It is of interest that the increase in the yield of 1butene with photon energy parallels the increase in the vield of n-butane, which as we have shown above, is formed in a reaction sequence involving H atom addition to ethylene. The presence of deuterium labeled butenes other than C4D, and C,D,H, in the photolysis and radiolysis of C2H4-C2D4 mixtures might be explained by the participation of diffusive recombination of vinyl and ethyl radicals in the overall 1-butene production. Increased diffusion would explain the observation that butenes such as C1H2D, C,D,H, CH Ds, and C3HD3 are formed at relatively higher yields in the liquid phase [3d] than in the solid phase radiolysis and that in the liquid phase they increase relative to the yields of C4H8, C4H4D, and CDs, with an increase in temperature [3d].

It is understood that proposed mechanism is a tentative one and that other mechanisms such as those proposed in previous studies cannot be ruled out.

Of the CHS products formed in the solid phase irradiations of ethylene (table 1) there is one product, cyclobutane, which seems to be formed via a C2H intermediate. This statement is based on the fact that the relative yield of cyclobutane is very small in the photolysis with 8.4 eV photons where presumably few ions are formed (cyclobutane was not even detected in the earlier study [1] at this energy), and increases by nearly an order of magnitude when the photon energy is raised to 10.0 eV and again increases when the energy is raised to 11.6-11.8 eV. This large increase in yield with energy can be contrasted with the yields of the butene products, which increase by less than a factor of two when the energy is augmented from 8.4 eV to 11.6-11.8 eV. Furthermore, the fact that approximately 90 percent of the cyclobutane formed in the photolysis and radiolysis of C2D4-C2H4 mixtures consists of C4H8, C4H4D4, and C4D8 (see Results) indicates that this product is formed mainly in a reaction of C2H, (C2D4) entities as has been shown. before [3d] in the liquid phase radiolysis of ethylene. Cyclobutane is also formed in the mercury photosensitized photolysis [18] of ethylene at a pressure of 700 torr, with a quantum yield of 3.8 × 10-6. In those experiments, the cyclobutane product was suggested to be formed as a result of a reaction between triplet state ethylene and ground state ethylene:

Cyclobutane has been reported [19] as a product (G~0.1) in the gas phase radiolysis of C2H4 (pressure: 100 torr) and ascribed to the participation of triplet ethylene. Reaction 17 cannot be operative in the 10 and 11.6 eV photolysis experiments, since the ejected electrons will have insufficient energy to bring about the optically forbidden transition to the lowest triplet state at 3.6 eV [20].

The 2-butenes are formed as minor products in all the solid phase irradiations given in table 1. Their yield shows little or no variation with increasing energy; hence, their formation is probably not associated with an ionic process. This is of interest, since, as indicated above, in the gas phase at pressures of 100 torr or less, it has been shown that reaction 10 of the ethylene ion with ethylene leads mainly to the formation of 2-C4H ions. Thus, at any rate, the absence of 2butene as an important product in any of the solid phase experiments demonstrates that C4H ions formed in reaction 10 do not undergo neutralization to form 2-C4H8, under these conditions.

Methylcyclopropane is also formed in small yields in all the solid phase photolysis and radiolysis experiments reported in table 1. In experiments carried out with C2H4:C2D4(1:1) mixtures, about 75-80 percent of this product consists of C4H8, C4H4D4, or CDs. In ethylene at 8.4 eV, the formation of methylcycloproan earlier study [1] of the solid phase photolysis of pane was attributed to a reaction of an ethylidene with ethylene:

Although this mechanism which requires a rearrangement of a long-lived excited ethylene molecule would indeed account for the isotopic distribution of methylcyclopropane products formed in the C2H4:C2D4 mixture, a free radical mechanism occurring in the cage may also explain the experimental observation. Addition of CH2CH to C2H4 occurs with a low activation energy (0.14 eV) for thermal CH2CH radicals and may involve a neighboring molecule, especially if, as in the gas phase, the CH2CH retains some internal energy after its formation. Such an addition process would lead to the formation of the 3-butenyl radical [3f] and conceivably also of the methylene cyclopropyl radical. These two radicals may capture a neighboring hydrogen atom to form dimeric 1-butene and methylcyclopropane respectively.

phase photolysis and radiolysis of ethylene. Indeed, the isotopic composition of the cyclopropane produced in the photolysis and radiolysis of a C2H4: C2D4 (1:1) mixture (see Results) indicates that more than 80 percent of the cyclopropane consists of c-C3H6, c-C3H4D2, c-C3D4H2, and c-C3D6. This distribution is consistent with the formation of cyclopropane through addition of CH2(CD2) to C2H4(C2D4). Insertion of a CH species into a C-H bond of ethylene would also lead to the formation of propylene

(20) CH2+ C2H,→ CH3CHCH2.

Propylene is formed as a product in these experiments (table 1), but has not been analyzed isotopically.

If indeed the cyclopropane and propylene products can be assumed to result from reactions 19 and 20, we can infer the occurrence of the primary process:

(21) C2H* → 2CH2.

This process requires less than 8.3 eV [7a] so there is enough energy available even in the 8.4 eV photolysis for its occurrence. It is seen that in the photolysis the relative importance of cyclopropane formation increases with increasing energy.

4.4. The Formation of Hexenes

Several C6H12 products are formed in the solid phase irradiations. The total yields of these products are listed in table 1.

The yields of the hexene products are relatively small in the 8.4 eV photolysis, but undergo a large increase when the energy is raised to 10.0 eV, and more than double again when the energy is further raised to 11.6–11.8 eV. These observations are very similar to those made above concerning the yields of the cyclobutane product. As in that case, we can infer that the large increase in yield when the energy is raised to a point where ionization is certainly of importance (10.0 eV) may be related to participation of ions in the formation of the products in question. It has been reported before [3c] that the 2-hexene product formed in the radiolysis of a C2D, -C2H, (1 : 1) mixture at 77 K consisted of more than 75 percent C6H12, C6H4D, C6H8D4, and C6D12. In the present study (see Results) it is seen that the 2-hexenes, as well as the 1-hexene product have a similar isotopic composition when an equimolar ethylene mixture is irradiated with 11.611.8 eV photons. That is, most of the hexene products are evidently made up of C2H4(C2D4) units.

In the earlier radiolysis study [3c] it was suggested that hexene is formed in a process initiated by reaction of the C2Hion with ethylene and terminated by electron recombination with a CH species. The present

photolysis experiments seem to substantiate this interpretation. Especially if one considers that an electron scavenger such as CCL, has a profound effect on the yields of the hexene (table 1). The actual role of CCI, in the enhancement is uncertain. It is however of interest to note that the yields of the C4 products are not seriously affected by CCl4.

5. References

[1] Tschuikow-Roux, E., McNesby, J. R., Jackson, W. M., and Faris, J. L., J. Phys. Chem., 71, 1531 (1967).

[2] For a review see Holroyd, R., in Fundamental Processes in Radiation Chemistry, Ausloos, P., Ed. (1968).

[3] (a) Chang, P. C., Yang, N. C., and Wagner, C. D., J. Am. Chem. Soc. 81, 2060 (1959); (b) Collinson, E., Dainton, F. S., and Walker, D. C., Trans. Faraday Soc. 57, 1732 (1961); (c) Wagner, C. D., J. Phys. Chem. 66, 1158 (1962); (d) Holroyd, R. A., and Fessenden, R. W., J. Phys. Chem. 67, 2743 (1963); (e) Wagner, C. D., Trans. Faraday Soc. 64, 163 (1968); (f) Fessenden, R. W., and Schuler, R. H., J. Chem. Phys. 39, 2147 (1963); (g) Brash, J. L., and Golub, M. A., Can. J. Chem. 46, 593 (1968); (h) Klassen, N. V., J. Phys. Chem. 71, 2409 (1967); (i) Wagner, C. D., J. Phys. Chem. 71, 3445 (1967).

[4]

Rebbert, R. E., and Ausloos, P., J. Chem. Phys. 46, 4333 (1967).

[5] Scala, A. A., and Ausloos, P., J. Chem. Phys. 47,5129 (1967). [6] Gorden, R., Jr., Rebbert, R. E., and Ausloos, P., Natl. Bur. Std. Technical Note 496 (1969).

[7] (a) Chupka, W. A., Berkowitz, J., and Refaey, K. M. A., J. Chem. Phys. 50, 1938 (1969); (b) Botter, R., Dibeler, V. H., Walker, J. A., and Rosenstock, H. M., J. Chem. Phys. 45, 1298 (1966); (c) Brehm, B., Z. Naturforschg. 21a, 196 (1966); (d) Al-Joboury, M. I., and Turner, D. W., J. Chem. Soc., 4434 (1964).

[8] Vermeil, C., Matheson, M., Leach, S., and Muller, F., J. Chim. Phys. 34, 596 (1964).

[9] Gorden, R., Jr., and Ausloos, P., J. Chem. Phys. 47, 1799

(1967).

[10] Tiernan, T. O., and Futrell, J. H., J. Phys. Chem. 72, 3080 (1968).

[11] (a) Stevenson, P., Rad. Res. 10, 610 (1959); (b) Ausloos, P., and Lias, S. G., Actions Chimiques et Biologiques des Radiation, 11, 1 (1967).

[12] (a) Sauer, M. C., and Dorfman, L. M., J. Chem. Phys. 35, 497 (1961); (b) Okabe, H., and McNesby, J. R., J. Chem. Phys. 36,601 (1962); (c) Ausloos, P., and Gorden, R., Jr., J. Chem. Phys. 36,5 (1962).

[13] Back, R. A., and Griffiths, D. W. L., J. Chem. Phys. 46, 4839 (1967); and Borrell, P., Cashmore, P., Cervenka, A., and James, F. C., J. Chim. Phys. 229 (1970).

[14]

[15]

Sieck, L. W., in Fundamental Processes in Radiation Chemis

try Ausloos, P., Ed. (1968).

Ausloos, P., Rebbert, R. E., and Sieck, L. W., J. Chem. Phys. in press and references cited therein.

[16] Meisels, G. G., J. Chem. Phys. 42,3237 (1965).

[17] Ausloos, P., Structure and Reactivity of Hydrocarbon Ions in Ion-Molecule Reaction, J. L. Franklin, Ed. (Plenum Press, 1971).

[18] Chesick, J. P., J. Am. Chem. Soc. 85,3718 (1963).
[19] Meisels, G. G., J. Am. Chem. Soc. 87, 950 (1965).
[20] Evans, D. F., J. Chem. Soc. 1351 (1957); 1735 (1960).
[21] Frey, H. M., Progr. Reaction Kin. 2, 131 (1964).

(Paper 75A3-657)

Kinetic

Mass

Spectrometric

Investigation

of the C5

lon

*

constants.

1. Introduction

Recent publications from this laboratory [1] have illustrated that a photoionization mass spectrometer is advantageous for studying the reactions of parent hydrocarbon ions. In some cases selective ionization of hydrocarbons in the presence of organic or inorganic additives may be achieved by proper utilization of the line emission (123.6, 116.4, or 106.7 nm) of rare gas resonance lamps. Furthermore, when ionization is induced by photoabsorption at energies only slightly above the ionization threshold, it is possible to investigate the reactions of parent ions with parent molecules in the absence of those complicating reactions associated with those fragment ions which would be produced at higher energies. The fact that parent ion-parent molecule reactions could be investigated at thermal kinetic energies in such an instrument at room temperature over a considerable pressure range was instrumental in the elucidation of the kinetics of formation and reaction of alkane dimer ions (CnH2n+2)[le, lf].

The present photoionization study differs from those mentioned above in that C4-C5 alkanes were irra

*This research was supported by the Atomic Energy Commission. "NBS Postdoctoral Research Associate 1968-70.

diated with photons of sufficient energy to induce unimolecular fragmentation of the parent ion. The reactivities of the resultant fragment ions towards the alkanes were derived from the variation of the composite mass spectrum of a particular system as a function of pressure. The ion-molecule chemistry occurring in some of these alkanes (n-C4H10, i-C4H10, and neo-C5H12) has been investigated previously by kinetic mass spectrometry using high energy electrons [2]. In those studies, however, the temperature of the ion source was approximately 500 K and the electron energy was well above that of the photons used in this investigation. Consequently, a considerable variety of fragment ions were produced and it was difficult in some instances to determine which of the many fragment ions were responsible for the formation of ionic products.

It was also considered of interest to compare rate data obtained from this study with that derived from analysis of the neutral products of ion-molecule reactions occurring in alkane systems [3]. Such a comparison is of particular importance since end product analysis has recently revealed the formation of alkyl and olefinic ions with more than one structure in the unimolecular decomposition of n-alkane and cycloalkane parent ions [4]. Since isomeric ions are known to exhibit different reactivities, one might expect to de

rive additional information from the contour of the ion decay plots obtained from the mass spectrometer.

2. Experimental and Results

The high-pressure photoionization mass spectrometer used in the present study has been described elsewhere in detail [la, b]. In all experiments the reaction chamber (ion source) was operated at 300 ±2 K in the absence of any internal electric or magnetic fields. All materials were purified by gas chromatography and subsequently distilled at low temperatures in order to remove traces of water. The techniques used for the evaluation of absolute rate constants for bimolecular reactions have been described in detail elsewhere [la, b]. Experimentally, the method involves a determination of the composite mass spectrum, including all reactant and product ions, as a function of sample pressure in the photoionization chamber. The logarithm of the percentage composition of the mass spectrum is then plotted versus pressure in the manner displayed in figures 1-6. The thermal bimolecular rate constant is then determined from the slope of the resultant decay curves found for primary

FIGURE 1. Composite mass spectrum obtained from the photionization (106.7-104.8 nm) of i-C4H10 versus pressure of i-C4H10.

PRESSURE (Nm ̄2)

FIGURE 2. Composite mass spectrum obtained from the photionization (106.7-104.8 nm) of n-C4H10 versus pressure of n-C4H10.

ions, which are straight lines for a single reactive species, and the calculated ionic residence time in the photoionization chamber. Rate constants are always derived from initial slopes since the ionic residence time may increase at higher pressures due to nonreactive scattering of ions exhibiting low overall reactivities. This latter condition will yield decay curves which exhibit an increased slope at higher pressures and are concave downward. Alternatively, a decay curve which is concave upwards at low pressures indicates that two or more empirically equivalent ions are present at this particular mass-to-charge ratio, each of which exhibits a different overall reactivity.

3. Discussion

3.1. Unimolecular Fragmentation

The primary mass spectrum resulting from photoionization of the C4-Cs alkanes by 11.6-11.8 eV photons may be obtained by extrapolating the parent and fragment ion curves given in figures 1 through 6 to

"zero" pressure. Because the "parent minus one" and "parent minus two" ions in each system are also the products of fast ion-molecule reactions, it was not possible to obtain an accurate estimate of their abundance in the primary mass spectrum. However, it is apparent (figs. 1 to 15) that the primary yields of these ions approach low values at low pressure. For n-C5H12 and i-C5H12, the Σ(C5H+CsH11) comprise approximately 2 and 10 percent of the primary mass spectra, respectively. As indicated by the summary in table 1, all of the major fragment ions have appearance potentials less than 11.2 eV at 300 K [5].Ions such as C2H1, C3H5, C2H3, etc., were not observed since the threshold energy requirements for formation of these species from C4-Cs alkanes are either very close to or in excess of [6, 7] the energies of the photons used in this study. The ions listed in table 1 are also the major species observed in a previous study of the photoionization of alkanes at energies approaching 11.5 eV [5]. For comparison, the relative abundances measured in the latter study at 11.25 eV are also included in table 1. As expected, for all compounds listed, the relative abundance of the parent ion is reduced at 11.6-11.8 eV. The other major difference be

*Reference [5].

**This work.

tween the mass spectral patterns is reflected in the greater probability for C-C scissions over four center olefin ion elimination processes when the energy is increased from 11.25 to 11.6-11.7 eV. In view of the slight differences in energy requirements for these two processes, this observation can largely be ascribed to the lower frequency factor associated with rearrangement reactions [8]. In the discussion which follows, the structures of the reacting ions will be considered.

3.2. Bimolecular Reactions

a. Alkene lons

The fractional intensities of the C3H ions produced in the photoionization of i-C4H10, n-C4H10, n-C5H12, and i-C5H12 decrease linearly as a function of pressure. Therefore, we may assume that this ion has only one structure, CH3CHCH2, at the time of reaction. The total rate constants for reaction of C3H with the various alkanes given in table 2 agree well with those obtained for reaction of C3D formed by photoionization of propylene-de with 10 eV photons [lb] in the presence of C4 and C, alkanes. In the case of i-butane, for which two reaction channels are possible: