MASSSPECTRA OF DEUTERATED BIPHENYLS
Oct., 1960
ever, appears much less reactive than the mercapto-pent,enyl radical. This latter result may in part be the result of assuming too high a value for
1359
MERCAPTAN CONVERSION, MOLECULES / l o o ev
0.6C
5
10
IS
I
I
1
x
io3
GR*
The G-value th:at is pertinent is actually the radiat,ion yield of mercapto-octenyl radicals, since it is assumed that these are the predominant chain ~ a r r i e r s . ~Radicals .~ that are produced by the radiation ‘but do not eventually give mercaptoocteiiyl radicals should not be included in GR. Olefins, as well as n-butyl mercaptan, can act as radical Thus, with octene-1 as a solvent, side reaot.ions may occur in which some of the radiat ion-produced radicals are converted to stable products without forming mercapto-octenyl radicals. COXl?ARISON
(l./mole/ sec.)
Mercapto-st-yryl 2 . 6 X lo3 5 X lo8 Mercapto-octenyl 3 . 6 x lo4 1 . 2 X 109 Mercapto-pcntenyl 1 . 4 x lo6’ 5 x 1010’ a Estimated viilues.
-I-
0.20
W
-I Y -I
s9
-a
0.10
0
z
;O.OE 0
z F 0.04 0
ka (l./mole/ W.)
C
w ,
TABLE I1 O F REACTION RATECOKSTAKTS k2
Radical.
0.4C
3 5
0
Ref.
3 This work
4
In addition to estimating the rate constants, the experimental data may be used to predict the effect of flow rate on the mercaptan-olefin reaction. The radical lifetime in the above experiments was sufficiently long that flow rate had a marked effect on mercaptan conversion. This point is illustrated in Fig. 3, using the alternative coordinate system. The reaction rate is given as the G-value for mercaptan conversion; the rotation rate of the sample tube is expressed in terms of the linrar flow rate of the reactants. The data il-
2
0.02
0.01
t
I
-
0
01
0.2
1
1
04
( $ / i d = (REACTION RATE) /(MAXIMUM
06
I O
REACTION RATE),
Fig. 3.-Comparison of theoretical and experimental relative reaction rates for n-butyl mercaptan-octene-1 system.
lustrate the use of a flow system to increase the efficiency of a fixed source of radiation.
THE MASS SPECTRA OF DEUTERATED BIPHEYYLS: MECHANISXS OF HYDROGES hXD CARBON LOSS PROCESSES’ BY J. G. BURR,J. M. SCARBOROUGH AND R. H. SHUDDE Research Ilepariwent oj“ dtonizcs International, A Dwision of IVorth American Aviation,Inc., Canoga P a r k , Cal. Brceaaed JaniLaTy 28, 1960
The monoisotopic mass spectra of biphenyl (I),biphenyl-4,4’-& (111,bipheny1-3,3’,5,5’-& (111),biphenyl-2,2’,6,6’-da (IV), bipheny1-2,2’,4,4’,6,6’-& (V), biphenyl-2,2’,3,3’,5,5’,6,6’-d~(VI) and biphenyl-d,, (VII), corrected for the contributions of less deuterated contaminants, are presented. The discussion is concerned with (1) the question of chemical selectivity in bond breaking where a possible preference for breaking of the para C-H bonds is shown; (2) the nature and use of the secondary isotope effcet in the loss of hydrogen and deuterium \+-herethis is defined in terms of normalized specific probabilities ( r and IIfactors); ( 3 ) the nature and significance of the primary isotope effect in the loss of hydrogen and deuterium from the molecule-ion; and (4)the factors governing the distribution of peaks within the peak groups corresponding to successive loss of carbon atom13 and also some factors probably governing the relative size of these peak groups.
I. Introduction Isotopic substitution of organic molecules has been used in ma,ss spectrometry2 both as a means for identif,ying l-he fragments observed and as a ( 1 ) Work performed under A E C Contract AT-(1 l-l)-GEh--S. This material was presented in part a t t h e Boston Meeting of the .4merican Cheniicd Society. April 5-10, 1959. ( 2 ) F. H. Firld and .J. 1,. Franklin, “Electron Impact Phenomena,” Academic I’ws.-, l n t . , Y e n Yorl;, S. Y., 1057, j J [ i . 204- ‘217 and Chap. \- in general.
means for verifying one or another of the theories of ionization and the dissociation of ions. Information thus gained about the identity of fragments formed has served to provide knowledge ahout the modes of decompositions of alcohols3 4 (7) (a) J. G. Bnrr, THISJOURNAL, 61, 1447 (1957): (b) W. H. M e Fadden, M. Lounsbnry and A . L. Wahrhaftig, Can J. Chem., 36, 990 (1958). (4) I, Friedman a n d J Tiirke\icli, .I .lm C h e m >S,IP , 74, 1666 ( 1952 1.
J. G. BURR,J. M. SCARBOROUGH A N D R. €I. SHUDDE
1360
Yol. 64
and the simpler straight chain alkane^.^,^ Similar interest is in the radiation chemistry of these mainformation has revealed the existence of extensive terials; some aspects of this are reported in rearrangements consequent upon ionization of an accompanying paper.16 We hoped that an several alkyl benzene^.',^ In several cases in- understanding of the molecule-ion dissociation formation thus obtained has been correlated with processes would help us understand the radiation data from the liquid and gas phase radiolyses of the chemistry. This expectation has been fulfilled same substances; evidence for the occurrence of only in part. similar dissociation processes in both the mass Many of our reasons for choosing biphenyl as spectrorneter and in the radiolyses was fo~nd.~.IOthe object of our efforts are discussed in the other The isotope effect observed in the dissociation paper.Ij We should like here to point out that of labeled organic molecules has been used to the ortho, nzeta aiid para positions of biphenyl are examine several semi-empirical treatments of mass distinctly non-equivalent to substitution by ionic spectra. Dissociation of isotopically labeled di- and free radical reagents; however the rateatomic molecules (such as hydrogen and nitrogen) determining step in these reactions is addition to has been explained successfully in terms of a the aromatic ring and the non-equivalence of the Franck-Condon model,2 but several effects in the several positions in substitution appears to be dissociation of more complicated molecules seem principally an empirical function of the quantum more easily treated by the quasi-equilibrium theory chemistry of the ring carbon atoms rather than a function of the strengths of the carbon-hydrogen of mass spectra.* The easier loss of hydrogen from partially bonds at these several positions. It is not known deuterated molecule-ions than from the completely whether these bonds do actually differ in strength,I6 protonated species, and the more difficult loss of and an a priori guess cannot be made as t o whether deuterium, ca 11 be expressed in terms of normalized these bonds should dissociate at different rates in specific relative probabilities for hydrogen aiid the various molecules-ions generated in the mass deuterium loss-the II and r factors’ as calculated spectrometer. from the mass spectra of a series of deuterated 11. Experimental methanes and ethylenes.”-I3 However, the effects of isotopic substitution upon The mass patterns reported here were obtained froni dissociation and rearrangement processes of mole- samples of the pure deuterated biphenyls synthesized here14 and of purified Eastman Kodak biphenyl. The samples cule-ions have been carried out chiefly upon were run in a modified Model 620 mass spectrometer, simple saturated and unsaturated aliphatic mole- manufactured by Consolidated Electrodynaniics Corporacules. With the exception of the admirable work tion a t a nominal ionizing potential of 70 volts. A series of of Meyerson and Rylander on the mass spectra of runs at nominal voltages from 9-70 volts WBS made from it was determined that fragmentation did not occur alkylaromatic molecules,7~slittle effort has been made which below 12 volts. All purity determinations were made a t 11 to use this technique to study the mass spectra volts. The temperature of the sampleinlet system was 250°, of arom:ttic molecules. and the temperature in the ionizing region was also 250’.
n
D
I
I1
I11
v
IV
We are presenting in this paper information which has been gained by examining the mass patterrie of the deuterated biphenyls, 11-VII, together with the mass pattern of biphenyl itself, I. The preparation and purity of these materials has been reported elsewhere. Our principal (5) F. E. Condon. J . Am. Chcm. Soc., 7 3 , 4675 (1951). (6) W. I-[. McFadden and A. L. Wahrhaftig, ibid., 78, 1572 (1956). (7) P, N. Rylander, S. hleyerson and 13. M. Griibb, ibid., 79, 842 (1957). ( 8 ) S. Meyerson and P. N. Rylander, THISJOURNAL, 62, 2 (1958). (9) J. G. Burr, ibid., 61, 1483 (1957). (10) J. G. Biirr. J . Ana. Chem. Soe., 79, 751 (1957). (11) V. H. Dibelei and F. L. Mohler, J . Research NatZ. Bur. Standards, 46, 441 (1950). (12) V. 13. Dibeler, F. L. Mohler and M. deHemptinne, ibid., 53, 107 (1954). (13) I+’. 1 .; RIohler, V. 13. Dibeler and E. Qiiinn, ibid.. 61, 171 (1958). (14) R. I. Akswii?, J. hf. Scarborough and J. G. Burr, J. O m . Chem., 24, 946 ( 1 9 9 ) reports the preparation of 11, 111, V, VII. T h e syn-
D
VI
1
VI1
111. Results Monoisotopic partial mass patterns of the biphenyls (I-VII) are shown in Tables I through IV; the patterns have been broken into sections simply for convenience in reading.” Table V thesis of IV and VI together with the catalytic driitpration of biphenyl witl be reported in a forthcoming note (R. I. :iBa\vie, J. Or@. Chem., in p r e s ) . (15) J. h‘l. Scarborough and J. G. Burr, THISJOCRN&L. 64, 1367
(1960).
(16) IS forthcoming papers by R. I€. Shudde and G. 1%‘. Lehman and by R . H. Shodde a n d J. M. Scarborough, on t h e vibrational modes of biphenyl i t will be reported t h a t the best fit between computed and experimehtal infrared spectra was obtained by using differing force constants for the C-H bonds in the ortho, mcta and p a r a positions. this, however, still does not provide useful information aboiit the dissociation enr-rgies of these bonds. (17) The complete raw and corrected I+titternr wilt be submitted to the American Petroleum Institute for publication in the Tables of Mass Spectral Data.
MASSSPECTRA OF
Oct., 1960 TABLEI RHEKYIS
m/e 164 163 162 161 160 159 158 157 136 ls55 154 153 1.52 151 150 149
TABLE 111
DEUTERATED Br(I-VII) PARENT REGION
THE MONOIBOTOPIC PATTERNS OF
THE
CIZHIDCizHsD CIZHODI CizHaDi CirHdDo CIIHZDS CIZDIO (I) (11) (111) (1%’) (V) (VI) (VII)
100 36 26 7 3 0
00 68 83 17 10 19
100.00 31.94 24.65 10.90 3.78 1.19 0 20
100.00 2.00 100.00 33.71 11.12 1.84 100.00 26.26 19.80 10.34 18.63 0.38 11.18 4.64 100.00 100.00 23.13 2.78 26.04 25.59 13.20 0.14 7.24 2 22 0.91 23.16 22.35 0 67 2.88 12.97 13 00 0.29 5.50 1.28 5.31 2.57 2.16 0.47 0.89 0.66 0.17 0.14
141
140 139 138 137
0.01 0.318 0.40 0 . 0 1 1.16 1 . 3 1 0.89 0.60 0.77 1..05 0.08 .14 1.97 0.14 .02 0.09 0.05
Loss of 4 carbons 110 109 108 107 106 0.13 105 0.90 104 103 0.23 3.05 102 3.87 0.99 101 1.33 0.50 100 0.35 99 0.61 98 1.06
0.18 0.30 1.48 2.72 0.91 0.47
0.57 2.74 2.39 1.05 0.51
0.28 1.51 1.97 1.06 0.34
0.35 2.35 1.56 0.77 0.38
0 . 3 3 3.80 2.s4 1.05 1.18 0 . Z 0.12 0.30 0 . 3
0 . GO 0.X
Loss of 5 carbons
TABLEI1 Loss of 1 carbon
146 145 144 143 142
d e
0.09 0.02 1.10 0 . 6 5 0.88 0 . 3 3 1.08 0.05 .99 0.30 .59
1.74 0 . 05 0 . 06
96 95 94 93 92 91 90 89 88 87 86 85
0.62 0.02 2.16 0.70 1.69 1.16 0.34
0.21 .30 .33 1.54 0.92 1.04 1.18 0.70 0 24
0.38 .;5 .57 1.52 0.90 0.88 1.11 0.84 0.43
0 . :39 0.51 0 71 0 49 2 47 .33 0 . 54 1.54 .70 0.51 1.30 0 . 4 9 0 . 6 7 1 . 6 5 1.10 0.75 .37 0.97 0.53 .58 .74 1 33 0.99 .78 .SO .53 1 . 1 3 .88 .72 .02 0 . 8 2 0.87 .28 .43 0.47 0.2;
TABLE IV
Loss of 2 carbons 136 135 134 133 132 131 130 129 128 127 126 125 124
4.13
0.02 3.9G 3.07 2.90 0.36 0.01
:2.35 2.97 2.49 1.83 0.41
1.34 :3.05 ‘2.43 ‘2.17 1.30 13.34
1.70 3.44 2.86 2.38 1.55 0.44
1.32 2.66 2.40 2.18 1.59
0.50 2.40 2.31 1.80 1.59 0.85 0.20
1.58 2.72 1.62 1.57 1.36 0.61 0.13
3.09 2.13 0.22
Loss of 6 carbons 84 83 82 81 80 79 78 77 76 75 74 73
1.87 1 . 4 8 13.64 14.30 29.57 41.76 20.40 5.82 3.20 4.75 2.51 0.39
0.73 1.37 1.98 2.07 1 5 . 2 1 16.02 22.79 23.37 8.66 9.01 4,32 3.55 3.69 3.14 1.36 1.37 0.21 0.24
0.70 3.98 11.60 21.40 17.70 6.00 2.72 1 i7
1..50 1 . 3 1 0 . 7 8 20.38 1.24 16.38 14.93 5 . 5 6 40.19 12.25 21.71 1 2 . 8 8 2.67 4.38 2.83 1 72 2.73 3.77 2.71 1.83 0.92 0 42 0 24 0.23 1 08
Loss of 7 carboiis Loss of 3 carbons
122 121 120 0.88 119 3.57 118 117 2.21 1 . 9 3 116 3 . 54 0.56 115 5.92 0 . 5 5 .49 .15 114 0.42 .60 113 0 . 8 1 .13
l . l j 0.93 4.30 2.80 2.17 1 . 6 8 0.59 0.22 .52 .19
0.15 1.77 3.45 1.00 0.34 0.41 0.18
0.225 6.10 3.51 2.29 0.36 0.23 0.20 0.68 0.44
shows the sum probabilities for the various carbon loss processes (Molecule-ion peak = 100) and also shows the “total ionization” for each molecule. 17:irious aspects of data presented in t,hese Tables i -jr1 are treated separately below. Tlie halfintegral peaks (double ionization) for biphenyl are omitted for the sake of clarity (but see Section B,
70 69 68 67 66 65 64 63 02 61
0.20 0.89 1 . 2 3 10.28 15.89 6.09 13.67 4.48 3.41 1.05 0.65 1.09
0.48 1.00 7.31 5.45 6.85 2.71 1 28 0 41
0.715 1.27 8.48 6.46 7.27 3.21 1.46 0.55
0.47 2.60 7.55 9.80 7 55 3.36 1.15 0.47
0.41 1.06 3.53 6.89 G.35 2.89 1.44 0.60
0.44 2.36 1.47 1.0:; 22.99 11.71 4 . 2 8 12.59 4.18 2.02 3 15 0.13‘3
0.70
0 70
footnote, below). The peaks for doubly ionized biphenyl-dlo species occur at integral mass numbers only. A. Isotopic Purity.-Isotopic purity was determined from the relative peak heights of the parent peak and the adjacent isotopic impurity parent peaks at a nominal ionizing potential of 11 volts. The isotopic piiritics obtained in this way :m recorded in the earliclr paper. l 4 I-iowei-cr. UPC of the isotopic puritv thus measured for biplieriyl-dlo (95.170 with 4iY0 biphenyl-&) did riot result iii
J. G. BURR,J. M. SCARBOROUGH AND R. H. SHUDDE
1362
TABLEV
VALUESFOR GROUP'AND TOTALIONIZATIONS' Group by C I P - CizHe- CizHs- CizHa- CizH4 CIZHPD4 Dc Da Dz D4 carbon HIQ loss (I) (11) (111) (IV) (V) (VI) Parent 1 7 3 . 2 3 172.71 170.69 169.86 166.88 164.77 15 2.10 2.15 2.15 2.57 2.07 2.11 2 10.97 10.45 10.73 12.43 9.82 9.61 3 7.59 7.51 7.87 9.29 7.24 9.67 4 7.62 7.41 7.66 8.95 6.76 6.45 J 6.77 G.73 7.68 7.90 6.71 6.06 6 C89.03 7 2 . 8 9 68.24 60.63 50.44 54.05 j5.63 24.62 25.70 29.22 23.33 26.79 8 55.07 25.94 30.18 37.32 27.54 29.56 9 13.46 9.98 11.83 15.12 10.12 11.38 Z 3181.48 340.39 332.73 353.29 310.91 320.46
-
CizDio
(VIG 163.64 1.88 9.59 7.41 6.76 5.61 72.86 41.24 37.41 12.65 359.05
By this we ;nean the group of peaks associated with the formation of an ion where 1,2,3,4 . . 9 carbon atoms have This is not been losi, from the biphenyl molecule-ion. precisely what is usually known as total ionization since we were unable to measure sensitivities for these samples; we know of no reation, however, why sensitivities should vary appreciably for :my of these samples.
Vol. 64
perimental peak heights corresponding to the loss of lH, 2H, 3H or 4H from biphenyl were taken as the total probabilities for the loss of 1, 2, 3 or 4 particles from the deuterated compounds. These probabilities then were multiplied by the appropriate statistical factor to give contributions of the several processes to a given m/e. In making these calculations the isotope effect in the relative rates of H and D loss was neglected since this matter is better dealt with by the methods described below. D. I' and IT Factors.-The calculation of I' and IT factors has been made according to the definitions2 probability of H loss from Cl2HnDl~-,, r = Specific Specific probability of H loss from C12H10
a
rI=
Specific probability of D loss from C12HnDloSpecific probability of H loss from ClzHlo
and the specific probabilities are defined in the following manner: the specific probability of H loss
TABLEVI SPECIFIC PROBABILITIES FOR HYDROGEN A N D DEUTERIUM Loss Co npound
a
rl
Loss of H or D El
1.00 (0.47)" Biphenyl ( I ) 1.10 .56 Biphenyl-dz (11) 1.18 .66 Biphenyl-dc (111)and (IV) 1.29 .73 Biphenyl-& (V) 1.51 .79 Biphenyl-d8 (VI) Biphenyl-dl,, (1.48)' .93 Values obtained by extrapolation; cf. Fig. 1 as an example.
complete removal of the peak at m/e = 163 in the monoisotopic pattern. We can only assign this deviation to slight instrumental instabilities since we also noted that the 13C/12Cratio varied slightly as the ionizing voltage was reduced. B. Preparation of Monoisotopic Patterns.The observed mass spect'rometer records were corrected for carbon-13 contributions and for contributions from the hydrogen impurity in the deuterium contaminant by a modification of the standard calculation. C . Calculated Patterns.-Calculated monoisotopic patterns of these materials that would represent the hydrogen (deuterium) loss processes and the carbon cleavage processes as purely statistical processes were prepared. These calculated patterns do not take into account any isotope effects; they are based on the isotopic composition of the matmerialsand a statistical loss of hydrogen and deuterium. The calculated values for one of the substances, IT,appear in Tables I1 through IV; calculated values for other substances have a similar agreement with observation and have been omitted for t'he sake of compactness and clarity. To obtain the calculated peak heights, the ex(18) Cf.,for example, G. P. Barnard "Modern iMms Spectrometry," t h e Institute of Ph,ysics, London, 1953, p. 201-202. T h e particular modification which we used is described in t h e report NAA-SRhIEN0-5125: our method r7as programmed in Fortran language for a n iE31 709 computer and details of this program mill be found in the N.1.1-SIL-ME~IO-5125 report,. T h e W/IzC ratio used was 0.1455 ithe averan,: value obtained from t h e low voltage patterns: t h e results were quite nsensitive t o t h e exact value of this ratio). T h e value of t h e H / V ratio used for each spectrum depended upon t h e size of the impurity p?ak in the low roltage patterns since our corrections for hydrogen content were confined t o t h e hydrogen impurity in t h e substitiient deriteriiirn.
7
7 -
LOSSof 213
IE = WE1
rz
( 2 . 12y 1.91 1 .so 1.72 1.91 (1.59)"
1.00 1.07 1.09 1.12 1.42 (1.15)"
+ HI) +IE2D--= r2/n2
n2 .. 0.47 .62 .67 .64 .74
.. 2 30 1 76 1.68 2.22 (1.56)"
from C12HnD10--n is equal to relatiw peak height of peak corresponding to loss of one H from CI2HnDlo-, divided by n. The only term in these expressioiis not directly available from the mass patterns is the probability for D loss. The value of this has been obtained by considering that the relative probability for the loss of one particle (H or D) from a molecule-ion is a linear function of the deuterium content of the molecule-ion. The peak height for the loss of one particle (RPHI) is established unambiguously for biphenyl by the size of the peak at mi e = 153, and for biphenyl-dlo by the size of the peak at m/e = 162. The corresponding values for the partially deuterated biphenyls were obtained from the expression RPHl = 36.68 - 0.0297N where S is the per cent. of deuterium in the molecule-ion. If the value of the RPHh for the peak corresponding to loss of one H is subtracted from this computed value, the difference is then equal to the value of the R P H d for loss of one D. The values I thus computed can be reduced to the r and I factors by application of the above definitions. The values thus obtained for the I? and II factors are listed in Table VI, and shown also on Fig. 1. The specific probabilities for double particle loss of hydrogen (Le., 2H l/zHD) and deuterium ( i e . , 2D l/,HD) can be estimated by a similar but longer and more approximate procedure. The steps involved in determining the loss of 2H HD 2D from a particular molecule-ion are shown below
+
+
+
+
RPH, = 2H
where N is the
+ HD + 2D = 26.83 - 0.070 N
yo deuterium
in thc inolccule-ion.
MASSSPECTRAOF DEUTERATED BIPHENYLS
Oct., 1960
RPHh and RPHd are obtained from the calculations of and II lis1ed above. Finally if we represent the peak at %,'e = (molecule-ion) - (one mass unit) as P - 1 and the peak at m/e = (moleculeion) - (two mass units) as P - 2, etc., then LOSSof 2H =: ( P - 2 ) - RPHd LOSSof HD = ( P - J) - R P H ~ I ~ LOSSoi 211 == RPHZ - (2H HD)
+
111'H3h T ~ ('stinluted S by multiplying the value of this peak in biphenyl (at m/e = 151) by the fractional hydrogen content of the particular moleculeion-probably the most approximate part of the calculations, but che least important since three particle loss from Diphenyl is not a large contributor to the ionization in the parent ion region. The ~-nlues for double particle hydrogen loss and double particle deuterium loss were normalized and made specific in the same way that this was accomplished for the single particle losses. The values for the factors and the isotope effects thus calculated are shown in Table VI, as rZ and TIz. IV. Discussion I. Preliminary Considerations.-For the purposes of the following discussion we are representing the dissociat Lon pr 3cesses of the biphenyl moleculeion by a simplified diagram
Inspection of Table I shows that chemical selectivity is not obvious, if present. If the ortho, meta and para carbon-hydrogen bonds in biphenyl dissociate at different rates, these differences must be of the same order of magnitude as the isotope effects and thus be masked by the isotope effects. The situation cannot be untangled by use of a priori information since there is no a priori quantitative information about the strengths of the seyeral carbon-hydrogen bonds, the effect of differences in bond strength upon rates of molecule-ion dissociation, or about the magnitude of the several isotope effects. A. The Secondary Isotope Effect.-The question of chemical selectivity can only be answered if this effect can be untangled from the tivo isotope effects. a r e shall make use of the normalized specific rate factors defined above (I' and IT factors) both for the purpose of disentaiigling isotope effects from chemical effects and also as a means for determining the magnitude and nature of the primary isotope effect. The manner in which we shall do this is qualitatively apparent from inspection of Fig. 1, the plot of I? and II factors as functions of the deuterium content of the moleculeion. It can be seen that each of these factors is a
/
L
I
1363
m/e = 154
[=I+
m/e =
77
11. Hydrogen :Loss Processes in the Moleculeion Region.-The ions produced by rupture of carbon-hydrogen and carbon-deuterium bonds in the molecule appear in Table I. The moleculeion whose dissociation produces this set of fragment ions has an undetermined structure; two of the most likely types of structure are shown on the preceding diagram (paths (1) and (2)). The losses of €[ and D atoms from a partially deuterated biphenyl molecule-ion are subject to the possibility that not all of the carbon-hydrogen bonds in the molecule-ion are of the same strength and that the weaE.er ones may have a higher probability of rupture this me shall define as chemical selectivity in bond rupture. The probability of carbon-hydrogen bond rupture is always higher than the probability of rupture of a similarly qituated carbon-dmterium bond; we define this as a primary isotope effect. Existence of a secondary isotope effect appears in the observation that the probabilities of C'-H bond rupture are higher in the deutera1,ed molecule-ions than the probability of C-H rupture in the biphenyl molecule-ion.
rather good linear function of the deuterium content, except that the r factor for biphenylds (T'I) is high, and the II factor is low (these two are not necessarily independent variables; cf. Section 111, D). This means that hydrogen is being lost abnormally easily from the 4,4'-position, and that deuterium is being lost with unusual difficulty from the other two sets of positions. One naturally then expects deuterium to be lost unusually easily from biphenyl-4,4'-dz; the point for the rI factor of this molecule-ion does not appear to deviate from the line. The primary isotope effect should, however, reduce the deviation of this point from the line to about one-third or less of the deviation of the gamma point for VI from the line thus making the effect of selectivity in I1 more difficult to detect. Before it can be said that one of these points deviates from a normal relationship, it is necessary to establish the nature of the normal relationship. I n Fig. 2 are shown the corresponding plots for the deuteriomethanesl1-l3 and the deuterioethylenes.11-13 I t n-ill be noted that in each case the
J. G. BURR,J. M. SCARBOROUGH AND R. II. S i I m D E
1363
\.-u1. 64
j - I
where B (eq. 2) is the change in non-fixed energy of the systems which results from progressive substitution of C-D bonds for C-H bonds, and where TABLEVI1 ENERGY LEVELS,ELECTRON DENSITIES AND BONDORDERS p--3\\ /'Q-T
4
Biphenyl, CI&
\>.-?
1 3
(+$
f'
'&
Energy levels: f2.27848; f1.89128; f1.31748; fl.OOOOp; *0.7046& Calcd. excitation energy: 1.40928 x 2.39 = 3.37 e.v. Calcd. ionization potential: 0.7058 X 2.39 7.18 = 8.87
+
Bond
1-2,143, 9-10,10-11 2-3,54,8-9,11-12 3-4,475,7-8,7-12 4-7
. . .. . . . . . . . . .. . . ...... . .
........
Atom
. . .. . .. . ....
Ground state
0.6601 ,6766 .6188 .... ,3697 1, 10 1.0000 2 , 6 , 9 , 1 1 1.0000 3,5,8,12 1.0000 4,7 1.0000
Excited state
Ion
0.5485 0.6043 ,7605 .7186 .4087 ,5137 .0161 ,4929 1 .OOOO .8415 1.0000 ,9803 1.0000 ,9105 1.0000 ,8768
Phenylhexatriene, C12H12,
I
5
20
40
60
80
IO0
PERCENT DEUTERIUM IN MOLECULE - ION. Fig. 1.--The r and rI functions calculated from the mm8 spectra of the deuterated biphenyls.
'I and I'I factors are smooth functions of the deuterium content of the molecule-ion; there is marked curvature to the methane plot and the ethylene plot appears to be linear although the data seem to distinguish between the two di-deuterioethylenes. The different 'I and TI factors for the two di-deuterioethylenes may possibly reflect an effect of the two different molecular structures. A rather more extreme example of such an effect is found in the mass spectra of the deuterated ethanols,2 where the almost corn plete preference of single hydrogen loss from the -CH2- group is accurately reflected in the jagged function of the 'I and II factors with deuterium content of the carbinols. A view of the mathematical nature of these factors can be obtained from the quasi-equilibrium theory of mass spectra1gf2where the rate constant for a particular dissociation process of the moleculeion is expressed (for a collection of loosely coupled oscillasors) as (19) fx. M. Rosenstock, A. L. Wahrhaftig and H. Eyring. Tech. Report. No. 11, June 25, 1952, Univ. of Utah, Inst. for Study of Rate Processes. Salt Latke Citv: H. M . Rosenstock, M. B.Wallenstein, A. L. Wahrhaftig and H. Eyring. Proc. Nat. Acad. Sci. U.S.,38, 667 (1952).
Energy levels : f2.15778; f1.7O1981 =I= .414?p; ) f1.00008; f1.00008; fO.a6588. Calcd. excitation energy: 0.7316 x 2.39 = 1.75 e.v. Calcd. ionization potential: 0.3658 X 2.39 7.18 = 8.05 e.v.
+
Bond
Atom
....
1-2 2-3 3-4 4-5 5-6 6-7 7-8,7-12 &Q, ll-12 9-10,10-11
....
.,.. .... .... .... ....
.... ....
....
1 2
....
....
3 4 5 6 7 8,12 9,11 10
.... .... ....
.... ....
.... ....
Ground state
Excited state
0.8639 ,4926 ,7633 ,5200 ,7764 .4461 .5954 ,6830 ,6550 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000
0.7000 0.7819 .6346 .5636 .4982 ,6308 ,7092 .6149 ,5254 ,6509 ,5618 ,5039 ,5309 .5632 ,7055 .6942 ,6809 ,6429 1.0000 .7759 1.0000 ,9700 1.0000 ,8318 1,0000 ,8956 1.0000 ,9148 1,0000 .8152 1,0000 ,9819 1.0000 ,9426 1.0000 .9978 1.0000 ,9341
Ion
C (eq. 3) is the difference in zero point energy between a C-H bond and a C-D bond (meaning of the other terms in the equations is the samc as in the original referencesLg). Thus progressive substitution of C-D vibrations for C-H vibrations mogressivelv increases the nonI
U
Oct., 1960
MASSSPECTRA OF DEUTERATED BIPHENYLS
fixed energy of the whole molecule and thus progressively increases the rate constant for dissociation of any particular C-H or C-D bondI9; on the other hand the rate constant for dissociation of a C-D bond is lower than the rate constant for a C-H bond in a given compound by an extent measured by the numerical value of C.19 The numerical value of C is about three orders of magnitude greater than the tiumerical value of B. To the extent that t,his approximation is qualitatively valid, it can be seen that the specific normalized rate factors (I' and II) should be smooth monotonic functions cf the deuterium content of the hydrocarbon. The several bits of evidence thus accumulated above all strongly suggest that these factors should be smooth monotonic functions of the deuterium content of the molecule-ions-as long as all of the C-H bonds in the molecule-ion dissociate at epuivalent specijc rates. If some of the C-H bonds in a non-symmetl-ical molecule-ion dissociate more casily than others then it seems reasonable to expect that this chemical selectivity will be reflected in deviations of these factors from the monotnnic relationship. From this point of view, we propose that the deviation (Fig. 1) of the specific rate factors for the biphenyl-d8 (VI) may represent L: distinct preference for the breaking of the C-H bonds in the para positions of biphenyl. We also vggest the general approach outlined above as a inethod for dirtinguishing isotope effects from rhemical selectivity when these two effects are of the same order of magnitude. If double partkle loss processes are considered 1/2HD) the cor( i e . , 2H l,12HD, and 2D responding normalized specific loss factors (I'z and nz;cf. Table ' E ccrtairily open to serious theoretical objectitms, hut calculation of excited molecule bond orders has been found qualitatively useful in several instances, 2 particularly in evaluating the effect of bond twisting upon the ultraviolet spectra of unsaturated molecules.24 The results are shown in Table T‘I[. for biphenyl and for l-phenylhexatriene-] .3,6. The izffects of excitation and ionization upon the mobile bond orders are particularly interesting. From Table VII, the interannular bond which is the neakest bond in the ground state of biphenyl appear$>strengthened by excitation or ionization, wherea
