BEHAVIOUR OF PARAFFIN HYDROCARBONS ON ELECTRON IMPACT SYNTHESIS AND MASS SPECTRA OF SOME DEUTERATED PARAFFIN HYDROCARBONS BY D. 0. SCHISSLER,* $ S. 0. THOMPSON? AND JOHN TURKEVICH * I- Received 26th February, 1951 A method of preparation of mom-, di- and tri-deutero paraffin hydrocarbons containing the deuterium on one carbon atom is described. The mass spectra are given for all the deuterated methanes, four deuterated ethanes and two deuterated propanes. A quantitative explanation is given for the mass spectra of the deuterated methanes. The study of the products produced on electron bombardment of various deuterated hydrocarbons in a mass spectrometer is useful for several reasons. It permits the determination of both the number and type of the various isotopic molecules present in a given system thus affording a means of using stable isotopes for the study of homogeneous and heterogeneous reactions.' Furthermore, the decomposition of a hydrocarbon into various ionic fragments on electron bombardment in the ionization chamber of a mass spectrometer is an example of a truly unimolecular process.A study of this process in a number of deuterated compounds will undoubtedly throw light on the primary act of a molec- ular decomposition. For this purpose deuterated hydrocarbons of known structure and purity must be synthesized and examined in a mass spectro- meter. In a previous publication from the Frick Chemical Laboratory,B the synthesis and mass spectra of monodeuteromethane, monodeutero- ethane and two monodeuteropropanes were reported. In the present communication, we wish to report a new method of introducing one, two, or three deuterium atoms on a given aliphatic carbon atom as exemplified in the synthesis of mono- di- and tri-deuteromethanes, mono-, asym- metrical di-, and asymmetrical tri-deuteroethanes, mono-deuteropropane, and 2 : 2-dideuteropropane.The mass spectra of these compounds are presented and a quantitative treatment of the deuteromethane spectra is given. Experimental The introduction of deuterium atoms into hydrocarbons has been previously effected by the use of Grignard reagents and Dz0,2* 3* 4 D 6 and by the reduction of alkyl halides either by lithium aluminium deuteride and D20,6 or zinc and C,H,OD.' The Grignard reaction is difficult to manipulate because of the * Chemistry Department, Princeton University.t Chemistry Department, Brookhaven National Laboratory. 3 Ethyl Corporation Fellow, 1950-51. Turkevich, Bonner, Schissler and Irsa, Faraday SOG. Discussion, 1950, 8. Turkevich, Friedman, Soloman and Wrightson, J . Amer. Chem. SOC., 1948, Wagner and Stevenson, J . Amer. Chem. Soc., 1950, 72, 5785. Evans, Bauer and Beach, J . Chem. Physics, 1946, 14, 701. Turkevich, McKenzie, Friedman and Spurr, J. Amer. Chem. SOL, 1949, 71, Dibeler and Mohler, J . Res. Nut. Bur. Stand., 1950, 45, 441. Benedict, Morikawa, Barnes and Taylor, J. Chem. Physics, 1937, 5 , I. 70, 2638. 4945. 46SCHISSLER, THOMPSON AND TURKEVICH 47 large amounts of solvent necessary to form the Grignard reagent and is reported by one set of workers to lead to the formation of considerable amounts of olefin by a disproportionation reaction., On the otber hand, lithium aluminium deuteride is not a readily obtainable reagent. The use of zinc and C2H,0D involves the previous preparation of isotopically pure C2H,0D.The method used in this work for the preparation of deuterated hydrocarbons consists of the reduction of the corresponding organic halogen compounds by metallic zinc in the presence of heavy acetic acid (CH,COOD). It has wider applicabilicy than the other methods and involves a relatively simple experi- mental technique. The reaction is carried out in a vacuum apparatus consisting of a reactor tube fitted with three side tubes for the storage of the organic halide, the heavy water and the zinc. An excess of redistilled acetic anhydride is placed in the reactor tube and some thoroughly out-gassed, granulated zinc added.Any evolved hydrogen gas is pumped off. The purified alkyl halide is then distilled into the reactor a t liquid nitrogen temperature. The mixtuxe is warmed to room temperature and any hydrocarbon, produced by the moisture in the system, is pumped off. A small portion (0.5 ml.) of de-aerated heavy water (99.8 yo D20) is distilled into the reactor. The reaction is allowed to proceed to exhaustion of the acid and the deutero hydrocarbon so formed is collected in a sample bulb. Successive small portions of D20 are added t o the reaction mixture and the deutero hydrocarbon evolved after each addition is collected as a separate sample. In this way, the major portion of the protium contamination in the system is removed.Examination of the mass spectra of successive samples shows in- creasing deuterium content which reaches a constant value. The isotopic purity of the product after the third addition oj D,O is usually greater than 95 yo. No measurable quantity of deuterium gas or olefins is produced by this reaction. A controllable reaction with a minimum of side reactions is obtained by the use of mono-iodidesJ dibromides and trichlorides. The halides used in the preparations are the following : for CH,D, CH,I ; for CH,D,, CH,Br, ; for CHD,, CHBr, and CHCl, ; for CH,CH,D, CH,CH,I ; for CH,CHD,, CH,CHBr, ; for CH,CD,, CH,CCl, ; for CH,CHDCH, CH,CHICH, ; and for CH,CH,CH,, CH,CBr,CH,. The method is not applic- able t o the preparation of tertiary deuterides or to cases when the halide atoms are not on the same carbon atom because of the formation of olefins.The tetradeuteromethane was made by the Fischer-Tropsch synthesis of D, and CO over cobalt thoria kieselguhr catalyst and was isolated from the reaction products by distillation through a low temperature Podbielniak column, The hexadeuteroethane was made by reaction of dideutero acetylene with deuterium over a 5 yo Pd on charcoal catalyst a t oo C. The unsaturates were removed by bromine vapour treatment €or 36 hours and the product was dis- tilled through a soda lime tube to a trap a t liquid nitrogen temperature. The 2 : 2-dibrompropane was prepared from g g yo.propylene by means of the following intermediates : I : z-dibrompropane (bromine and propylene), methyl acetylene (I : a-dibromide and butyl alcohol + KOH),s z : 2-dibrompropane (methyl acetylene and HBr on activated charcoal a t 180O C).s The mass spectrometer is an all-metal 60' sector Nier type instrument.The accelerating potential is 2000 V and magnetic scanning is employed. The bombarding electron voltage is usually 75 V and the trap current is 40 PA. There is no pusher or drawing out potential in the ionization chamber. The positive ion currents are amplified by a Victoreen 5803 miniature tube with a grid leak of 2 x 10l0 ohms. The output of this tube is amplified through two stages and fed into a Brown electronic recorder. We wish t o thank Mi. W. Higinbotham for the design of the amplifying circuit. The isotopic purity of the compounds was determined in the following way.The presence of impurities of higher deuterium content was shown t o be neglig- ible in all our preparations by the absence of masses higher than the parent of the desired deutero compound. The amount of compounds of lower deuterium content was determined by measuring mass spectrum a t electron bombardmeat voltages just necessary to produce the ion of the parent mass.l0 (Olefin content of all preparations was shown to be negligible in the same way.) The impunties when present (usually less than 10 yo) and a CIS correction of 1-15 yo per carbon atom was subtracted t o give the results presented below. Hurd, Meinert and Spence, J . Amer. Chem. SOL, 1930, 52, 1138. Wilson and Wiley, J . Chem. SOC., 1941, 14, 596.10 Stevenson and Wagner, J . Amer. Cheun. SOC., 1950, 72, 5612.48 MASS SPECTRA - - - 100.0 73'4 16.9 4-02 1-80 1-06 - - Results Methane .-The mass spectra of all the deuteromethanes are presented in Table I. These patterns can be compared with those of protium methane in the following way. To the first approximation, the probability of a given process, e.g. the removal of a hydrogen atom, is the same in all the deuteromethanes. One can therefore calculate the abundance of ions of various masses on a pure statistical basis. The formulae for the various masses are given in Table 11, if one sets the constants a and b equal to one. An evaluation of how nearly the calculated spectrum approximates the experimental spectrum is obtained by TABLE I - - 100'0 59.0 27'5 3'53 3-56 0-94 1-01 195.54 2 400 - Mass - 20 I9 18 I7 16 I5 14 I3 Sum Sens div., mm.I2 CHZDS calc. z = 1.48 b = 0.65 CHDI calc. & = 1-80 b = 0.76 - 100'0 35'8 46.6 4-82 2-32 0'34 1-07 2'10 - - 1'02 % :I'I yo CD4 calc. b = 0.90 CH4 expt. CHDS expt. - 100'0 36.2 46'3 3'05 1-91 0'44 1-07 191.07 2-10 2200 CDa expt . 100.0 - 72.8 - 4'24 1-60 - - 1-10 179'74 - CHsD expt. - - 100'0 73'5 17'3 3'52 1-65 0.8 7 196.84 2830 b = 0.55 -I- ~~ - - 100'0 58-9 28.8 5-10 2-62 0.89 1-06 - - 1-14 "/c r2.6 yo COO'O - 72.8 - 6.72 2-24 0.78 - I - - 1-40 % 3'8 % - I - - 100'0 79.6 7-91 2.84 1.07 191.43 2580 Weighed statistical deviation . Pure statistical deviation . 0'3 % 8.0 Yo TABLE I1 CH3D CHzDa Ions Ion Currents Ions Ion Currents Mass 18 I7 16 I5 I4 I3 I2 CH,D$ CH,C++ CD,+ CHD+ CD+ + CH,+ CHf C+ CHD,+ 1-000 M,, ~ 5 0 0 ~ M15 0-5oob M1,+o*167a2 M,, o.667ab MX4 o.500a2b M1,+o.r67b2 M,, o.500ab2 M13 a2b2 M,, - CH,D+ CH,D+ CH,+ +CHD+ CH,++CD+ CH+ C+ CHD, CD, CD4+ CD,+ CD,+ CD+ C+ - - - - 20 19 18 I7 16 I5 14 13 I2 - CHD3+ CD,+ CHD,+ CD$ CHD+ CD+ CH+ C+SCHISSLER, THOMPSON AND TURKEVICH 49 taking the square root of the sum of the squares of the deviations and dividing it by the total’number of ions produced.The figuxe of merit so obtained for the pure statistical calculation is given in Table I. While This approach reveals the qualitative features of the experimental spectra of the various deutef-omethanes, it is not satisfactory from the quantitative point of view. The second approximation consists in weighing the rupture of a C-H bond in a deutero compound by a factox a and the rupture of a C-D bond by a factor b.49 The resulting iormulae are listed in Table 11.Using these formulae, the con- stants a and b are so chosen as to give a minimum value for the figure of merit. The values of a and b that give the best fit t o the experimental data, the corresponding cal- culated spectra, and the figure of merit are given in Table I. It is seen that the calculated spectra are in good agreement with those obtained experi- mentally. Furthermore, the values of a and b chosen t o give the best fit vary linearly with the deuterium content of the deuteromethane (Fig. I). Application of this method t o the data of Dibeler and Mohler on deuteromethanes 6 gives a good approximation t o the ex- perimental values except that all deviations are positive when the values of a and b herein reported are used.The mass spectra of the deuteromethanes were also measured a t an electron bom- bardment potential of 15-2 V. At this electron energy, the methane molecule suffers the rupture of only one carbon hydrogen bond and the positive ions produced are those of the parent mass and the parent mass less one or two depending on whether a hydrogen atom or a deuterium atom is lost. Thus, rhe isotopic factors a and b can be measured directly under these conditions. Thk I FIG. I.-Relationship of isotopic factors to deuterium content of methanes. values found are 1.12 and 0.5G for CH,D, 1-14 and 0.59 for CH2D2 and 1-58 and 0-79 for CHD,. It is evident that these factors differ from those used in the calculation of the spectra characteristic of 75-V electrons, although they do show the same tendency to increase with increasing deuterium content of the deutero- methane.Ethane .-The mass spectra of ethane, monodeuteromethane, dideutero- ethane I-d2, trideuteroethane I-d,, and hexadeuteroethane are presented in Table I11 with the spectra for monodeuteroethane and dideuteroethane I -d, calculated by a method similar to that used for methane and based on the formulae given in Table IV. The agreement between the calculated and ob- served spectra is satisfactory. The greatest discrepancies occur in masses 29 and 30 in dideuteroethane and, as we shall see later, this discrepancy may be associ- ated with the process of removal of two hydrogen atoms from the ethane mole- cule.The paxtern for trideuteroethane I-d, calculated in the same fashion is very different from the observed one. This difference may arise from a re- arrangement of the deuterium atoms in the preparation of trideuteroethane 1-d3, giving rise t o some trideuteroethane 1-d2, 2-d or may be due to a selectivity in the process of dissociation of ethane on electron bombardment. The dissociation of ethane leads to the production of a very large ethylene ion current (320 units) which is caused either by a e + C2H, -+ C2H4+ + H2 + 26,50 MASS SPECTRA or by a e + C2H, 3 C,H,+ + 2H + 28 process. The fact that the appearance potential for the ethylene ion is lower than that for the ethyl ion and is equal to that for the production of the C2H,+ ion suggests that the process of formation of the C,H,+ involves the removal of a H, molecule rather tban two H atoms.This H, molecule may be formed by the removal of two hydrogen atoms from the same carbon atom or from adjacent carbon atoms. The latter process appears favourable on chemical grounds be- cause i t leads to the formation of CH,=CH,f, while the other process would lead t o a CHs-CH+ species. TABLE I11 Mass C2H6D expt. 36 34 33 32 31 30 29 28 27 26 25 24 - - - - - 100'0 70.76 320.0 95'6 63'7 9-50 2'00 - - - - 100'0 62-4 133.0 63.0 33'7 6.1 1-54 256.7 Weighed statistical deviation Pure statistical deviation 18 I7 16 15'5 I5 14'5 I4 13'5 I3 I2 Sens. div. / mm. -- - - 0-28 10.44 3'43 0.17 1-19 0.56 - 8-49 I 240 . . - - 5'87 9-17 4'67 3'40 1-13 0.71 - 0'12 1005 CHaCHD, expt. - - - 100'0 55'9 181.7 196.5 77'6 52-1 28.3 5'1 1.50 - - - 4-14 2-30 2.68 5-69 2.85 2.61 0'10 0.95 0.78 I010 CHsCHD1 calc.a = 1.15 b = 0.80 - - - 100'0 54'3 187.9 186.0 81.8 55'6 25'9 6.1 2'22 2'0 yo 4'8 % - - - - - - - - - - - CH3CDs expt . - - 100'0 43'9 59'6 285'9 131-8 82.7 61.2 2 6.5 5.20 2.40 - - 2-50 4'77 3'14 6.19 1-62 3'42 1-20 0'10 1-20 1-19 830 C2H6 100'0 61-4 - 343'8 - 78.2 - 53'8 - 6.26 1.07 - 10'1 10'2 2'00 0.07 0.32 1-98 - - - 0.75 I 250 -- We must consider, then, three different processes for the removal of two bydrogen atoms : (I) the removal of two hydrogens from the same carbon atom, (2) the removal of two hydrogens from adjacent carbon atoms, and (3) a random process involving a statistical combination of the two.It emerges that there is no difference in the statistics of these three processes for the case of monodeuteroethane. For dideuteroethane I-d, and trideuteroethane I-d%, 2-d, the differences are very small. However, for trideuteroethane I-dg, the removal of two hydrogens from the same carbon gives the ratio of 0.5 : 0.0 : 0.5 for H, : HD : D,, from adjacent carbon atoms the ratio is 0-0 : 1-0 : 0-0, and random removal gives 0.2 : 0.6 : 0-2. The experimental value obtained from the observed spectrum is 0.06 : 0.71 : 0.23.* *These values were obtained by subtracting from the current of mass 31 the statistical value of the current due to the loss of one deuterium atom ; from mass 30, the very small contribution from the loss of three hydrogen atoms and, from mass 29, the statistical value of a (2H + D) process.SCHISSLER, THOMPSON AND TURKEVICH 51 In an attempt to gain further insight into the process of hydrogen removal from ethane, a study was made of the ions produced by electrons of 13'0V energy.At this potential, one obtains only the ions of the parent mass and the ions resulting from the loss of two hydrogen atoms. There are observed, there- fore, mass 29, 30 and 31 resulting from the loss of D,, HD and H, uncomplicated by contributions from one atom or 3 atom processes. With trideuteroethane I-d3, the following ion currents were measured : 5-1 for mass 31 ; 87.0 for mass 30 ; and 31.0 for mass 29. Thus, the ratio of removal of H, : HD : D, is 0.04 : 0-71 : 0.25 in agreement with the value calculated from the 75 V data.It is curious to note that the removal of D, is favoured over H, in this compound. If one uses this ratio of two hydrogen removal obtained at low potential to calculate the two atom removal process for the CH3CH, spectrum and the statistical approach with u and b equal to one for the production of all the other ions, one obtains a calculated spectrum that is very similar to the one observed experimentally. This is, however, not entirely satisfactory and addi- tional deuteroethanes must be synthesized in order to obtain a complete under- standing of the dissociation process of ethane in the mass spectrometer. 112-0 for mass 33 ; TABLE IV CH3CH,D Mass 32 31 30 29 28 27 26 25 24 Ions C,H,D + C,H,D + C,H,+ + C2H3D+ C,H,+ + C,H,D+ C,H,+ + C,D+ C,H + C2H3+ + CgHD+ c2 + 32 31 30 29 28 27 26 25 34 Ion Current - M30 0 .8 8 3 ~ M29 o-167b MBY + 0 . 6 6 7 ~ ~ M,, o~500a2b M,, + 0 . 3 3 3 ~ ~ M,, o.667a3b M,, + 0 . 1 6 7 ~ ~ M,, o.833a4b M,, 0'333Ub M28 + 0*500U3 M27 afb M24 CH ,CHD, C2H4D2+ C,H,D + +C,H,Ds + C2H3D2+ C,H3D+ +C,HD, + C,H,++C,H,D++C,I C,H +C2HD + C,H + C,H, + C,D + c2 + '2 + %I 0 . 6 6 7 ~ M29 0.533Ub M28+O*ZOOU3 M27 o.067b2 M,,+o.600u2b M,, +0.067a4 M,, 0.400ab2 M,,+o-533u3b M2, o.400a2b zM26+o-333u*b M,, o.667a3b2 M,, a4b2 M,, 0'333b M,g+o'400a2 fif,, The ion currents corresponding to the C, fragments as shown by the mass spectrum taken at 75 V electron bombardment are determined by the molecular structure of the isotopic species, but are complicated by the occurrence of doubly charged ions of the C, group and possible rearrangements of the hydrogens during the fission of the carbon-carbon bond.The following may be stated about the doubly charged ions. The number of doubly charged ions is not proportional to the corresponding singly charged ions. The parent ion is not capable of carrying a double charge as shown by the absence of the 15-5 peak in C2H,D and 16.5 peak in C,H3D If one lowers the bombarding potential t o 17.5 V the doubly charged ions $sappear and so also does the fragmentation of the methyl ion into methylene and methyne ions. A pure methyl ion spectrum is obtained under these conditions. If the compound preserved the methyl group structure on electron bombard- ment one would obtain for CH,CD3 only masses 18 and 15 in equal amounts, and for CH,CHD2 masses 17 and 15.The occurrence of masses 17 and 16 in the methyl ion spectrum of the CH,CD, compound and mass 16 in the methyl The results are presented in Table V.52 MASS SPECTRA ion spectrum of the CH3CHD2 compound signify either that there is a reshuffle of the hydrogen atoms when the caxbon-carbon bond is broken or that the com- pounds prepared, while containing the expected number of deuterium atoms, did not have the deuterium atoms in the desired positions due to rearrangement in the synthesis. It is hoped that infra-red examination of the compounds will clear up this point. TABLE V Mass CHaCDs 18 7'2 17 2'5 16 2'1 1.5 7'1 CHaCHDz - 9'2 3'0 9'5 - - - 11'0 10'2 21'2 Propane .-The mass spectra of propane, monodeuteropropane zdl and dideuteropropane 2d2 are given in Table VI.The mass spectra of monodeutero- propane 2dl and monodeuteropropane 2dz were previously discussed.* No attempt shall be made a t this time to develop a scheme ot calculating the spectra. Some observations will be made extending those given in the previous communication. The removal of the first hydrogen from C,H, must come predominantly from the secondary hydrogen in the molecule because the value of the ion current of mass I lower in C,H,D2-2d2 is very low, 14.8 compared to 83-4 for C3H, and 76.6 for C,H,D-2dl. The value of 221-1 for the mass 31 in the spectrum of C3H,D2-2d, indicates that the two deuterium atoms are on the central carbon atom. The equal values for mass 30 in C2H,D, and 28 in C,H, signifies that when methane is formed from a propane molecule the methyl group picks up a hydrogen not from the central carbon atom but from the end carbon atom.An examination This must wait the synthesis of more deuteropropanes. TABLE V1.-PROPANE Mass 46 45 44 43 42 4= 40 39 38 37 36 31 30 29 28 27 26 25 24 16 15 I4 I3 12 - I 100'0 12'0 38.1 10.6 83'4 4'32 7-8 I 36-5 1'0 - - 244'8 151-2 82-3 14.5 1-26 0'20 0.26 5.86 2-27 0.81 0'45 CsH,D-zdl - 100'0 76.6 18.5 27'9 15'9 19'5 25'5 9'0 1.14 5-48 - 267.0 172.4 77'0 45'9 8-88 1-06 0.39 2-54 6'47 1.93 0.71 0.59 100'0 14.8 27.2 15.6 I 6.6 18.1 14'4 9-81 5'59 3-52 0.84 221'1 157.4 70'7 61-5 23'7 4'32 0.46 2-64 4-09 1-54 0.53 0.32 0'10SCHISSLER, THOMPSON AND TURKEVICH 53 of the appearance potential of the mass 28 in C,H, disclosed that i t had approxim- ately the same value as the appearance potential of the parent ion and a lower potential than that necessary for the removal of a methyl group (mass 29). This would seem to indicate that when the parent ion can form an olefin ion and a molecule, e.g. ethylene and methane, this will take place very readily. It should be noted that the formation of propylene and hydrogen is not a favoured process as is the formation of ethylene and hydrogen from ethane. We wish to thank Dr. Oliver A. Schaeffer and Mr. Adolph P. Irsa for assistance in the mass spectrographic work. The work presented has been carried out in part under the auspices of the U.S. Atomic Energy Com- mission. Chemistry Department, Chemistry Department, Princeton University, Brookhaven National Laboratory, Princeton, New Jersey, Ufiton, Long Island, U.S.A . New York, U.S.A.