Dec., 1942
IONIZATION OF
ISOBUTYLENE, PROPANE
AND PROPYLENE BY
[CONTRIBUTION FROM WESTINGHOUSE
ELECTRON IMPACT
2769
RESEARCH LABORATORIES]
Ionization and Dissociation by Electron Impact: Isobutylene, Propane, and Propylene BY D. P. STEVENSON’ AND JOHN A. HIPPLE It will be noted that while in the ethylene In a recent paper the results of a mass-spectroscopic investigation of the dissociation products series there is a difference of over 2.2 volts beof normal and isobutane were reported.2 We pre- tween the appearance potentials of R-H+ and H, the difference is less than 1.2 e. v. for sent here the resultsQofa similar investigation of R + isobutylene. As a part of an extensive study of the saturated hydrocarbons. A similar situation the form of ionization efficiency curves, we have obtains when one compares the appearance poexamined certain processes in propane and pro- tentials of R-CH3+ and R+ CH3 for R olefin or pylene. Inasmuch as our results on these latter saturate. These observations are consistent with molecules differ from those of a previous investi- the assumption which is usually made with regation, we also report appearance potentials of spect to the relative strengths of single bonds to some of the ions in the mass spectra of propane unsaturated and saturated carbon The ionization efficiency curve for C4H8+ from and propylene. The instrument and general technique have isobutylene shows inflections a t -12.5 e. v. and been described in the preceding article.2 The gas a t -18 e. v. The inflection a t 12.5 probably samples were given to us by the Standard Oil is to be attributed to the removal of the electron from a C-H bond orbital instead of from a C-C Company of Indiana. The appearance potentials, determined from double bond orbital. Although we can make no the “initial breaks,”z are summarized in Table I. assignment of the higher ionization potential The results of Delfosse and Bleakney3 on pro- (-18 e. v.), i t should be remarked that isobutane pane and propylene are included for comparison. also has an ionization potential about 9 volts We can offer no explanation for the discrepancies above the appearance potential. The ionization between our results and those of Delfosse and efficiency curves for C4Hs+ from i-C4H8 and for Bleakney. Careful checks show the results we C4H10+ from i-C*HlO are shown in Fig. 1. Barring molecular rearrangement, the ion report are consistent with our previously reported work on ethane and the butanes. CsH6+ formed from isobutylene has the structure The significance of the appearance potentials of CHsCCHz+. One might expect the C3&+ the various ions in the spectra of propylene and formed from propylene to be the isomeric CHZpropane has been discussed by Delfosse and CHCHz+ since the bonds to carbon atoms atB l e a k n e ~ . ~Our values are not sufficiently dif- tached to unsaturated carbon atoms are genferent to change their interpretation in any sig- erally weaker (more reactive) than the bonds to nificant detail. an unsaturated atom. The fact that A(C3H6+) The value of the appearance potential of the from C3He is 2.2 e. v. greater than A (C3He+) while parent ion in the isobutylene spectrum, A (C4H8+) A(C3H7+) from C3H8 is but 0.5 e. v. greater than = 8.9 e. v., which is to be associated with IvertA(C3HS+) suggests that the hydrogen atom lost (i-C4Hs) is surprisingly low. The substitution of by propylene was attached to one of the ethylenic carbon atoms. I n either case we can write two methyl groups in ethylene reduces T,rU)by -2.0 e. v. The effect of methyl for hy- (a) C3H6 -+ CHaCCHs+ + H + EA(CaHs+) 2 11.96 drogen substitution is more marked in the ethylene series than in the corresponding ethane series, (b) X d H s +CHaCCH2’ + CHs + EA(CsHs)+ = 11.51 = 11.7 e. v. and Ived(C4H10) = since Ivert(C%H~) (c)‘ i-CaH, 2H = C& CHa AH&.,= -4.93 10.4 e. v . ~The Iver,(ethylenes) decrease more or less smoothly with methyl substitution, while These equations lead to D(CHs--H) I4.48 e. v. the ethane series shows a greater decrease from This is in good agreement with the value of 4.38 propane to butane than from ethane to propane. e. v. deduced from other data.6 This agreement
+
+
+
+
(1) Westinghouse Research Fellow. Present address, Shell Development Co., Emeryville, California. (2) D. P. Stevenson and J. A. Hipple, THISJOURNAL, 64, 1588 (1942). (3) J. Delfosse and W. Bleakney, P h y s . Rev., 56, 256 (1939).
+
(4) F. 0. Rice and K. K. Rice, “The Aliphatic Free Radicals,” Johns Hopkins Press, Baltimore, Md., 1935, p. 75. ( 5 ) F. D. Rossini, Chem. Rcu., 87, 1 (1941), 1 kcal./rnole = 0 04337 e. v. (6) D. P. Stevenson, J . Chem Phys., 10, 291 (1942)
2770
1701. 64
.A(X+) e
V b
10.0 *o. 2 1 1 . 8 1 .f! 11.31 . 2 11.91 .2 12.31 .2 12.21 .2
a The ionization potential ol argon. I = 13 i b e v , nas takeii it1 accordance with the new conversion factor, 1 e. v = 8U66 cm -1, The appearance potentials it1 reference 2 should all be raised 0.07 e. v. We should like to thank the referee who called attentiori to the inconsistency oil thib point which existed in our original manuscript. Ref. 3 of text.
'
suggests either that the urisyninietrical structure for the C3H5+ from C3HB is correct or that the isonierization energy of the reaction CH2CHCH' --+ CH&CHz+ is quite small. The latter seems very likely, since the allene methylacetylene reaction involves but -0.Oi e. v.
-
14
lti
2!J
?:!
34
28
v- +. Big. l.---Ionizatio~i efficiency curves for the reactioiis t-C4Ha -+ CIHs'; i-CSHw -+. C4H7' H ; i-C4H10 C4H,o'. T h e voltage scale is uncorrected. The crdinates for isohuiylcnt. and isobutaiie are not coni-
+
+
I~arahle.
[d) E - C ~ = H ~CHzCCH2
or id') Z-C~HS = CHaCCH
il
12
The C3H4+from isobutylene may have either the allene or the methylacetylene structure. From the data summarized by R o ~ s i n i ,we ~ can wri tc
+ CH,
+ CHa
AH&
AH&
= 1.36 e v = 1.29 e v
Subtracting from i-CdHe; A(C3H4+) = 11.62, we have either I(CH2CCHz) = 10.26 or I(CH3CCH) = 10.3;je. v. Delfosse and Bleakney3found from direct measurement on allene, A(C3H4+) = 9.9 e. v. The ionization potential of methylacetylene has not been measured. In acetylene, A(C2H2+) = 11.2 e. v . ~ By analogy with the observed decreases in the vertical ionization potential brought about by substituting a methyl group into ethane and ethylene, one may guess that I,r,,t(CH&CH) is 1 e. v. less than (C2H2)or IVert(CHsCCH)= 10.2 e. v. Thus we 'ire unable to reach a decision with regard to the structure of the C3H4+ in the isobutylene spectruni. I t may be noted that regardless of the structure of this ion, the methyl and hydrogen which form the methane come from adjacent carbon atoms rather than from the same atom.* l'hr relatively low values of the appearance potentials of the ions CzHb+ and C2H4+ in the isobutylene spectrum indicate that a minimum number of bonds are lost in the course of the rearrangement and dissociation reactions through which they are formed. The essential sharpness 17) 1 I 11933:
%k, L-PI
T a t e , P T Smith and 4 1. Vaughan, Phys Rev , 48, 5'23 tlir discuswon in
ref, 2 and 3
Dec., 1942
IONIZATION OF
ISOBUTYLENE, PROPANE AND PROPYLENE BY ELECTRON IMPACT
2771
with which the ionization efficiency curves (see Fig. 2) rise to their maximum values suggests that but one process is involved in each case, i. e., one set of products. The low precision of A(C~HS+) precludes the unique assignment of the un-ionized fragments which accompany the formation of C2Hs+. From the data of RossinP we can write AHZOS8.1 = 1.23 e. v .
(e) i-C4H8 = 2CzH4
while Tate and co-workerss give (f)
CZHI+C2H4+
+ e-
A(C*Hd+) = 10.8 e. v.
Adding, we find (g) i-C4Hs +CzH4+ -I- C2H4
+
dodod.
e-
(CzH4’)
= 12.0 e. V.
in excellent agreement with the observed value (Table I). If, instead of the electron impact value of A (C2H4+), we had used Price’s’O spectroscopically determined TZ(C2H4) = 10.4 e. v., the value of A (CZHh+) calculated for (g) would be 0.4 e. v. below the observed value. This suggests 20 30 40 60 70 80 0.4 e. v. as a limit to the activation energy of the v- +. reverse of reaction (g). We have assumed that Fig. 2.-Ionization efficiency curves for various processes the products are of the ethylene rather than the in isobutylene. (CdHs+)l is t o be compared with the Cz ethylidene structure, an assumption which is unmasses. (CaHs+)z is to be compared with the Ca masses. warranted a t present. Until the energies of ethyliThe form of the ionization efficiency curve of dene and its ion are known from other data, this CH3+ from propane is of a character intermediate ambiguity in the foregoing discussion must reof this ion in ethane and the butanes. to those main unresolved. The formation of the methyl ion in the ethane The ionization efficiency curve of the ion of spectrum involves mainly single ionization and in mass 27 (primarily C2Ha+) of the isobutylene spectrum is similar in form to the curves for this the butanes only double ionization. The value ion in the spectra of ethane and the butanes. The of A(CH3+) from propane, while only rough, rather large range of electron energies through indicates that a process involving only single which the slope increases suggests that one or ionization contributes. The very pronounced more secondary processes, involving more frag- curvature of the ionization efficiency curve in mentation than the primary process, are involved. the vicinity of 26 volts indicates double ionizaCombining reaction (e) with CzH4, A (C2H3+)= tion to be the more important source of methyl ions. The CH3+ current from propane attains 14.1 e. v . , ~we find about 10% of its maximum value for 30 volt i-C4Hs+ C2H4 + CzH3+ + H + E(h) doalod.(C&+) = 15.3 e . v. electrons, while for similar electrons in ethane, in good agreement with the observed value of CH3+ is 50% of its maximum value. The methyl A(C2H3+) = 15.(2) e. v. We cannot exclude ion current is but 6% of its maximum value for 30 C2H5 as the un-ionized product which accompanies volt electrons in the butanes. The relative abundances of the various ions the CzH3+. The association of CzH4 and H to C2Hj in the mass spectrum of isobutylene are given, for gives off about 1.7 e. v., thus for the reaction round values of the bombarding electron energy (h’) i G H s --f CzHs + CZHO * + ein Table 11. Qualitatively, the distribution of inone would estimate A(C2H3+) r l 3 . 6 e. v., well tensities is very like that observed in the butane below the observed value. The 1.5 e. v. discrepAs is the case in the spectra of propylene, spectra. ancy could easily be assigned to an activation propane and the butanes, the principal reaction is energy of the reverse of reaction (h’). (9) P Kusch,A. Hustrulid and J. T. Tate,Phys. Rcv.,ti2,843 (1937). (10) LV. C Price, ibid., 47, 444 (1933).
R-CHI
+Rf
+ CHI +
E-
In the Cd region of the spectrum, isobutylene
2 773
THOMAS
E. MOOREAND
-rHR ?l!/
