Electron Spin Resonance Study of Substituted Nitrobenzene Negative Ions Produced by Electroreduction

Transcription

1 822 Taitiro FUJINAGA, Yasuo DEGUCHI Kisaburo UMEMOTO [Vol. 37, No. 6 Electron Spin Resonance Study Substituted Nitrobenzene Negative Ions Produced by Electro By Taitiro FUJINAGA, Yasuo DEGUCHI Kisaburo UMEMOTO (Received December 14,1963) A number organic radicals, by chemical ls, have been studied by means electron spin resonance (). Since formation free radicals can be expected in course electrochemical s, it seemed to us radical formation could be conveniently detected by measurements. Austen et al.1) electrolyzed certain organic compounds, such as anthracene anthraquinone benzophenone, observed spectra free radicals in a frozen state in a dimethylformamide (DMF) solution. This first time electrolytic formation organic free radicals proved directly. However, since measurements were performed in frozen solid state, hyperfine structure (hfs) lost in broadened absorption-line width, no detailed information about electronic structure could be. A similar method used by Geske Maki,2) who carried out electrolysis in cavity recorded radical in an acetonitrile solution at room temperature. These authors3,4) also used this "in situ" method for ir study many para-substituted radicals, interesting results. Ever since, electrolytic technique has been employed by many authors5-8) as a convenient method for preparation free radicals for measurements. Recently some investigators have used measurements for elucidation electrode 1) D. E. G. Austen, P. H. Given, D. J. E. Ingram M. E. Peover, Nature. 182, 1784 (1958). 2) D. H. Geske A. H. Maki, J. Am. Chem. Soc., 82, 2671 (1960). 3) A. H. Maki D. H. Geske, J. Chem. Phys., 33, 825 (1960). 4) A. H. Maki D. H. Geske, J. Am. Chem. Soc., 83, 1852 (1961). 5) P. H. Rieger G. K. Fraenkel, J. Chem. Phys., 37, 2795 (1962) ; 39, 609 (1963). 6) J. Gendell, J. H. Freed G. K. Fraenkel, ibid., 37, 2832 (1962). 7) P. H. Rieger G. K. Fraenkel, ibid., 37, 2811 (1962). 8) I. Bernal, P. H. Rieger G. K. Fraenkel, ibid., 37, 1489 (1962). 9) P. Ludwig R. N. Adams, Anal. Chem., 34,917 (1962). 10) H. Lee R. N. Adams, ibid., 34, 1587 (1962). 11) P. H. Rieger, I. Bernal, W. H. Reinmuth G. K. Fraenkel, J. Am. Chem. Soc., 85, 683 (1963). processes by studying radicals formed during electrolysis;9,10) Rieger et al.,11) for instance, elucidated mechanism many aliphatic aromatic nitrile compounds by studying spectra at different electrode potentials. The present communication will deal investigation hfs in spectra a series substituted radicals, by electrochemical. The mechanism electrolytic some halos will also be studied.. Experimental Nitrobenzene, a series ortho-, para- substituted derivatives, in form ir chloro, bromo, iodo, methyl, methoxyl, amino, hydroxyl, cyano aldehyde compound, as well as anthraquinone naphthoquinone, were used. Most reagents were an extra pure grade were supplied by Maruwaka Chemicals Co., Osaka; an y were used out furr purification. Acetonitrile an extra pure grade used as a solvent in all measurements unless orwise mentioned. It first dried phosphorous pentoxide n anhydrous carbonate. All solutions were approximately I mm contained 0.1 M tetra-n-propylammonium perchlorate as supporting electrolyte. The electrolytic cell similar to used by Geske Maki.2) Purified nitrogen gas used for de-oxygenation solutions ; electrolysis carried out at a constant potential, value which chosen in accordance results polarographic measurements. The spectra chloro, bromo iodonitro benzene were observed in both acetonitrile a DMF solution. The DMF dried anhydrous potassium carbonate, fraction distilling at used. The spectrometer an X-b instrument, manufactured by Japan Electron Optics Co., Model JES-3B-type; it employs a rectangular reflection cavity, 100 kc./sec. field modulation, a JES electromagnet 300 mm. pole pieces. For all field measurements, peroxylamine disulfonate radical, which shows three lines a 13.0 gauss separation, due to nitrogen splitting, referred to as stard. Results Spectra.-Nitrobenzene. Discussion -The is shown in Fig. 1;

2 June, 1964] Study Nitrobenzene Derivatives 323 ( b) Fig.1. The (b) Spectrum calculated radical. based upon shown in Table I. (b) (c) Fig. 2. spectra p- (b) m- (c) o-nitrotoluene radicals.

3 824 Taitiro FUJINAGA, TABLE I. Yasuo DEGUCHI HYPERFINE it is very similar to obserbed by Geske Maki2) though re is a slight difference in hyperfine. This difference might be due to presence traces water in solvent used, in accordance results we reported previously dealing solvent effect on width line due to nitrogen in cases diphenyl oxide.13) tion nitric oxide12) di-p-anisyl nitric It shown re nitrogenconstant in an aqueous solularger than in acetonit- rile. Piette et al.14) also examined dependence nitrogen constant nitro compounds on water content solvent showed presence small 12) Y. Deguchi, This Bulletin, 35, 260 (1962). 13) K. Umemoto, Y. Deguchi H. Takagi, ibid., 36, 560 (1963). 14) L. H. Piette, P. Ludwig R. N. Adams, J. Am. Chem. Soc., 84, 4212 (1962). Kisaburo COUPLING [Vol. UMEMOTO 37, No. 6 CONSTANTS quantities water considerably increases magnitudes constant. Nitrotoluene.-The spectra electrolytically reduced o-, m- p-nitrotoluenes are shown in Figs. 2a, 2b 2c. The given in Table I are in good agreement observed spectra. The were assigned to each proton by referring to results Hiickel's molecular orbital (MO) method calculation. It found all three methyl protons had same constant; this shows re is a hyperconjugative interaction between methyl group rotating at high speed ring π system. would In expect case o-nitrotoluene, molecule will one have an unequal spin distribution in ring system molecular asymmetry; however, two protons in position to nitro group showed same constant.

4 June, 1964] Study Nitrobenzene 825 Derivatives (b) Fig. 3. (c) spectra p- (b) m- (c) o-nitrochlorobenzene In case m-nitrotoluene, two ortho protons were found to be equivalent. Hence, spin distribution in molecule seems to be little affected by substitution a ring proton a methyl group. The nitrogen constats are a little larger than, showing methyl group has an electron-repelling effect. The fact ortho para s have a larger nitrogen-splitting constant than compound explains why electronrepelling effect methyl group is directed towards ortho para positions. Chloro.-The spectra reduced ortho, para s are shown in Figs. 3a, 3b 3c. Once again shown in Table I agree well observed to obtain an hfs Vivo,15) who radicals. spectra. studied It not possible chlorine. Wertz spectra chloro compounds in an aqueous solution, were also unable to observe hfs chlorine small nuclear magnetic moment this atom. As in case nitrotoluene, two protons ortho two ortho protons have same constant. Hence, chlorosubstitution bution seems to molecule affect very little. spin The gen- were somewhat than electron-attracting effect 15) J. E. Wertz (1955). J. L. Vivo, J. Chem. Phys., distrinitro- smaller chlorine 23, 2441

5 826 Taitiro TABLE * Data from II. FUJINAGA, Yasuo 14N COUPLING DEGUCHI CONSTANTS Kisaburo OF SUBSTITUTED UMEMOTO [Vol. 37, No. 6 NITROBENZENE Ref. 3. atom. The nitrogen- ortho s nitrotoluene nitrochlorobenzene are little larger than those corresponding para s, indicating a small steric effect. It has been pointed out by Geske Ragel16) an value in an aromatic radical approaches aliphatic counterpart nitrogroup is twisted out plane benzene ring by steric effects bulky substituents. Bromo.-The spectra reduced para s were very similar to those corresponding chloro compounds. However, ortho showed same as radical, indicating elimination bromine atom occurs prior to formation free radical. Iodo.-The electrolytic products ortho, para s all showed same as radical. This once again indicates iodine is eliminated prior to nitro group. Or Substituted Nitrobenzenes.-The or substituted radicals are also listed in Table I. The different nitrogen- are given in Table II; substituents are arranged according to ir increasing electron-attracting effect. Electron-attracting substituents generally reduce nitrogen- constant, whereas electron-repelling substituents increase it. In both cases, electron effect is more pronounced in case ortho para than in case substitution. onitrobenzaldehyde gave an abnormally large nitrogen-splitting constant, usual order observed electron-attracting groups, viz. para<ortho<, does not hold; this seems to indicate a strong ortho effect aldehyde group. As regards nitrophenol, authors have discussed effect hydrogen bonding in a previous publication.17) Huckel's Molecular Orbital Calculations.Using shown in Table I, spin density distribution could be calculated from following relation: aia= 16) D. H. Geske L. Ragle, J. Am. Chem. Soc., 83, 3532 (1961). 17) T. Fujinaga, Y. Deguchi K. Umemoto, This Bulletin, 36, 1539 (1963). QCHHρiπ, methyl where group, QCHH=28 ach3h=qcch3hρ gauss18) π, where, for QCCH3H= 30 gauss.19) By Huckel's MO method, spin density in molecules could also be calculated. The results are listed in Table III, where parameters employed in this calculation are also listed. The calculated spin density in ortho para positions relative to nitro group is in fairly good agreement experimental values; by contrast calculated position is not in agreement experimental results. Rieger Fraenkels) calculated spin density distribution for some nitrile nitro compounds by both Huckel's MO method McLachlan's procedure, which employs concept approximate configuration interaction ; se authors also found Huckel's MO method generally gave good results for ortho- para- but not for -substitution, while McLachlan's approximation method generally gave very good agreement experimental results. Electrochemical Process. -Nitrobenzene. -The radical generated by electrolysis at -1.3V. vs. SCE (saturated calomel electrode), is, at a potential slightly more negative than first wave in polarogram. The radical stable for about 30 min. It is known at potential second wave mono negative ion is furr reduced to dinegative radical, which is generally rapidly protonated by reaction solvent. Therefore, no would be cxpected a priori. However, an observed in potential range second wave, although intensity reduced to some extent. This shows process corresponding to second wave is slower than radical formation. When phenol added as a proton donor to 1 mm DMF solution, radical formation at -1.3 V. vs. SCE not affected until concentration added phenol reached about 2 mm; at higher potentials, stability radical decreased. Upon furr addition phenol, rate radical formation at potential first wave 18) H. S. Jarrett, J. Chem. Phys., 25, 1289 (1956). 19) D. J. E. Ingram, " Free Radicals as Studied by Electron Spin Resonance," Butterworths Scientific Publications, London (1958), p. 116.

6 June, 1964] TABLE III. CALCULATED SPIN AND Study Nitrobenzene EXPERIMENTAL DENSITY Derivatives 827 decreased as well, indicating rapid protonation, hence, a decreased life time free radical. Chloro.-The reducttion products three s chloro, prepared by electrolysis at -1.3 V. vs. SCE in an acetonitrile solution, showed an. In DMF para s gave ir characteristic spectra. When applied potential changed from -1.2 V. to -2.0 V. vs. SCE generated radical species showed little change, although intensity decreased at potentials lower than -1.8 V. vs. SCE. When ortho electrolyzed at -1.3 V. vs. SCE in a 1 mm DMF solution, its could be observed soon after electrolysis began, but after about fifteen minutes showed beginning formation radical, which after 45 min. major component present. When applied voltage changed to -1.5 V. vs. SCE became identical radical, indicating a dehalogenation reaction according to: The ease dehalogenation in case ortho might be due to steric reasons. When phenol added in about an equimolar quantity, dehalogenation impeded, a quantity about three times greater added, no dehalogenation observed at potentials above -1.7V. vs. SCE ; no radical formation could be detected at lower potentials tonation. Bromo.-The para characteristic spectra electrolysis at -1.3V. vs. s SCE in trile solution, whereas under same conditions, dehalogenation in DMF In not Parameters used in this calculation are as follows. show DMF any solution sign enhanced proproducts gave ir prepared by an acetoni- ortho derivative, showed, indicating solution. dehalogenation in did potential range from -1.1 V. to -2.0 V. vs. SCE, whereas ortho completely dehalogenated. The product para showed its own characteristic prepared by electrolysis at -1.0 V. vs. SCE in a 1 mm solution; at V. vs. SCE beginning formation 20) 21) 1953, 22) J. I. F. Alonso., Compt. rend., 233, 403 (1951). C. A. Coulson U. A. Crawford, J. Chem H. H. Jaffb, J. Chem. Phys., 20, 279 (1952). Soc.,

7 828 [Vol. 37, No. 6 radical could be observed, while as electrolysis proceeded, formation increasing quantities could be seen. At -1.6 V. vs. SCE a clear alone could be observed. When concentration para raised to 5 mm, reduced para alone observed in potential region from -1.0 V. to -1.4 V. vs. SCE, whereas at lower potentials dehalogenation became evident. These results suggets following electrode process: Indo.-All three s showed formation radical, even in a 5 mm solution at -1.2 V. vs. SCE both in acetonitrile in DMF solution. Hence, complete dehalogenation occurred in all cases. The reaction can be represented by: The experiments show ease which halos are dehalogenated during electrolytic follows order: p,m-cl, p, m, o-i m-br<o-cl<p-br<o-br, This series seems to indicate ease dehalogenation depends on steric considerations. Details polarographic studies halos will be reported later.23) The authors wish to acknowledge helpful suggestions Visiting Pressor B. Breyer University Sydney. Chemistry Institute Faculty Science Kyoto University Sakyo-ku, Kyoto 23) T. Fujinaga T. Arai, J. Chem. Soc. Japan, Pure Chem. Sec. (Nippon Kagaku Zasshi), in press.