Reinterpretation of the Electron Spin Resonance ...

Reinterpretation of the Electron Spin Resonance Spectrum of the oXylene Negative Ion J. R. Bolton Citation: J. Chem. Phys. 41, 2455 (1964); doi: 10.1063/1.1726286 View online: http://dx.doi.org/10.1063/1.1726286 View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v41/i8 Published by the American Institute of Physics.

Additional information on J. Chem. Phys. Journal Homepage: http://jcp.aip.org/ Journal Information: http://jcp.aip.org/about/about_the_journal Top downloads: http://jcp.aip.org/features/most_downloaded Information for Authors: http://jcp.aip.org/authors

Downloaded 20 Nov 2012 to 129.128.216.34. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

15 OCTOBER 1964

VOLUME 41, NUMBER 8

THE JOURNAL OF CHEMICAL PHYSICS

Reinterpretation of the Electron Spin Resonance Spectrum. of the o-Xylene Negative Ion J.

R. BOLTON*

Department of Chemistry, Havemeyer Hall, Columbia University, New York, New York

(Received 1 June 1964) The electron spin resonance spectrum of the a-xylene negative ion has been re-examined in view of a discrepancy between one of the previously reported splitting constants and the predictions of molecular orbital theory. This splitting constant has been found to be 1.81 G instead of "'0.2 G. The new values of the hyperfine splittings reported here are in excellent agreement with the predictions of both molecular orbital and valence bond theory. A technique is described by which molecules such as o-xylene can be reduced.

INTRODUCTION

approximation to the electron spin distribution in the o-xylene negative ion. Similar conclusions concerning the spin distribution in this molecule are obtained from a valence bond treatment. 5 ,6 The odd electron in the m-xylene negative ion should also reside in the symmetric orbital. However, in this molecule, the methyl groups do not bring about a large shift of the hyperfine splittings from the values given in Fig. 1 (a). It is thus surprising that one of the ringproton splittings reported by Tuttle4 is only "'0.2 G. If his proton splittings are assigned as in Fig. 1 (b), this value of ",0.2 G must be compared with the expected value of ,....,,1.9 G. A valence bond calculation predicts 2.13 G for this splitting.6 It has been found for methyl- and alkyl-substituted benzenesl ,7 that if I Q (H) 1=22.5 G2 (from the total extent of the benzene negative-ion spectrum) is used, the spin densities calculated from the McConnell relationships add up very closely to one. In particular, consider the m-xylene negative ion: Here we need a value of Q(CHa) which can be obtained from the normalization of spin density and I Q (H) 1=22.5 G

HE electron spin resonance (ESR) spectra of the methyl-substituted benzene negative ions are of considerable interest because they provide a very clear picture of the electron distribution in the two degenerate antibonding molecular orbitals of benzene.! Until recently, only three methyl-substituted benzenes had been successfully reduced, namely toluene, p-xylene, and m-xylene. l - a Several attempts were made to reduce o-xylene but without success,l-a However, in 1962 Tuttle' was able to obtain an ESR spectrum from the reaction of potassium with o-xylene in 1: 2-dimethoxyethane (DME) at -80°C. This spectrum was readily interpretable as that of the o-xylene negative ion. Following the molecular orbital treatment of substituent effects on the degenerate orbitals of benzene developed in Ref. 1, it is clear that the two methyl groups in o-xylene will result in the symmetric orbital being of lower energy than the antisymmetric orbital. Hence the electron distribution in the symmetric orbital [illustrated in Fig. 1 (a) ] should provide a good

T

0

7 .50)

I

i2 1

12

!

(0)

I

ji(1.88)

1

6.95

-0.20' CH~·OI 6

I

4

CH

3 OK

=-0.2

(b)

O 5

I

4

2

12

6.93, 1.81 0

3

CH 2 .000 3

2

3

OK'

CH

3

0.17,

(e)

FIG. 1. (a) Spin density distribution in the symmetric E,u orbital of benzene. Numbers in parentheses are the hyperfine splittings that would be obtained if I Q(H) 1=22.5 G. (b) Hyperfine splittings reported by Tuttle' for o-xylene reduced with potassium in DME at -80°C. (c) Hyperfine splittings found in this work for the o-xylene negative ion. Conditions the same as in (b).

* Supported in part through the U.S. Air Force Office of Scientific Research through Grant AF-AFOSR-285-63. Author's present address: Department of Chemistry, University of Minnesota, Minneapolis, Minnesota. 1 J. R. Bolton and A. Carrington, Mol. Phys. 4, 497 (1961). IT. R. Tuttle, Jr., and S. I. Weissman, J. Am. Chern. Soc. 80,5342 (1958). 3 V. V. Voevodskii, S. P. Solodovnikov, and V. N. Chibrikin, Doklady Akad. Nauk SSSR 129, 1082 (1959). ·T. R. Tuttle, Jr., J. Am. Chern. Soc. 84, 2839 (1962). iT. H. Brown, M. Karplus, and J. C. Schug, J. Chern. Phys. 38,1749 (1963). 8 T. H. Brown and M. Karplus, J. Chern. Phys. 39,1115 (1963). 7 J. R. Bolton, A. Carrington, A. Forman, and L. F. Orgel, Mol. Phys. 5, 43 (1962). 8 H. M. McConnell, J. Chern. Phys. 24, 633, 764 (1956); H. M. McConnell and H. H. Dearman, ibid. 28, 51 (1958); H. M. McConnell and D. B. Chesnut, ibid., p. 107. 2455


J. R. BOLTON

2456

(a)

I: i

)

5 gauss

FIG. 2. (a) First derivative of the ESR absorption spectrum of o-xylene reduced with potassium in DME at -80°C. Field increases to the right. (b) "Stick plot" reconstruction based on the split tings given in Fig. 1 (c). (c) Stick plot reconstruction based on the splittingsgiveninFig.l(b).

(bl

.1111

.Ii

II I

I

:111111 II

III

11III

1,1111

II

1111111

II

1111.

Ie)


ESR SPECTRUM OF o-XYLENE NEGATIVE ION

using the results from Ref. 1. The m-xylene negative ion should be a good model for the o-xylene negative ion; however, using Tuttle's results the spin density adds up to only 0.835. These considerations prompted us to re-examine the ESR spectrum of the o-xylene negative ion. EXPERIMENTAL

In view of the difficulty in preparing the o-xylene negative ion found previously by the author l and other investigators,2.3 a special technique was developed to remove interfering impurities both from the o-xylene and from the walls of the apparatus. o-Xylene (chromatographic purity) was obtained from Matheson, Coleman, and Bell and was used without further purification. The negative ion was prepared according to the techniques described by Bolton and Fraenke19 with the following modification: A solution of the tetracene negative ion was made in the apparatus by reacting tetracene with a sodium mirror in DME. The walls of the apparatus were thoroughly washed with this solution which was then tipped into a sidearm. By distilling DME from the side arm into the main apparatus, the remaining tetracene negative ion was rinsed into the sidearm. Potassium metal was then distilled from the end of the ESR sidearm so as to partially cover the walls of the sidearm. Finally, 30 III of o-xylene was distilled first into the tetracene solution (o-xylene does not react with the tetracene negative ion), and then the DME and o-xylene were distilled into the main apparatus. The sample tube was sealed off under vacuum. Temperature control was achieved using a flow of dry nitrogen through a coil immersed in liquid nitrogen and then passing through Dewar tubing to a quartz Dewar in the microwave cavity. The temperature was measured using a copper-constantan thermocouple. When the temperature of the quartz Dewar reached -80°C, the solution of o-xylene in DME was tipped into the side arm containing the potassium and immediately put in the cold Dewar. An ESR spectrum began to develop after about 30 min, and gradually grew in intensity over about 2 h. It then maintained a steady intensity. At this time the spectrum shown in Fig. 2 (a) was taken. The broad line marked A is apparently due to an impurity in the o-xylene. 9 J. R. Bolton and G. K. Fraenkel, (1964) .

J. Chern. Phys. 40, 1815

2457

The ESR apparatus was a superheterodyne spectrometer described previously.lO.n INTERPRETATION OF THE SPECTRUM

A "stick plot"l2 of the spectrum to be expected from Tuttle's results is given in Fig. 2 (c). In this reconstruction every group of lines consists of six lines with relative intensities 1: 3: 4: 4: 3: 1. However, a careful perusal of the spectrum shows that the groups vary both as to the number of lines and the relative intensities. A stick plot based on the splittings given in Fig. 1 (c) is given in Fig. 2 (b), and this gives a much better account of the experimental spectrum both as to line positions and intensities. It appears that the same mistake has been made here as was made with the Wurster's Blue ion spectrum,13 namely, the difference between two very similar splitting constants has been mistaken for one of the basic splittings. A least-squares analysis 12 of all the observed lines yielded the splittings in Fig. 1 (c). The 95% confidence level for the splittingconstant errors obtained from the least-squares analysis is about ±0.005 G. The g value was measured with a transfer oscillator, frequency counter, and proton resonance field indication. It was found to be 2.00288±0.00004. DISCUSSION AND CONCLUSIONS

Referring to Fig. 1, it is apparent that the new splittings for the o-xylene negative ion are in excellent agreement with Tuttle's earlier results except for the ring .positions opposite the methyl groups, where the modIfied value of 1.81 G should be used. These hyperfine splittings are much more reasonable since now the total sum of spin density is 0.992. Furthermore, the value of 1.81 G is in excellent agreement with the predictions of both molecular orbital and valence bond theories. ACKNOWLEDGMENTS

I would like to thank Professor George K. Fraenkel for many helpful discussions and Mr. M. Kaplan for the use of his computer programs. 10 J. M. Hirshon and G. K. Fraenkel, Rev. Sci. Instr 26 34 (1955). . , 11 J. W. H. Schreurs and G. K. Fraenkel J. Chern. Phys 34 756 (1961). , . , 12 The ."stic~ diagram~" !lnd the least-squares analysis of the chart calibratIOn and splittmg constants were calculated with an IBl~ 7090 computer using programs developed by M. Kaplan. J. R. Bolton, A. Carrington, and J. dos Santos-Viega, Mol. Phys. 5. 615 (1962).
