219 J. Electroanal. Chem., 236 (1987) 219-238 Elsevier Sequoia S.A., Lausanne - Printed ELECTROCHEMISTRY IN ACETONITRILE MECHANISTIC SUBSTITUENT ERIC W. TSAI and MARTIN Department (Received OF SOME N-SUBSTITUTED ASPECTS, EFFECTS l, MALGORZATA POMERANTZ of Chemisiry 23rd January in The Netherlands zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP l ELECTROPHORIC K. WITCZAK l*, PHOSPHA-X5-AZENES GROUP KRISHNAN BEHAVIOR, RAJESHWAR AND *** ** The Unrversrty of Texas at Arlmgton, Arlrngton. TX 76019-006.5 (U.S.A.) 1987; in revised form 19th May 1987) ABSTRACT Cyclic voltammetry (CV) data are presented for the oxidation and reduction of a series of N-arylP, P, P-tnphenylphospha-X5-azenes. The electrochemical behavior of these compounds is compared with the corresponding N-sulphonyland N-acyl-compound series. In all three cases, the electronic mfluence of the substituent, R, in the aryl ring system is quantified in terms of Zuman plots of the CV oxidation or reduction potential vs. the Hammett substituent constant. It is shown that insertion of the SO, and CO spacer groups between the aryl ring system and the N reaction site attenuates the electronic influence of R on the electrochemtcal oxidation response. The latter, therefore, is largely centered at the N reaction site. By way of contrast, the influence of R on the reduction response is minimal. These observations lead to the categorization of two types of electrochemical responses. One is intrinsic to the molecule as a whole, and the other is caused by the presence of an ‘electrophoric” group. We show herem how these N-substituted phospha-X5-axenes represent good model systems for defining and dlustrating the concept of an electrophoric group; i.e. a portion of a molecule giving rise to an electrochemical response that is characteristic (fingerprint) of that moiety itself rather than of the molecule as a whole. Finally, the oxidation mechanism involving dimerization of the parent compound in the N-aryl series is elucidated via the combined use of CV, constant potential coulometry, “P NMR spectroscopy, and HPLC. Similarities in the oxidation behavior of this compound with anilines are pomted out. Taken in part from the dissertation of E.W.T. submitted to the Graduate School of UTA in partial fulfillment for the D.Sc. degree, 1986. l * On leave from the Center of Molecular and Macromolecular Studies, Pohsh Academy of Sciences, Lodz, Poland. Current address: ZakIad Chemii FarmaceutyctyczneJ. Akademia Medyczna, Lodz, Poland. l ** To whom correspondence should be addressed. l 0022-0728/87/$03.50 0 1987 Elsevier Sequoia S.A. 220 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA INTRODUCTION zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Phospha-X5-azenes are an important class of compounds for both theoretical and practical reasons. They are useful as fungicides, herbicides, defoliating agents, oil additives, light stabilizers and flame retardants in rayon [l]. Polymeric phospha-X5azenes are used for hoses, gaskets, in non-burning foam rubber articles [2,3], and as lubricants, fibers, and varnishes [4]. With amine side chains they are water soluble, and are being examined as potential carriers for biologically-active molecules [2,5]. The recently discovered high ionic mobility in these compounds [6] additionally opens up a host of application possibilities in the energy storage and chemical sensor areas. In this paper, we present an electrochemical study of three series of phospha-A5azenas, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1, 2, and 3: R 2 1 aR=p-NO, eR=p-CO,CH, j bR=m-NO, f R=p-Br k R=p-OCH, c R = p- CN g R = p- C1 I d R = p-CF, h R = p-F m R=p-NM e, zyxwvutsrqponmlkjihgfed 3 R=p-CH, R=p-NH, i R=H Specifically, we show that: (a) notwithstanding the presence in these compounds of two potential sites for electron removal and addition, namely N and P, 1, 2, and 3 may be viewed in an electrochemical sense as essentially equivalent to the corresponding anilines, sulphonamides, and carboxamides respectively; (b) 1, 2, and 3 are good model systems for the concept of an “ electrophoric” group, which as defined previously [7], is a portion of a molecule that gives rise to an electrochemical response that is characteristic (fingerprint) of that moiety itself rather than of the molecule as a whole; and (c) the electronic influence of the substituent in the N-aryl ring system in 1 is attenuated by the insertion of the SO, and CO spacer groups as in 2 and 3 respectively. In a previous paper [8], we presented briefly the oxidation of 1 and its derivatives and reported that 1 was largely resistant to electrochemical reduction except for the cyano- and nitro-derivatives. In this paper, we describe in more detail the oxidation mechanism of the parent, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB li, and also show how the reduction response is modified by the presence of the SO, and CO electrophoric groups. Few electrochemical studies exist on compounds of the type 1, 2, or 3. However, the electrochemical reduction of li has been reported by Pak and Gulick [9]. Polarographic reduction of the nitro-derivative, la, has been described by Penkovskii et al. [lo]. The reduction of cyclic phosphazenes has been studied by previous authors [3,11], and triphenylphosphine has undergone electrochemical scrutiny by 221 Bontempelli and co-workers [12] and Bard and co-workers pertinent studies will be referenced at appropriate junctures [13]. These and other in what follows. EXPERIMENTAL Methodology Cyclic voltammetry (CV) and constant-potential coulometry (CPC) were performed in dry acetonitrile obtained by double-distillation of HPLC-grade solvent (Fisher Scientific) from PzO,. Tetrabutylammonium perchlorate (TBAP) (Southwestern Analytical Chemicals) was used as supporting electrolyte (0.1 M). The usual three-electrode cell geometry with positive feedback iR compensation was used in all the cases for CV. The working electrode was either Pt or C and had a nominal geometric area of 0.2 cm’ for CV, and ca. 10 cm2 for CPC. Pretreatment of the working electrode surfaces prior to use followed usual procedures [14]. Frequent monitoring of the cyclic voltammogram peak shape and peak separation potential, A E, for the ferrocene/ ferricenium ion redox couple, provided a convenient means of assessing the efficacy of these pretreatment procedures. The reference electrode was Ag/O.l M Ag+ in acetonitrile; all potentials in this study are reported with respect to this reference. The counterelectrode was a Pt spiral. The potential scan-rate was 0.1 V/s unless otherwise stated. The sample concentration was in the 1 mM to 5 mM range. An EG&G (Princeton Applied Research) electrochemistry system assembled from Model 173, 179, and 175 modules was used in conjunction with a Houston Instrument Model 2000 X-Y recorder for acquiring CV data. Fast-scan (> 0.5 V/s) cyclic voltammograms were recorded on a Tektronix Model 5441 storage oscilloscope. The CV oxidation and reduction response of each compound was probed in the usual manner by scanning first positive up to ca. 2.5 V and then in the negative were performed at direction to ca. - 3.0 V starting from 0 V. All measurements ambient temperature. Chromatographic analyses were performed on a Waters Model 204 HPLC system fitted with a C,, reverse-phase column. The mobile phase composition was methanol + 15% acetic acid (SO/SO v/v) and was delivered at a nominal flow rate of 4 ml/mm. 31P NMR spectra were taken at 80.99 MHz on a Nicolet NT-200 wide-bore spectrometer. External 85% H,PO, standard, CDCl, solvent, and 12 mm tubes were employed for this purpose. The nominal sample concentration was 40 mg/ml. The chemical shifts reported herein are not corrected for concentration effects (cf. refs. 8, 15), except for the iV-acylphospha-X5-azenes (3). The ‘H NMR spectra were obtained on either a 60 MHz Varian T-60 spectrometer or on the aforementioned 200 MHz wide-bore system. IR spectra were obtained on a PerkinElmer 599 B spectrometer using KBr discs. UV/vis spectra of the electrolyzed and parent compounds were obtained on a Cary 14 instrument. Elemental analyses were by Galbraith Laboratories, Knoxville. TN. Melting points are not corrected. 222 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA C~mp5und sy ntheses N- Ary lsuljony l- P, P,P- tripheny lph5spha- ~‘- azenes (2) were prepared as described previously [16,17]. All new compounds were completely characterized spectroscopically and yielded correct C, H, and N elemental analyses. N- Acy l- P,P,P- tripheny lphospha- A5- azenes (3). General. Two different procedures were employed for the syntheses of the derivatives of 3: Procedure A [26]. To a stirred and cooled solution of the benzamide (1 or 2 mmol) and t~phenylphosp~ne (1 or 2 mmol) in dry tetrahydrofuran (5 or 10 ml; distilled from LiAlH,) under dry argon, was added diethyl az~ic~boxylate (1 or 2 mmol) dropwise, by syringe, through a rubber septum. The solution was allowed to warm to room temperature and then to stir for 12 hr. The solvent was removed in vacuum, 2 ml of MeOH was added and the solid which formed was filtered and recrystallized from the solvent indicated. Procedure B [I??/. To a cold (5-10 ’ C), stirred solution of triphenylphosphine in dry tetrahydrofuran (distilled from LiAlH,) was added, dropwise, the aroyl azide, prepared from the corresponding aroyl chloride and sodium azide [19,20], in dry tetrahydrofuran. The mixture was stirred for 4 additional hours, warmed to 40” C, and stirred for 1 more hour. The solvent was removed in vacuum and the residue recrystallized from the indicated solvent. N- p- Nitrobenzoy I- P,P,P- tripheny ~phuspha- ~~- azene (34. Procedure A was used. 0.333 g (2 mmol) of p-nitrobenzamide, 0.524 g (2 mmol) of t~phenylphosp~e, and 0.34% g (2 mmol) of diethyl azodicarboxylate gave 0.63 g of zyxwvutsrqponmlkjihgfedcbaZ 3a (recryst. from MeOH); 74%; m.p. 201-205“C (ref. 16, 206-207°C). IR (KBr): F/cm-‘; 1585 (C=O); 1530 (NO,); 1330 (P=N). 31P NMR (CDCl,): S 22.27. N- p- Cy anobenzoy l- P,P,P- tripheny lphospha- ~sJazene (3~). Procedure A was used. From 0.146 g (1 mmol) of p-cyanobenzamide, 0.262 g (1 mmol) of triphenylphosphine, and 0.174 g (1 mmol) of diethylazodicarboxylate there was produced 0.250 g of 3c (62%) which was recrystallized from MeOH; m.p. 178-179OC. IR (KBr): v/cm-‘; 3060 (C-H); 2235 (C=H); 1590 (C=O); 1340 (P=N). ‘H NMR (CDCI,); 6 7.0-8.0 (m, 17 H, ArH), 8.27 (d, 2 H, ArH, J = 8 Hz). 31P NMR (CDCL,): 6 22.21. Anal. calcd. for C,H,,N,OP: C, 76.83; H, 4.71; N, 6.89; found: C, 76.80; H, 4.91; N, 6.61. N- p- Bromobenzoy l- P,P,P- tripheny iphospha- X5- azene zyxwvutsrqponmlkjihgfedcbaZYXWVUT (3 f). Procedure B was used. 0.345 g (1.5 mmol) of p-bromobenzoyl azide and 0.400 g (1.5 mmol) of triphenylphosphine in 10 ml of dry THF gave, after recrystallization from MeOH, 0.43 g (61%) of 3f; m.p. 145-146°C (ref. 21, 140-142°C). IR (KBr): ?/cm-‘; 3050 (C-H); 1600 (C=O); 1330 (EN). ‘H NMR (CDCI,): 6 7.3-8.1 (m, 17 H, ArH), 8.17 (d, 2 H, ArH, J = 8 Hz). 3’P NMR (CDCI,): S 21.48. 223 zyxwvutsr N- p- Chlorobenzoy l- P,P,P- tripheny lphospha- A5- azene (3g). Procedure B was used. From 0.250 g (1.4 mmol) of p-chlorobenzoyl azide and 0.356 g (1.4 mmol) of triphenylphosphine in 10 ml of THF was obtained, after recrystallization from MeOH, 0.30 g of 3g (53%); m.p. 146-150°C * (ref. 22, 152-154” C). N-p-Fluorobenzoyl-P, P, P-triphenylphospha-X-azene (3h). Procedure B was used. From 0.248 (1.5 mmol) of p-fluorobenzoyl azide [23] and 0.390 (1.5 mmol) of triphenylphosphine in 10 ml of THF was obtained 0.31 g of 3h (52%) m.p. 178-180°C after recrystallization from MeOH. IR (KBr): S/cm-‘; 3055 (C-H); 1608 (C=O); 1330 (P=N). ‘H NMR (CDCl,): S 7.02 (t, 2 H, ArH, J = 8 Hz); 7.2-8.1 (m, 15 H, ArH); 8.31 (dd, 2 H, ArH, J = 6 and 8 Hz). 31P NMR (CDCl,): 6 21.27. Anal. calcd. for C,,H,,FNOP: C, 75.18; H, 4.79; N, 3.51; found: C, 75.53; H, 4.89; N, 3.38. N- Benzoy l- P,P,P- tripheny lphospha- A5- azene (3i). Procedure A was used. Reaction of 0.121 g (1.0 mmol) of benzamide and 0.262 g (1.0 mmol) of triphenylphosphine with 0.174 g (1.0 mmol) of diethyl azodicarboxylate gave 0.136 g of 3i (36%) which was recrystallized from MeOH, m.p. 190-195°C (ref. 16, 195-196°C). N-p-Toluoyl-P,P,P-triphenylphospha-X5-azene (3j). Procedure B was used. Reaction of 1.64 (10.0 mmol) of p-toluoyl azide with 2.62 g (10.0 mmol) of triphenylphosphine in 20 ml of THF gave 2.85 g of 3j after recrystallization from MeOH (73%); m.p. 153-155°C (ref. 16, 152-153°C). 3’P NMR (CDCI,): S 20.87. N-p-Anisoyl-P,P,P-triphenylphospha-h5-azene (3k). Procedure B was used. Reaction of 0.885 g (5 mmol) of p-anisoyl azide with 1.31 g (5 mmol) of triphenylphosphine in 15 ml of THF produced, after recrystallization from ethyl acetate, 1.3 g of 3k (63%); m.p. 158-16O’C. IR (KBr): ?/cm-‘; 3060, 2840 (C-H); 1605, 1595 (C=O); 1330 (P=N). ‘H NMR (CDCI,): 6 3.83 (s, 3 H, OCH,), 6.87 (br. d, 2 H, ArH, J = 9 Hz), 7.2-8.0 (m, 15 H, ArH), 8.27 (d, 2 H, ArH, J = 9 Hz). 31P NMR (CDCl,): 6 20.71. Anal. calcd. for C26H22NOZP: C, 75.90; H, 5.39; N, 3.40; found: C, 75.58; H, 5.68; N, 3.39. N,N ‘- (4,4’- Biphenvldiy l)- bis(P,P,P- tripheny lphospha- X5- azene) (5) was prepared by the method of Horner and Oediger [24]. To a stirred solution of 10.48 g (40.0 mmol) of triphenylphosphine in 100 ml of Ccl,, cooled to O-5” C (ice-salt bath), was added 6.4 g (40.0 mmol) of Br, in 25 ml of Ccl,, dropwise. To this suspension was added 15 g of Et,N in 20 mL of Ccl, and the ice-salt bath was removed. Then 3.66 g (40.0 mmol) of benzidine was added and the solution was refluxed with vigorous stirring for 30 min. The mixture was filtered and washed 3 times with 25 ml of Ccl.. The solid was stirred with 100 ml of 30% NaOH, warmed with MeOH (50aC) The melting pomt of the chloro-phosphazene have been 152-153O C. l 3g was incorrectly reported m Table 1 of ref. 16. It should 224 for 20 min, filtered and this crude product was recrystallized from n-butanol/ chlorobenzene (4 : 1). Yield: 8.9 g (32%), m.p. 266-268” C (ref. 24, 269-270 o C). IR 3050,302O (C-H); 1323 (P=N). ‘H NMR (CDCl,): 6 6.76 (d, 4 H, (KBr): p/cm-‘; ArH, J = 8 Hz); 7.17 (d, 4 H ArH, J = 8 Hz); 7.35-7.9 (m, 30 H, ArH). 31P NMR (CDCl,): S 3.89. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDC RESULTS AND DISCUSSION Mechanistic aspects As reported previously [8], the oxidation of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO li is electrochemically irreversible for potential scan-rates up to at least 10 V/s. Figure 1 illustrates the CV oxidation behavior of li at a scan-rate of 0.1 V/s in CH,CN + 0.1 M TBAP. The peak current corresponding to the major oxidation wave at ca. 0.60 V is proportional to u112 (U = potential scan-rate), Fig. 1 (inset). This is diagnostic of the absence of adsorption complications (but see below). Repeated scanning however, results in the evolution of a pair of reversible CV waves at ca. 0.35 V and an irreversible electrochemical response centered at - 0.35 V. A typical example is shown in Fig. 1. These additional features obviously reflect the consequence of chemistry accompa- 1 400 ‘P/PA 200 100 3000 LL 0 0.1 0.2 v’ ‘/ 0.4 0.5 I 1 0.6 0 3 (V/S)“2 0.6 04 I 0.2 Polenl~al/V 0 B -0.2 -0.4 AgIAg+ Fig. 1. Cyclic voltammograms of Ii m M eCN+O.l M TBAP. The first and steady-state cyclic scans are shown. The mset shows a plot of peak current vs. the square-root of potential scan-rate. 225 nying initial electron-transfer from zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG li. The excellent correlation that was previously observed ‘by us [8] between the CV oxidation potentials for li and its derivatives and the reported polarographic half-wave potentials for the corresponding anilines [25], was taken as suggestive of similarities in the oxidation mechanism in the two cases [8]. We now present more detailed evidence that this is indeed the case. First, the possibility that the initial oxidation of li involves the abstraction of electrons in P-centered orbitals, can be dismissed by the fact that the CV oxidation potential, Epa, for Ph,P is 0.83 V at 0.1 V/s [12], a value which is obviously quite disparate from the 0.47 V measured for li, Fig. 1. Our previous MO calculations [8] also argue against the participation of P-centered orbitals in the electrochemical oxidation. Assuming therefore for now, that electrooxidation of li involves the removal of electron(s) from the N lone-pair, we have reaction (1) as the primary oxidation step: Ph3P=i H -e- H The cation-radical formed in reaction such as those shown below (4a-d): H - Ph&ij / a- 4a (1) may be represented by resonance forms H \ - 4b H - Ptl&ti h - 4c \ - * Pt+-Fi =e l-l - 4d Of these, 4d is likely to predominate in terms of relative contributions to the resonance hybrid (see below). Two of these units can undergo tail-to-tail coupling via reaction (2) to yield the dimer, 5: This dimer, because of the possibility of extended conjugation, would be readily oxidized via an ECE mechanism, at potentials more negative than that of the parent compound, li, reaction (3): 226 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Another route to the electrogeneration of 5 is the initial abstraction of both of the N lone pair electrons to yield the dispositive species, 6, via reaction (4): Ph3P=Fj ’ -0 - \ Ph,P=,+=& + Ze- - (4) 6 Subsequent coupling of 6 with the neutral compound (reaction 5): zyxwvutsrqponmlkjihgfedcbaZYXWV I i would yield the dimer 5 Subsequent oxidation of 5 would be as before (cf. reaction 3). Precedence exists for this pathway in the case of dimethylaniline, which has been claimed to be oxidized by an initial 2-electron process with subsequent coupling to a neutral parent molecule [26]. The two alternative possibilities can be discriminated by Tafel-type analyses of the CV data. Figure 2a shows a log i-E plot for zyxwvutsrqponmlkjihgfedcbaZY li. The slope of this plot, which corresponds to anF/2.303RT (n is the charge number of the reaction), yields an an value of 0.49. Figure 2b shows a plot of log i vs. log c, where c is the variable bulk concentration of li. The least-squares fitted line yields a slope (reaction order) of 0.97. Taken as a unit, these CV analyses are consistent with a first-order reaction with 1 e-transfer in the rate-determining step. Thus, the intermediate formation of 6 (involving 2 e) is considered an unlikely possibility. Head-to-tail radical coupling formally involving 4b and 4d (reaction 6), would yield cation 7, which should be less susceptible to oxidation than the starting compound, li. (6) 7 In contrast to the previous ECE case (formation of 5), the expected mechanism is an EC dimerization pathway. This type of head-to-tail coupling is preponderant in the unsubstituted aniline case [26]. To elucidate the oxidation mechanism of li better, an authentic sample of 5 was synthesized (cf. Experimental) and characterized by CV. Figure 3 summarizes the 221 (4 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 2.40 1 160 ‘7 0.35 0.37 0.39 . Potential/V vs. AgIAg+ (W 3.00 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 2.60 ;i 2.60 a .- 2.40 %I S 2.20 1.603, -2.00 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM log(c/ moll-‘1 Fig. 2. (a) Tafel plot for zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA li. Raw CV data (not shown) were obtained for this plot on a 5 X zyxwvutsrqponmlk 10e3 M solution of li. The solid line is a least-squares fit of the data (r = 0.989). (b) Plot of log i vs. log c for li. The sohd line is a least-squares fit of the data (r = 0.998). salient features in the electrochemistry of 5. A set of kinetically-sluggish CV waves at 0.36 V (very close to the CV feature seen on oxidation of li, Fig. 1) diminishes in amplitude systematically and becomes more drawn out on repeated cycling, Fig. 3a. If the scan is carried past the irreversible anodic wave at 1.58 V, this set of CV waves reappears as shown in Fig. 3b. An irreversible anodic wave is also seen for li at 1.60 V; this feature is not shown in Fig. 1. Constant-potential coulometry on li further illustrates the complexity of its oxidation mechanism. The CPC analyses were carried out at 0.60 V (cf. Fig. 1). The n values decreased systematically with increasing concentration of li; a typical trend is contained in Fig. 4. “ The inset illustrates the morphology of log i-t plots and their dependence on the initial concentration of li. The non-linearity of these plots is clearly diagnostic of coupled chemical reactions accompanying the electrochemistry [27]. The original colorless solution turned light-orange to a dark-brown hue depending upon the initial concentration of li. Concomitantly, the UV/ vis 228 I 2.0 1.6 1.2 0.8 Potential/V 0.4 0 -0.4 Irg, AglAg+ Fig. 3. (a) CV of 5 mM solution of 5 in M eCN +O.l M TBAP from -0.60 to 1.00 V at 0.100 V/s. The numbers refer to the scan sequence (1 = first cycle, etc.). (b) CV scan on the same solution over a wider potential window from -0.60 to 1.90 V at 0.100 V/s. The dashed line mirrors a terminal scan in (a). spectrum of the electrolyzed product showed a broad featureless absorption commencing at 600-650 nm. Perceptible filming of the electrode surface was also observed at the higher concentrations shown in Fig. 4. We interpret these data in terms of oligomer formation, and subsequent oxidative cleavage of these species to regenerate the dimeric species (eqn. 5) in the oxidized form. The wave at + 1.60 V is assigned to this latter process. Oligomer formation has ample precedence in the aniline system [28]. Solutions of the electrolyzed starting compound, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ li, the electrolyzed dimer, 5, and the neutral compounds li and 5 were subjected to 31 P NMR analyses. Solutions for 229 zyxwvutsrq l.O- zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1 l 0.6 - zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0.6 - zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA n 4 l l 0.4 - . 0.2- ot -5.5 1 I -5.0 -4.5 -4.0 -3.5 log(c/ moll-‘1 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM Fig. 4. Dependence of the CPC n value on concentration for li. The inset shows log I-C plots with li concentration as a parameter. The labels denote corresponding data. NMR analyses on the electrolyzed samples were prepared in CDCl, after vacuum evaporation of MeCN. Two drops of acetic acid were added in each case to enhance sample solubility. Chemical shifts of 6 33.04, 32.88, 3.73, and 3.89 ppm (averaged from three replicate measurements) were obtained for these samples, respectively. The first two values fall in the range characteristic of phosphonium compounds such as 8, where SS1, is 32.0 [29]. Ph,P+-NHPh Br- 8 The downfield shift relative to the neutral species, li and 5, is also consistent with the lowered electron density at P. The slight difference between the chemical shifts in the first two samples is most likely due to concentration effects (cf. ref. 15 and see above). Reversed-phase HPLC analyses support the formation of polar species from the oxidation of li. The chromatograms for the electrolyzed sample of 5 also show a marked resemblance to those obtained for electrolyzed li. From these and the electrochemical data which were discussed above, it seems safe to conclude that the electrooxidation of li proceeds by a complex mechanism generating the dimeric species, 5 and 7. Further coupling of these dimers with the parent (neutral species) generates oligomers. High molar mass polymer formation via the electrochemical 230 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA route, however, is impeded by the proclivity of these oligomers to undergo oxidative cleavage to regenerate the dimeric species (see above). Experiments are in progress to unravel the details of this complex mechanism further. In particular, we have not yet addressed the question as to what extent the head-to-tail coupling route (leading to 7) competes with the tail-to-tail route (generating 5). Note that the Tafel data, Fig. 2, would be consistent with both tail-to-tail as well as head-to-tail coupling routes. The N-substituted anilines such as N,N-dimethylaniline and triphenylanine are known to couple predominantly tailto-tail [26]. On the other hand, the subsequent oligomer formation observed by us for zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA li, is more reminiscent of the electrochemical behavior of unsubstituted anilines. zyxwvutsrq Electrophoric group behavior and substituent effects In Table 1 we show the CV oxidation and reduction potentials for zyxwvutsrqponmlkj 1 [8] and corresponding N-arylsulfonyl- and N-aroyl-P, P, P-triphenylphospha-A5-azenes (2 and 3) respectively. These compounds may be ordered in terms of relative ease of TABLE 1 Cyclic voltammetry X5-azenes Compound (CV) oxidation and reduction 1’ ; g h i j k I m a of N-substituted-P, P, P-triphenylphospha- E,, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ /V vs. Ago’+ b E,,/V vs. Ago’+ b Substituent a b c d potentials P-NO, m-NO, p-CN P-CF, 2 3 lC 2 0.84 2.22 1.83 - 1.73 - 2.79 * - 2.88 * 0.75 2.21 nd ’ - 1.61 - 2.78 * nd e 0.74 2.20 1.82 -1.77 - 2.77 0.74 nd e nd e N= nd e 3 f - 2.79 f nd e nd e 2.17 nd e nd e -2.76 f nd e nd e nd e nd e - 2.62 f -2.80 f 1.73 0.51 2.10 nr e -2.70 f 1.73 p- Cl -2.80 ’ 1.73 nr e 0.48 2.10 - 2.75 - 2.80 P-F 0.47 2.07 H - 2.83 1.71 nr e - 2.74 0.37 1.99 1.65 nr ’ - 2.76 - 2.83 P-CH3 nr e 0.21 1.70 g 1.35 g - 2.90 - 2.82 P-~H, nd e 0.77 h nd e nd e - 2.72 nd e P-NH, -0.20 ’ nd ’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA p-NM e, nd e nr e nd e nd e p-COOCH p-Br 3 a Nominal error is f 10 mV. b Oxidation and reduction potentials, at 0.10 V/s scan rate in the positive’ Taken from ref. 2. E,,, and E,, taken as peak potentials (E,, and E,, respectively) and negative-going CV scans, respectively (see Experimental). d Third wave in CV reduction branch, cf. Figs. 6b and 7b. e “ml” denotes not determined and f Second “N” denotes no reduction response up to at least - 3.00 V. wave in CV reductron branch. s This potential refers to the oxidation of the -OCH, h Thrs potential refers to the oxidation of the -NH, ’ This potential refers to the oxidation of the -NM e, electrophore, refer to the text. electrophore. electrophore, cf. Fig. 5a. 231 oxidation: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1~ 3 > 2. Deferring for the moment the electronic influence of the R substituent on these potentials, three other points are worthy of note in the data collected in Table 1: (a) electrooxidizable entities such as -NMe, in the case of 1 and -NH, in the case of 2 show noticeably disparate Eo, values, (b) similarly, electroreducible moieties such as NO, and CN modify (see below) the reduction-resistant response which is characteristic of the Nary1 compounds, 1, and (c) after allowing for the influence of these electroreducible entities, the reduction responses of 1, 2, and 3 reveal remarkably uniform patterns across each series. Thus, compounds in series 1 are resistant to reduction up to at least -3.00 V, impounds in series 2 reduce at - 2.76 + 0.14 V, and those in series 3 reduce at - 2.80 + 0.04 V. (b) 1 1 1 I zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFED zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 .3 0 .1 -0 .1 -0 .3 9 .5 -0 .7 a l.0 -1 .4 -1 .8 -2 .2 -2 .6 -3 .0 L 0.5 I I Pot e nt ia l]V vs. Ag* ‘O Potential/V vs Ag/Ag* Fig, 5. (a) Cyclic voltammogram of tetramethylphenylenediamine in M eCN+O.l voltammogram M TBAP. (b) Cyclic zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA of lm in the same electrolyte. Fig. 6. Comparison of the CV reduction response for selected compounds in the 2 series! (a) Zi: (b) 2a; Cc) 2~ (4 2~ 232 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA The latter two values, of course, represent the reduction potentials of the SO, and CO moieties respectively. These points when taken collectively, define and illustrate the concept of an “ “ electrophoric group” in organic electrochemistry [7,30]. Analogous to the role of a functional group in IR spectroscopy, an electrophoric group manifests a fingerprintable response in the voltammograms of the host compound. A rather striking example of the electrophoric group effect is contained in Fig. 5. This figure underlines the close similarity that exists in the voltammetric behavior of the -NMe, derivative, lm, and another entire@ djffe~~nt compound containing the -NMe, moiety, namely tetr~ethylphenylenedi~ne. Figures 6 and 7 contain the CV reduction responses of selected compounds in the series, 2 and 3. Comparison of the voltammograms of the derivatives with those of the parent compound in each case (cf. Fig. 6a and Fig. 7a) shows clearly how the presence of the electrophoric group introduces new features in them. For example, r (a) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG 0 I -1.o I -1.4 I I I -1 .a -2.2 -2.6 Potential/V vs AglAg+ 1 -3.0 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ Fig. 7. Comparison of the CV reduction response for selected compounds m the 3 series: (a) 3i; (b) 38; w 3c. 233 (4 I1 II -2.1 11 -2.2 11 -2.3 Potential/V 11 -2.4 I -2.5 vs Ag+‘O Fig. 8. CV data over a more limited potential span than in Figs. 6 and 7 illustrating the reversibility of the reduction response of p-NO, and p-CN electrophores in (a) 2a and (b) 3c respectively. The CV parameters from these and similar data are contained in Table 2. zyxwvutsrqponmlkjihgfedcbaZYXWVUT as with the previously reported -NO, derivatives of 1 [8] (la,b), a reversible CV response is observed for the electroreduction of the -NO, electrophore in 2a, 2b, and 3a as illustrated in Fig. 8. The corresponding CV parameters for the -NO, waves as well as for other electrophores such as -CN and -COOCH,, are collected in Table 2. In terms of chemical reversibility on the CV time scale, the compounds can be ordered as follows: 2% 2b, 3a, 3c > 2c > 2e. Thus, interestingly enough, the CV response due to the -CN electrophore does seem to reveal subtle differences depending on the rest of the compound matrix (i.e., compare column 5 for 2c with 3c in Table 2). In terms of electrochemical reversibility, however, a quick glance at the E, column in Table 2 reveals the following order: 3c >> 2e > 2a, 2b, 3a x=- 2c. The above sequence suggests that the differences between 2c and 3c may be due to differences in charge transfer kinetics rather than intrinsic chemical dissimilarities. 234 TABLE 2 CV parameters for reduction of 2 and 3 derivatives CV parameter a Compound -qX/V 2a -1.39 d -1.39 -1.39 - 1.39 -1.32 -1.32 -1.32 - 1.31 2b - 1.43 - 1.43 -1.44 -1.44 - 1.37 1.37 1.36 1.36 - - 2.25 2.25 2.23 2.23 0.08 0.11 0.12 0.83 0.91 0.93 0.94 2c 2.33 2.34 2.34 2.35 b Era/V b W/V i&r,, 0.07 0.07 0.07 0.08 0.92 1.00 1.01 1.00 zyxwvutsrqponmlkjihgfed 0.08 0.08 1.00 1.03 1.04 1.07 2e -2.32 -2.32 -2.32 -2.32 - 2.25 2.25 2.25 2.25 0.07 0.07 0.07 0.07 0.73 0.79 0.89 0.97 3a -1.44 -1.44 -1.45 -1.45 - 1.37 - 1.37 -1.37 - 1.36 0.07 0.07 0.08 0.09 0.99 0.98 1.02 1.06 3c -2.2-l - 2.27 - 2.27 - 2.28 - 0.06 0.07 1.05 1.05 1.09 1.08 2.21 2.21 2.21 2.21 ’ A The subscripts ‘p’. ‘c’, and ‘a’ denote peak, cathodic and anodic respectively. A E, = 1E, - E,, I. b Potentials vs. Ago’+ m acetonitrile (see Experimental). ’ Ratio computed according to ref. 31. d The four rows for each compound denote data at 0.02, 0.05, 0.10, and 0.20 V/s potential scan rate, respectively. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA As is well established in electrochemical methodology, the CV responses such as those depicted in Fig. 8, are to be attributed to one-electron electrochemistry due to the initially-formed radical anion. The voltammetric profiles for a given chromophore are quite similar in the two compound series. Figures 6 and 7 contain sample voltammograms illustrating this similarity for the p-NO, (Figs. 6b and 7b) and p-CN (Figs. 6c and 7c) compounds. Furthermore, the -NO, electrophore evinces a reversible CV response at potentials very close to those seen for the present compounds (Fig. 8 and Table 2) in an entirely different compound matrix, namely the C2-substituted thioxanthone system [32]. These similarities again reinforce the concept that electrophores manifest 235 /) 1 .2 s.Ag"" zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCB Fig. 9. Correlation of the CV oxidation potentials (cf. Table 1) with o for the 1,2 and 3 compound series. The substituent constants were obtained from ref. 37. responses in much the same manner as chromophores in optical spectroscopies. Other electrophoric groups such as -Cl and -Br also modify the reduction response of 2 and 3. In these cases, a shoulder superimposed on the main cathodic feature signals the occurrence of cathodic C-Cl or C-Br bond cleavage in much the same manner as were observed previously for halo-substituted thioxanthones [32]. The terminal reduction waves in Figs. 6a and 7a are attributable to the -SO, and -CO spacer groups in the 2 and 3 series respectively. These cathodic features, however, are altered by the presence of other electrophores such as -NO,, -CN, and -COOCH, (cf., Figs. 6b-d, Fig. 7b and c) as a consequence of the initial reduction of these latter moieties, and the possible interaction of the products of these electrolyses with the parent compound. From the findings of previous studies on sulphonamides [33], and carboxamides [34], we infer that the electroreduction of 2 and 3 proceeds with rupture of the S-N and C-N bonds respectively. The electronic influence of the N-aryl substituent, R, on the electrochemical behavior of 2 and 3 was examined in the form of Zuman plots [35] of the CV oxidation or reduction potential vs the Hammett substituent constant, (I. Figure 9 compares such plots for 2 and 3 with the previously reported data [8] for 1. Immediately obvious from this comparison is the reduced sensitivity of zyxwvutsrqponmlkjih Epa to u in 2 and 3 relative to 1. In other words, the SO, and CO spacer groups attenuate the electronic influence exerted by R at the N reaction site. The reduced sensitivity is 236 -0.8 +0.8 Fig. 10. Correlation of the CV reduction potentials (cf. Table 1) with 0 for the 2 and 3 compound senes. The substituent constants are from ref. 37. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJI quantifiable in the ordering of the reaction coefficients, p, in the Zuman plots: p = 0.40, 0.20, and 0.15 V for zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ 1, 2, and 3 respectively. The exclusion of the p-0CH3 derivative for the correlations in the 2 and 3 series, is suggested by the noticeably disparate on values that were measured in these cases. (Obviously the Era values measured for 2k and 3k (cf. Table 1) reflect those corresponding to the -OCH, electrophore.) For the same reason, the p-NMe, derivative was omitted from the correlation in the 1 compound series. In all three cases in Fig. 9, the correlation of zyxwvutsr EPa with u was either excellent or satisfactory *. It should be pointed out that the correlation coefficients using 6+ (0.980 and 0.984 for 2 and 3 respectively) are better than those using u (0.966 and 0.969). The significance of this, however, is not clear since we are unable to use the strongly electron-donating groups (OCH, and NH,) in the correlation. In contrast to the oxidation case, the reduction potentials should not correlate with the substituent constant because of the complications introduced by the electrophoric group effect. The analogs of the Zuman plot for the reduction of 2 and 3, which are shown in Fig. 10, bear out this expectation. Obviously, a corollary is that the validity of a Zuman plot rests with the #bse~ce of an electrophoric group effect. ACKNOWLEDGEMENTS One of the authors (M.P.) thanks the Robert A. Welch Foundation (Grant Y-684) for financial support of this research. We also thank the Organized Research * We adopt the convention here that for an excellent satisfactory correlation, 0.95 < r i 0.99 1361. correlation, r lies between 0.99 and 1.00 and for a 237 Fund of UTA and the National Science Foundation (Grant PMR 8108132) for partial funding support and Mr. Whe-Nam Chou, Ms. Christina G. Smith, Mr. 0. Robert Davis and Prof. Shmuel Bittner of Ben-Gurion University, Beer Sheva, Israel, for some compound syntheses and NMR analyses. REFERENCES 1 Bad&he Anilin und Soda Fabrik, A.G.. Fr. Demande 2,001,704, Chem. Abstr., 72 (1970) 66992; S.J. Kuhn. U.S. Patent 3557,208, Chem. Abstr. 75 (1971) 6096; N.A. Khodak, V.A. Gilyarov and ML Kabachnik, J. Gen. Chem. USSR (Engl. Trausl.). 44 (1974) 24; R.M. Washburn and R.A. Baldwin, U.S. Patent 3,189,564, Chem. Abstr. 63 (1965) 9991; L.E. Godfrey and J.W. Schappel, Ind. Eng. Chem. Product Res. Dev., 9 (1970) 426. 2 H.R. Allcock and F.W. Lampe, Contemporary Polymer Chemistry, Prentice-Hail, Englewood Cliffs, NJ, 1981. 3 H.R. Allcock, Phosphorus-Nitrogen Compounds, Academic Press. New York, 1972. 4 D.L. Herring, U.S. Patent 3,182,028, Chem. Abstr., 63 (1965) 1897. 5 H.R. Allcock, P.E. Austin and T.X. Neenan, Macromolecules 15 (1982) 689: H.R. Allcock, T.X. Neenan and W.C. Kossa, ibid., 15 (1982) 693. 6 P.M. Blonsky, D.F. &river, P. Austin and H.R. Allcock, J. Am. Chem. Sot., 106 (1984) 6854. 7 M.M. Baizer and H. Lund, Organic ~~tr~he~st~, Marcel Dekker, New York, 1983, Ch. 1. 8 M. Pomerantz, D.S. Marynick, K. Rajeshwar, W.-N. Chou, L. Throckmorton, E.W. Tsai, P.C.Y. Chen and T. Cain, J. Org. Chem., 51 (1986) 1223. 9 C.M. Pak and W.M. Gulick, Jr., J. Korean Chem. Sot., 18 (1974) 341. 10 V.V. Penkovskii, Yu.P. Egorov and G.S. Shapoval, J. Gen. Chem. USSR (Engl. Transl.), 41 (1971) 742. 11 H.R. Ailcock and W.J. Birdsah, J. Am. Chem. Sot.. 91 (1969) 7561; Inorg. Chem., 10 (1971) 2495. 12 G. Sciavon, S. Zecchin, G. Cogoni and G. Bontempelli, J. Electroanal. Chem., 48 (1973) 925. 13 K.S.V. Santhanam, L.O. Wheeler and A.J. Bard. J. Am. Chem. Sot., 89 (1967) 3386; K.S.V. Santhanam and A.J. Bard, ibid., 90 (1968) 1118. 14 R.N. Adams, Electrochemistry at Solid Electrodes, Dekker. New York, 1969, Ch. 7, p. 187. 15 M. Pomerantz, B.T. Ziemnicka. Z.M. Merchant, W.-N. Chou, W.B. Perkins and S. Bittner. J. Org. Chem., 50 (1985) 1757; D.G. Gorenstein m D.G. Gorenstein (Ed.), Phosphorus-31 NMR: Principles and Applications, Academic Press, Orlando, FL, 1984, Ch. 1. 16 S. Bittner, Y. Assaf. P. Krief, M. Pomerantz, B.T. Ziemmcka and C.G. Smith, J. Org. Chem., 50 (1985) 1712. 17 M. Pomerantz, W.-N. Chou, M.K. Witczak and C.G. Smith, J. Org. Chem., 52 (1987) 159. 18 Yu.G. Gololobov, I.N. Zhmurova and L.F. Kasukiun, Tetrahedron, 37 (1981) 437. 19 P.A.S. Smith. Org. Reactions, 3 (1946) 337. 20 M.E.C. Biffen, J. Miller and D.B. Paul, in S. Patai (Ed.), The Chemistry of the Avdo Group, Wrley. London, 1971, Ch. 2. p. 86ff. 21 G.I. Derkach and ES. Gubnitskaya. J. Gen. Chem. USSR (Engl. Transi.), 34 (1964) 605; Zh. Obshch. Khtm, 34 (1964) 604. 22 M.R. Ktrchedorf, Synthesis, (1972) 695. 23 B.F. Malichenko, O.N. Tsypnia and V.F. Kulik, J. Gen. Chem. USSR (Engl. Transl.), 37 (1967) 1711; Zh. Obshch. Khim., 37 (1967) 1796. 24 L. Homer and H. Oedtger, Liebigs Ann. Chem., 627 (1956) 142. 25 J.C. Suatoni, R.E. Snyder and R.O. Clark, Anal. Chem., 33 (1961) 1894. 26 (a) T. Mizogucht and R.N. Adams, J. Am. Chem. Sot., 84 (1962) 2058; (b) 2. Galus and R.N. Adams, ibid., 84 (1962) 2061; (c) Z. Galus, R.M. White, F.S. Rowland and R.N. Adams, ibid., 84 (1962) 2065: (d) E.T. Seo. R.F. Nelson. J.M. Fritsch, L.S. Marcoux. D.W. Leedy and R.N. Adams, ibtd.. 88 (1966) 3498: (d) J. Bacon and R.N. Adams, ibid., 90 (1968) 659. 238 27 28 29 30 31 32 33 34 35 36 37 A.J. Bard and L.R. Faulkner, Electrochemical Methods, Wiley, New York, 1980, Ch. 11, p. 429. J.C. Chiang and A.G. MacDiarmid, Synth. Met., 13 (1986) 193 and references therein. T.A. Albright, W.J. Freeman and E.E. Schweizer, J. Org. Chem., 41 (1976) 2716. L.L. Miller and E. Riekema, J. Org. Chem., 34 (1969) 3359. R.S. Nicholson, Anal. Chem., 38 (1966) 1406. E.W. Tsai, L. Throckmorton, R. Mckellar, M. Baar, M. Kluba, D.S. Marynick, K. Rajeshwar and A.L. Ternary, Jr., J. Electroanal. Chem., 210 (1986) 45. D. Benedetti, R. Andreoli, G.B. Gavioli and G. Grandi, J. Electroanal. Chem., 68 (1976) 243; A. Rastelli, R. Andreoli, G.B. Gavioli, G. Grandi and D. Benedetti, ibid., 89 (1978) 207. See also references therein. L.V. Vesheva, R.A. Ovshirmikova and L.S. Reishakhrit, J. Gen. Chem., USSR (Engl. Transl.), 41 (1971) 980; L.V. Vesheva, O.S. Zatulina and L.S. Reishakhrit, ibid., 42 (1972) 1457; L.V. Vesheva, L.N. Prosvimova and L.S. Reshakhrit, ibid., 42 (1972) 2388. P. Zuman, Substituent Effects in Organic Polarography, Plenum, New York, 1967. J. Shorter, Correlation Analysis of Organic Reactivity with Particular Reference to Multiple Regression, Research Studies Press, Chichester, 1982. 0. Exner, in N.B. Chapman and J. Shorter (Eds.), Correlation Analysrs in Chemistry: Recent Advances, Plenum, New York, 1978, Ch. 10.