the 1358 cm'] band observed in the spectrum of HCECCHNZ. v4 and its counterpart in the spectrum of diazoaceto- nitrile (55% 15N in the NN group) are shown on the previous page. This band is strong in both IR and Raman. In the latter it is observed at 1 1333.9 cm" (pure solid). 1170 cm-1 1150 cm V5 The band observed at 1162 cm'1 in NNCHCN is assigned to the (v 5) C=N stretching fundamental. This assignment finds support in the nor- mal coordinate analysis results, which also show a significant contribu- tion to this frequency from the CH wag, and the Similar assignment of 1 band observed in HCECCHN and the 1170 cm" band observed the 1165 cm' 2 76 in CH2N2. This band is pictured onpage.75..'95 is.weak in.the IR but strong in the Raman where it is observed‘at.1156.4‘Cm71 (pure.solid). V6 No band attributable to V6 was observed in the matrix Spectra. 1 Normal coordinate analysis results predict it to be at 907.9 cm" and assigns it to the CC stretch. This vibration is expected to be very 51 1 weak in the IR but rather strong in the Raman. The band at 993.4 cm' observed in the spectrum of the pure solid is a likely candidate for this band, although lies quite far away from the predicted frequency. 600 cm- 640 cm-1 ' v 7 The band observed at 620 cm-1 in NNCHCN and shifted to 612 cm'] in 13 NNCH CN is assigned to the (v7) CNN in-plane bending fundamental. This band is very weak in the IR but of medium intensity in the Raman '(623 3 cm“). It lies considerably higher than that of CH 421 cm*‘) 2N2( while no useful correlation can be made with that of HC'='CCHN2 due to the extensive internal mode coupling in the latter.This assignment is supported by the normal coordinate analysis results which show a signi- ficant contribution of the CC stretch in this frequency. This band (v)7 is pictured above. 77 ‘8 No band which' could be assigned tov8 was observed in the matrix or pure solid spectra. The nOrmal coordinate analysis predicts V8 to 1 be at 433.1 cm- and assigns it to CCEN in-plane bending vibration. 1 172 cm- 157 cm- V9 The band observed at 164 cm.1 in the Raman spectrum of NNCHCN in a N2 matrix is assigned to the (09) CC=N bending vibration. This assign— ment which is supported by the normal coordinate analysis results and the similar assignment of the 168 cm-1 band in HCENCNNZ, is in agree- ment with the results of the MW experiments.65 In the latter, based on relative intensity measurements, a low-frequency deformation mode (predicted to be at 150 :30 cm']) was tentatively assigned to the CC=N bending vibration. The 164 cm'1 band is pictured near the top of the 1 in the Raman spectrum of the page. 09 is also observed at 193.6 cm- pure solid. Its IR analog could not be observed due to the fact that it lies outside the frequency range covered by our IR spectrophotometer. v9 along with the v2 are the only two fundamentals which could be obser- ved in Raman matrix spectra. 78 ozzeowcooaoaONaoo :6 z__>o zzo: oeaza-eo-ozo zmoo .om 6238:: 284630012010sz 9 n8 moo AZo\oomiZUIoz 20 n— n— 35 8.6 08 084 008 08.6 8 630012075220 2 n. .66 a? 00v : .66 mm... 000 330m... 2000220 n5 moo AZUIQZZV LEE: 7:6 moo 006 r. oom coo coo 79 A" Class V10 No band assignable to 010 was observed in the matrix or pure solid spectra. 010 is predicted by the normal coordinate analysis at 719.1 cm'1 and assigned to the CNN out-of—plane bending vibration. This frequency like its in-plane analog, lies considerably higher than that of CH N (564 cm-]) and that of HCECCHN 475 cm_]). 2 2 2 ( V11 The band observed at 488 cm-1 in NNCHCN, based on its isotopic fre- quency pattern and the normal coordinate analysis results, is assigned to the (v11) CCEN out-of—plane bending vibration. 011 and its counter- parts in all five isotopes of diazoacetonitrile are Shown in Figure 20. Normal Coordinate analysis shows a significant contribution to this fre- quency from the CH out-of—plane wag vibration. 011 is strong in the IR but weak in the Raman where it is observed at 507.5 cm”? (pure solid). It is interesting to note that 011 lies very close to the frequency of the CCN bend in cyanoacetylene (500 cm'1).44 v12 The band observed at 362 cm'1 in NNCHCN is shifted to 292 cm‘1 in NNCDCN. This large deuterium Shift alone implies that this is a hydro- gen motion. Since the only hydrogen motion remaining is the CH out-of- plane wag vibration, it is assigned as such (912). This assignment is supported by the n0rmal coordinate analysis results and by a similar 1 assignment made for the 406 cm" band in CH2N2. The analogous band in 8O HCECCHN2 has not been.observed. 012 and itS‘13C.counterpart are pic- tured just below. I\ J ' 335 cm- 370 om‘1 / 1 The Raman analog of this fundamental is probably the 394.8 cm'] band (observed in the Raman spectrum of the pure solid). There is how- ever a remote possibilitythat the latter is the Raman analog of the V8 fundamental (Predicted at 433.1 cm'1 but not seen in matrix spectra). A depolarization ratio measurement of the 394.8cm‘1band would probably solve the assignment problem; Unfortunately measurements of this nature cannot be made on pure solid sampleS(unleSs they are single crystals) Since these samples scramble polarizations thoroughly.66 81 Other Bands 1 The band observed at 728 cm” in NNCHCN is assigned as the overtone of 012 (362 x 2 = 724). Its isotopic frequency pattern supports this assignment fully (582 cm" in NNCDCN and 720 cm" inNNCH‘3CN). 1 The band observed at 852'cm' in NNCHCN is assigned to the combi- nation of 011 and 012 (488 + 362 = 850). This assignment is also sup- 1 l ported by the isotopic frequency pattern (770 cm- in NNCDCN, 840 cm- in NNCH13 CN). The band observed at 3062 cm"1 is assigned to the CH stretch of a polymer (perhaps a dimer) of NNCHCN (formed by hydrogen bonding). This assignment is supported by the large deuterium shift (792 cm") and the intensity changes of the band with variation of the M/G ratio. Hydrogen bonding is known to occur in CH2N2 and is expected to be even more pro- nounced in NNCHCN due to the presence of the CEN group which Should in- crease the acidic character of the hydrogen atom. 1 Two bands (504 cm' and 367 cm'1) observed as shoulders of 011 and 012 respectively are also assigned to a polymer of NNCHCN. These assignments are supported by IR spectra of low M/G ratios where the only bands observed in those frequency regions correspond to those high frequency shoulders. The band at 2154 cm_] which appears sometimes as a shoulder of the NEN stretch and disappears upon photolysis could not be assigned to any- 1 thing specific. The 1019.5 cm- band observed in the Raman spectrum of pure solid is assigned to a combination of the 623.3 cm'] and 394.8 cm'1 bands (623 3 + 394 8 = 1018.1). Two other bands observed in the same 1 spectrum (2087.6 cm- and 525.1 cm']) cannot be assigned to anything 82 specific at this time. ' The vibrationa1 Potentiai‘runation at oiaaoaoetonitriie The force constant of the CC stretch is calculated at 5.13 mdyn/A and is very similar to that of HCECCHNé (5.0 mdyn/A). Its value is a little larger that that of a CC Single bond which is in accordance with the CC bond length obtained by the microwave spectra. The force constant of the CEN stretch is calculated at 16.99 mdyn/A and is a little smaller than that of a typical CN triple bond, in accor- dance with the MW results. The force constant of the NEN stretch is cal- culated at 17.642 mdyn/A, which is larger than those of CHZN2 (16.89 mdyn/A) and HCE'CCHN2 (14.14 mdyn/A). This correlates well with the fact that the NEN bond is shorter in diazoacetonitrile than in CHZNZ' The force constant of the C=N stretch is calculated at 7.26 mdyn/A and is considerably smaller than that of CHZN2 (8.34 mdyn/A) which is in disagreement with the relative order of the bond lengths in the two mol- ecules.65 The force constant for the CH in-plane wag is calculated to 0.48 O O mdyn°A and is very close to that of HCECCHN (.456 mdyn-A). The force 2 constant for the CC=N bend is calculated at 0.408 mdynoA and is much Tower than that of HCECCHN2 (0.905 mdyn'A). The f0rce constants of the CNN in-plane (.836 mdyn'A) and out-of- plane (.882 mdyn-A) bends are much larger than those of CHéNé (0.477 and .533 mdyn-A) and HCECCHN2 (.418 and .324 mdyn°A). The Same situation is encountered for the CH out-ofeplane wag f0rce constant which is 0.121 mdyn-A for diazoacetonitrile and 0.045 mdyn-A 83 forCHZNZ. .The value for diazoacetonitrile is however closer to that of ketene(0.086 mdyn-A),Iwhilestill.much lower than that of ethylene (0.23 mdynQA). _ The f0rce constants of the CCEN in-plane (0.341 mdyn‘A) and out-of- plane(0.430 mdyn-A) bends are not atypical of those of aliphatic ni- triles, while the interaction force constants FC=N,NN (1.502 mdyn/A), FC=N,CHwag (-0.396 mdyn) and FCNN,CHwag (0.008 mdyn-A) for diazoaceto- nitrile are very similar to those of CH2N2 (1.23, -0.467 and 0.006 respectively). Although several resonance structures can be written for CHZN2 and NNCHCN, there is one in particular which can be written for the latter but not for the former: Fl 1 /‘\ 1+ \ This structure weakens the CN bond of the CNN group and strengthens the (I —/ fl bl CC bond. Thus it could be the reason for the difference between the KC=N force constants of the two molecules. This resonance structure can also explain the value of the KCC force constant which is a little larger than that of a CC Single bond. The Slightly low value of the force constant for the CEN bond can also be explained by this resonance Structure. 84 Finally, it should be mentioned thatonly ll:out of the.12 di- Vagonal force constants could be varied SimUltaneously for the normal coordinate analysis program to converge. This is not unexpected.when the experimental information available is not Complete (Which is the case here, due to the 3 missing frequencies). In Such a case the max- imum amount of information, with regard to the vibrational potential function, that can be extracted from the experimental information con- sists of the following: a) the force constants that can be varied Sim- ultaneously and (b) the changes in the values of these force constants when the force constants whose values are assumed are changed. This al- lows one to see how accurate the values of the variable force constants are. When the change is large, with respect to the value of the force constant itself, it implies that the latter is not very accurate and when the change is small, the implication is that it is accurate. The Table on the following page shows the accuracy of variable force constants as obtained from the normal coordinate analysis. Thus as one can see, the values of the following f0rce constants should not , and F" L The CNN,CHwag' inaccuracy of the HCC=N force constant must be the reason for the large be considered very accurate HCC=N’ HCCEN’ HCHwag discrepancy between the latter and its analog in thelHCECCHN2 molecule mentioned earlier in this section. 85 Table XV. - P Matrix Elements of Diazoacetonitrile Force Constant P Element f Value 3f/3HCNN KCC 5.130 -0.065 KCH 5.281 0.017 KCEN 16.990 -0.008 KC=N 7.260 —0.016 KNN 17.642 -0.001 HCHwag 0.480 0.025 HCC=N 0.408 -0.288 HCCEN 0.341 -0.049 HCNN 0.882 -0.011 HCCEN 0.430 -0.317 HCHwag 0.121 -0.087 FC=N,CHwag -0.396 -0.012 FC=N,NN 1.502 -0.004 F" 0.008 0.08 CNN,CHwag 86 COnClUSion Nine of.the twelve fundamental frequencies of diaancetonitrile were observed and properly assigned, based on the isotopic frequency pattern, group frequency cencept and the results of normal coordinate analyses. The three miSsing frequencies are also predicted by the latter and accordingly assigned. The valence force field is surprisingly accurate, despite the 3 missing fundamentals, as was shown in the pre- vious section. Overall, the results of the vibrational study are in verygood agreement with the results of the MW experiment. The f0rce constants of diazoacetonitrile, in general, correlate well with those of HC'='CCHN2 and CH2N2. Whenever large discrepancies occur, these can be explained by invoking the concept of resonance or the results of the study on the accuracy of the obtained force constants from the normal coordinate anal- ysis. Despite the difference between the refined force constants for the three molecules, they are considered transferrable from the point of view that the initial set of calculated frequencies of diazoaceto- nitrile based on the force constants of CHZN2 and HCECCHNZ correlates very well with its frequencies which are observed experimentally. The group frequency concept seems to hold for the stretching mo- tions in the CNN group. However, due to the intensive mixing of the internal valence modes, the group frequency c0ncept does not seem to hold for the bending motions asSociated with the CNN group. Finally, it should be mentioned that, if a better estimate of the f0rce field of diazoacetonitrile is to be obtained, further investiga- tion for the determination of the missing frequencies is necessary. 87 The Electronic.Aborption Spectrum of Diaancetonitrile AlthoUgh irrelevant to the vibrational work ondiazdacetonitrile, it seems appropriate to at least report the data obtained regarding the electronic spectrum of diazoacetonitrile. These data were obtained in the process of trying to observe the electronic spectrum of HCCN on one hand and establish the wavelength range of the photolyzing radiation on the other. The electronic Spectrum of diazoacetonitrile in an Ar matrix at 15 K was scanned between 185 nm EHKIBOO nm.‘ There were no bands obser- ved in the visible part of the spectrum. One very strong band was ob- served in the UV part and is shown in Figure 10 on page 47 (dotted line). This band (2280 A) disappears with photolyzing radiation of X> 3500 A. The electronic spectrum of diazoacetonitrile in CH2C12 at room tem: perature was also scanned between 185 and 800 nm. Two bands were ob- served. One was very broad and weak covering the wavelength range from 'v4800 A to ~3000 A and seemingly peaked at ~4050 A. The other was a very strong band which began at ~3500 A and went off scale at .02500 A. As the concentration of diazoacetonitrile in CH2C12 was lowered by di- lution, the band in the visible became weaker and finally disappeared while the band in the UV was still off scale. This is the reason why the band in the visible could not be observed in the matrix, since the light scattering becomes a severe problem when the thickness of the ma- trix exceeds a certain point. The electronic spectrum of diaancetonitrile seems to follow the 67 pattern exhibited by other diazo compounds. The weak, broad band in the visible must be the one responsible for the photodisSociation of 88 diazoacetonitrile when it is irradiated with'l> 3,500 A. Indeed the frequency range covered by this band corresponds to energies higher 68 than the disSociation energy of the C=N bond and is certainly outside the frequency range that corresponds to energies close to the disSocia- 69 tion energy predicted for the CH bond. The latter is in agreement with the fact that no other free radicals were observed in the matrix. APPENDIX APPENDIX A The uncertainties listed in parentheses next to the fOrce constants in Tables VII and XIV are obtained from the following equation: where: CZ 33 ~ -l 1/2 NM. 5 (n-m) h) = uncertainty in force constant Kh number of frequencies number of force constants diagonal matrix whose elements give the weights assigned to corresponding observed frequencies Jacobian matrix given by the equation Av=JAK (where Av is the frequency change vector and AK is the force constant change vector) weighted sum of squared deviations defined by (Vobs ' vcalc.)(w)(vobs - Vcalc) column frequency vector row frequency vector 89 REFERENCES 1o. 11. 12. 13. 14. 15 REFERENCES E. Herbst and w. Klemperer, Physics Today, 32, June 1976. C.B. Moore and G.C. Pimentel, J. Chem. Phys.,‘fig, 342 (1964) F.K. Chi and G.E. Leroi, spectrochim. Acta,‘3lA, 1759 (1975). A. Poletti, G. Paliani, M.G. Giorgini, and R. Cataliotti, Spectrochim. Acta,'§lA, 1869 (1975). R. Cataliotti, A. Poletti, G. Paliani, and A. Foffani, Z. Naturforsch., 27b, 875 (1972). R.A. Bernheim, R.J. Kempf, P.W. Humer, and P.S. Skell, J. Chem. Phys.,‘4l, 1156 (1964). R.A. Bernheim, R.J. Kempf, J.V. Gramas, and P.S. Skell, J. Chem. Phys., 43, 196 (1965) R.A. Bernheim, R.J. Kempf, and E.F. Reichenhecher, J. Magn. Resonance, 3, 5 (1970). E. Nasserman, w.A. Yager, and V.J. Kuck, Chem Phys. Letters, Z, 409 (1970). a) A.J. Merer and D.N. Travis, Can. J. Phys., 43, 1795 (1965). b) A.J. Merer and D.N. Travis, Can. J. Phys., 44, 353 (1966). R. Hoffmann, G.D. Zeiss, and G.w. Van Dine, J. Amer. Chem. Soc., 99, 1485 (1968). P.S. Skell, S.J. Valenty, and P.w. Humer, J. Amer. Chem. Soc., 95, 5041 (1973). w.J. Hehre, J.A. P0p1e, w.A. Lathan, L. Radom, E. Nasserman, and Z.R. Nasserman, J. Amer. Chem. Soc.,'g§,4378 (1976). R.R. LUCchese and Henry F. Schaefer III, J. Amer. Chem. Soc., (to be published). . J.F. Harrison (private communication). 90 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 91 E. Whittle, D. A. Bows, and G. C. Pimentel, J. Chem. Phys.,‘22, 1943 (1954). H. P. Broida and A. M. Bass:, Formation and Trapping of Free 'RadiCals, Acad. Press, New York, (1960); (b) A.J. Barnes and H. E. Ha11am, "IR Studies of Matrix ISolated species," Quarterly Reviews Chemical Society, London (1969); p.392.; (c) D.E. Milligan and M.E. Jacox, Physical Chemistry, an Advanced Treatise, Chapt. 5, ' 4, 193 (1971); (d) L. Andrews, Annual Review of PhySical Chemistry, Chapt. 5,'22, 109 (1971); (e) A.J. Barnes, "Applications of Matrix Isolation IR Spectroscopy," Reviews in Analytical Chemistry, 1, 193, (1972); (f) 0.5. Milligan and M.E. Jacox, International Re- view of Science (Spectroscopy), Chapt 1, P. Chemistry Series One, _§, 1 (1972); (g) Vibrational Spectroscopy of Trapped Species, ed. H.E. Hallam, Wiley and Sons, London (1973); (h) A.J. Downs and S.C. Peake, "Molecular Spectroscopy? ed. R.F. Barrow, D.A. Long, and D.J. Millen, (Specialist Periodical Reports), The Chemical Society, London, 1, 523 (1973); (i) S. Cradock and A.J. Hinchcliffe, "Matrix Isolation,” Cambridge University Press, Cambridge (1975); (j) B.M. Chadwick, "Molecular Spectroscopy," ed. R.F. Barrow, D.A. Long, and D.J. Millen, (Specialist Periodical Reports), The Chemical Society, London, 3, 281 (1975). E.B. Wilson, S.C. Decius, and P.C. Cross, Molecular Vibrations, McGraw Hill, New York (1955). D.D. Phillips and W.C. Champion, J. Amer. Chem. Soc., 18, 5452 (1956). G.W. Robinson, J. Mol. Spectrosc., 6, 58 (1961). M.J.S. Dewar and R. Petit, J. Chem. Soc., 2026 (1965). The Sadtler Standard Spectra, Infrared Spectrum #7886, Sadtler Research Laboratories, Inc. (1976). G.P. Groggins, Unit ProCesses in Organic Synthesis, McGraw Hill, New York, 4th ed., (1952). R.E. Rondeau, J. Chem. Eng. Data, 1, 124 (1965). R. Adams and W.D. Langley, "organis Synthesis," ed.H.Gilman, J. Wiley and SOns, Inc., New York, Coll. Vol. 1, 355 (1932). ‘26. 27. 28. 29. ' 3o. ' 31. 32. 33. 34. 35. 36. 37. 38. 39. 4o. 41. 42. '43. 92 W. K. Anslow and H. King, "Organic synthesis,“ ed. .HGilman, J. Wiley and Sons, Inc. , New York, C011. Vol. ,298 (1932). - M. E. Jacox and D. E Milligan, Chem. Phys“, 4, 45 (1974). .. .........-.........._................................--.- ................... F. A. Cotton, Chemical Appl1cat1ons Of Group TheOr , Interscience, New York, 2nd ed. , (1971). J.S. Shirk, J. Chem. Phys., 55,‘3608(1971). F.K. Chi and L. Andrews, J. M61. spectr05c.,'52, 82 (1974). G. Herzberg, Infrared'and Raman SpeCtra, Van Nostrand Reinhold, New York (1945). T. Shimanouchi, Computer Programs for Normal Coordinate Treatment of Polyatomic Molecules, Tokyo: Univ. Tokyo (1968). T. Shimanouchi, "Physical Chemistry: An Advanced Treatise," ed. H. Eyring, D. Henderson, and W. Jost, Academic Press, New York, 4, 233 (1970). T. Shimanouchi and I.Suzuki, J. Chem. Phys., 42, 296 (1965). a) L.A. Woodward, Introduction to the Theory of M01ecular Vibrations and Vibrational Spectroscopy, Oxford University Press, London (1972). 6) P.)Grans, Vibrating Molecu1es, ed. Chapman and Hall, London 1971 c) D. Steele, Theory of Vibrational Spectrosc0py, ed. Saunders, London (1971). L.H. Jones, Inorganic VibrationalISpeCtroscopy, Vol. 1, Dekker, New York (1971). W.H. Fletcher and W.T. Thomson, J. Mol. Spectrosc., 25, 240 (1968). ‘6. Herzberg and D.N. Travis, Can. J. Phys., 42, 1658 (1964). F.K. Chi, Ph.D. Thesis, Michigan State University (1972). M.E. Jacox and D.E. Milligan, J. Chem. Phys.,‘43, 3734 (1965). D.E. Milligan and M.E. Jacox, J. Chem. Physt,‘4§, 3149 (1965). D.E. Milligan and M.E.-Jacox, J. Chem. Phys.,'45,f1387 (1966). 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 93 %.C.iurre11, W.D. Jones, and A. Maki, J. Chem. Phys., 25,.1544 1957 Sons, Inc. , New York (1972). T.B. Freedman and E.R. Nixon, Spectrochim. Acta, 28A, 1375 (1972). P. D. Mallinson and L. Nemes, J. Mol. SpectrOSc. , 59, 470 (1976). G. Zerbi and M. Gussoni, J. Chem. Phys. , 41_, 456 (1964). R. Yamadera and S.Krimm, Spectrochim.Acta,‘244, 1677 (1968). M. E. Jacox and D. E. Milligan, J. Chem). Phys.,‘40, 2457 (1964). (1975). J.W. Linnett and M.F. Hoare, Trans. Faraday Soc., 45, 844 (1949). R.N. Dixon, Phil. Trans. Roy. Soc., (London), 4252, 165 (1960). R.N. Dixon, Can. J. Phys., 55, 10 (1960). G. Herzberg, The Spectra and Structure ofSimple Free Radicals, Cornell University Press, Ithaca (1971). N.M. €therton, Electron Spin ReSOnance, Halsted Press, New York 1973 . R.A. Bernheim, T. Adl, H.W. Bernard, A.Songco, P.S. Wang, R. Wang, L.S. Wood, and P.S. Skell, J. Chem. Phys., 54, 2747 (1976). R.N. Dixon, Can. J. Phys., 52, 1171 (1959). C. Thomson, J. Chem. Phys., 55, 841 (1973). E. Wasserman, L. Barash, and W.A. Yager, J. Amer. Chem. Soc., 51 2075 (1965). S. H. Glarum, J. Chem. Phys-‘ , 39, 3141 (1963). G.R. Williams, Chem. Phys. Letters, 25, 602 (1974). T. Curtius, Ber. Deut. Chem. Ges.,'5l, Part II, 2489.(1898). T. Dominh, 0.P. Strausz, and H.E. Gunning, Tetrahedron Letters, 5237 (1968). C.C. Costain and J. Yarwood, J. Chem. Phys., 45, 1961 (1966). "66. 67. 68. 69. M.C. R.K. A.H. H.J. 94 Tobin,L5§eriRam§ntSpeCtr55copy, Wiley-Interscience, (1971). Brinton.and D.H. Volman, J. Chem. Phys.,'45, 1394 (1951). Laufer and H. Okabe, J. Amer. Chem. Soc., 55, 4137 (1971). Bernstein, Spectrochim. Acta, 18_161 (1962). , :7 7 ~ , . _ a - E ”News .. RAP , 'i x ' ' . . ' 3 9 030710317 7 : . . ~11 , . 1 .. , 1...,3114“. ., H . .., 1..., t,\ a v. 1:0:1‘” ; .331 15:. ‘I":"I'v'5(\‘\~"' M a ~14 V‘V: V\'\\N‘\‘.";V xv we a \V-mV “w“ .. I; “‘V‘ w ”535-1.?“ " V . .. l - ‘., ‘3:3-‘.-“‘ .1 1‘,- “.4 ‘I‘a‘a-i“ .- . , . . ‘ ‘Ils V\V'\-\‘:fl :‘VW‘V \V\\ \V a Vewa»: ,a...,-,w‘tn- A Vi“ .. _ ‘C :‘HN‘ -u~-i~u:«s--\~V\-i ~- 4‘ \HHW _ ‘kV‘ v-VV-i- ‘4‘: a: \-- \‘V'nu‘ ‘VVV m< » \ I h ,4 '\~‘\ irreVV Hm ~ I . k ' “VV1‘* ~“I1‘::Vt\‘:‘l\‘li I Y . “I Vm‘mw‘;‘ ‘V V“ V'C' \‘\ Jig-:74 “I ‘ _ , ,5 “‘ , *‘&‘£‘ «a "3.. ’VL‘SQF‘W :12" 3:" ““ . - -. 'I‘;“~'1"‘V\" ‘ v . -| - ‘ _; 1. J.‘3“3:.-. “€343,133. 9. a... n“ 74 the“. , ~ 'v «:1. , «(‘3‘ «a: ”L writ . r.» VV ‘5‘» ‘W‘ mt.» \VI': “u a \a a a \ aim-“M. “45‘ SAW...“ 1‘“ lv ;\~i Ino-w.‘\. k; >(::y ..\ . ‘ \ V-ve 13"“I‘KW‘. Vs u‘u“ i -,... \ a . 1 , . ‘ ‘ ~ "at“ "w w ,1»... .-:L‘.a5.‘-t.., 1,01,”: \ ; h- M. , ; X wax“ “3:: ‘3 K, VIC-\il‘vV“: . 13.3.: C3114“ 1N" +31)“. “:1; \ 1': U . y, . 3 ._. . p1,.‘.,.. ‘4 ...-, n. \w‘. mm :‘2-3 "‘ , -‘- - x “3' ‘ ‘ ' ‘ ‘.‘.'.‘.‘,- . -' $5.3? :fiaxnxazaa 31:3...” ‘3 “r a ‘ 3:“: ‘~?““ ‘1 =- ~ - .‘ . N.‘.‘.:;1r.x,u\-‘}m “‘ " “ ix“ '9: 4“ .‘k}‘ ”at“: :‘Ii' ‘51; I :“E‘ 5‘.” : l:\.v‘::\\k'll~‘n ‘ -:\ \ 61"“; r‘w "‘:,“,‘,1;:‘i':rcs““\" «‘ ' ' ‘ ‘ 1,‘ ‘1 1 .‘ , . ~, \1.‘ 1V - . . m:§?::&‘¢: ‘::~:4£?r-«1“~*“ -'-> ”'“35 fi.‘ .. "QYA N ‘ .‘ .\ . v0 \ . . . “A. u m 1:: 3‘s.-.r:w::"~ mi. .4; q... :.~. ~;\ ~\« (.33.; _ .. is: . we: u-I. ‘ :.;:.tm,.‘:~.;::..\‘:m':‘a‘, u, “at? “$1719; _ “13:... .,M,,..:.“,,1er\1r”‘a“‘-E «3 ‘75? 3332511., :cv .. I" .\ u 1; “tux“‘mw‘cc‘m ‘ ., . 3.3"» 3,, ~,_.:(_“:, 5.1;; ., W. ‘. ' N \ I a ‘ -\ u‘ I ~ u : m":“13é~m«h‘?“fi%m%Wufis‘s‘kflfi‘qmfin‘i‘w‘ ' ‘ . ‘ - w 3" “"‘ w ‘ L "““'S. :35}: "" ‘ V“; ‘| 54mm 3; ”A“ . .. . amp-:2 «“m 1 «V \r . u i. :‘4 lanai-{Tun :im $335525.“ 4" ‘Mw‘l: ‘1‘.» \u 7"“ \r, H . ‘ .1 ”NT, 6 ‘\_| .4 n...“ . -u 3.32.“,31Mmm ‘L :"-“-‘~\ 9 1 :“\,~‘;\¥;t .~\\1\\*\ ' .. - {war 31.1453,» - 5 , 7'1““ « «can fim‘t‘ 0 331th n ‘ ‘ .~ 'V 1.1 ‘1 \ ~ . . Ham‘s: vs ‘ A\ ‘9‘ Eth‘muy .;.n‘1;.\""'»~-..\\1ur .- “1'59““ “3"» - . x34.“ ,..,,,,.1.. W, 'K‘ mixer: “=3“ 113‘ r; «a... is:- .. . .1 x. 28m." 1... ~31 ,. ..1..-\ 1.. . . ,: {4.311 n. i x; “a “.311, ‘ . (515:1:531‘ 1 2‘“ ,' ‘\ ’1 ; V an“ em ‘Vilul\1-\-IIH\ 1 , n a u \‘ wr.‘ 93.x, 1. ._ a; _ “, x . 11 MC; “amine; \' ”‘4‘ 3.5:: 1‘ ; ; i- ‘11:”? 1' . 4 ‘5‘ ' ' 0‘ \ ID > I r I“ ’ ‘ ‘- 7 ‘7: V5» " 1"" ~-~\~~ ”‘34“ - 1 1' H" \x"' “ « 1... \‘1\"llin‘ ‘ “'1 :17“. -£~ u‘ \‘$§17§: a. ., ..,.\ , :H..1y. 7 I \ V «an :9 u. H r V. n «VV-‘m‘n fiw‘x’fl‘ ‘ \‘m‘Km‘MQVVV1‘:‘\:V‘V v 4 ‘ ; ‘7 a: "§VI'\VV {ha-smug l ‘N 2‘ ‘ . n 5 ;mmm‘un 3:31“ (ilk: nu": HS...» ‘11.“. 4 . . {mummy-”Mum...