The reactions of three bis(viologen) tetraquaternary salts and their

Keiichi Tsukahara, Naoko Sawai, Satoko Hamada, Kayoko Koji, Yoshiaki Onoue, Takashi Nakazawa, Ryoichi Nakagaki, Koichi Nozaki, and Takeshi Ohno...
0 downloads 0 Views 816KB Size
J . Am. Chem. SOC.1986, 108, 3380-3385

3380

Reactions of Three Bis(vio1ogen) Tetraquaternary Salts and Their Reduced Radicals Stephen J. Atherton,’. K. Tsukahara,lband R. G. Wilkins* Contribution from the Department of Chemistry, New Mexico State University, Las Cruces, New Mexico 88003. Received September 9, 1985

Abstract: Monoradical trications X3+-and diradical dications X2+-were produced by reduction of three bis(vio1ogen) X4+ (1,I”-alkanediylbis( l’-aIkyl-4,4’-bipyridinium))tetraquatemary salts. The spectral properties of X”+. and its disproportionation kinetics were examined by e,200 pM, violet-blue solid precipitates. eObeys Beer’s law at X = 470, 534, and 560 nm, 6-200 pM, and pH 7.2.

of 1 and of one- and two-electron reduced species 2 and 3 have been recently reported.9J3-17 In continuation of our examination of the kinetics of reaction of viologen specie^,^^^'^ we report on the electrochemical,spectral, and equilibria characteristics of the bis(vio1ogens) 1-3, generally designated X4+, X3+., and X2+- and on the kinetics of a variety of reactions of these species involving reduction, oxidation, and disproportionation.

Experimental Section

2

-3 (1) (a) Center for Fast Kinetics Research, The University of Texas, Austin, 78712. (b) Department of Chemistry, Shimane University, Matsue, Shimane 690, Japan. (2) (a) Bard, A. J.; Ledwith, A.; Shine, H. J. Adv. Phys. Urg. Chem. 1976, 13, 155. (b) Bird, C. L.; Kuhn, A. T. Chem. Soc. Rev.1981, IO, 49. (c) Summers, L. A. Adv. Heferocycl. Chem. 1984, 35, 281. (3) Laane, C.; Ford, W. E.; Otvos, J. W.; Calvin, M. Proc. Nafl. Acud. Sci. U.S.A. 1981, 78, 2017. (4) Graetzel, M. Arc. Chem. Res. 1981, 14, 376. (5) Darwent, J. R.;McCubbin, I.; Porter, G. J. Chem. SOC.,Furuduy Trans. 2 1982, 78, 903. (6) Kalyanasundaram, K. Coord. Chem. Rev. 1982, 46, 159. (7) Tazuke, S.;Kitamura, N. Pure Appl. Chem. 1984, 56, 1269. (8) Schoot, C. J.; Ponjd, J. J.; van Dam, H.T.; van Doom, R. A.; Bolwijn, P. T. Appl. Phys. Leff. 1973, 23, 64. (9) Bruinink, J.; Kregting, C. G. A,; Ponj&, J. J. J . Elecfrochem. SOC. 1977, 124, 1854. (10) Cieslinski, R. C.; Armstrong, N. R. J . Elecfrochem.Soc. 1980, 127, 2605. (1 1) Summers, L. A. The Bipyridinium Herbicides; Academic: New York, 1980. (12) Sassoon, R. E.; Gershuni, S.;Rabani, J. J . Phys. Chem. 1985, 89, 1937 and references therein.

0002-7863/86/ 1508-3380$01.50/0

Chemicals used were the purest commercial product. The three bis(viologens) were prepared by a slight modification of the method described by Furue and N0~akura.I~ The l ,l”-poly(methylene)bis(4,4’bipyridinium) dibromide compounds were converted to the corresponding 1,l”-poly(methylene)bis( l’-methyL4,4’-bipyridinium) salts by using a large excess of methyl iodide in DMF at 90 OC for 24 h. After recrystallizing from water, the iodide salts were converted to the chloride salts using a Dowex 1 x 4 (Cl- form) column, and then the perchlorate salt was precipitated by adding NaC104 to a concentrated solution of the chloride. The absorption spectral data for ETQ(C1O4),, PTQ(C104)4, and BTQ(C104)4 were in good agreement with those reported in the literature (Table I). Cyclic voltammetry was carried out in an argon atmosphere in an aqueous Tris/H2S04buffer (pH 7.2) with a BAS Model CV-1B instrument. A three-electrode system was used with a Pt auxiliary electrode and a Pt working electrode against a Ag/AgCl (3 M NaC1) reference electrode. Voltammograms were recorded on a Houston Instruments Omnigraphic 100 X-Y recorder at scan rates from 20 to 200 mV s-l. The instrument was checked using methylviologen which showed a reversible wave ( E o = -0.45 V and AEp = 60 mV).’* (13) Furue, M.; Nozakura, S.Chem. Leu. 1980, 821. (14) Furue, M.; Nozakura, S.Bull. Chem. Soc. Jpn. 1982, 55, 5 13. (15) Deronzier, A.; Galland, B.; Vieira, M., Nouu. J . Chim. 1982, 6, 97. (16) Deronzier, A,; Galland, B.; Vieira, M. Electrochim. Acta 1983, 28, 805. . ~~

(17) Attalla, M. I.; McAlpine, N. S.;Summers, L. A. 2.Nufurforsch., B 1984, 398, 74.

(18) Tsukahara, K.; Wilkins, R. G.J . Am. Chem. SOC.1985, 107, 2632. (19) Tsukahara, K.; Wilkins, R. G. Inorg. Chem. 1985, 24, 3399.

0 1986 American Chemical Society

J . Am. Chem. Soc., Vol. 108, No. 12, 1986 3381

Bis(uio1ogen) Tetraquaternary Salts Reactions The comproportionation constants were determined by treating various concentrations of FTQ4+(0.1-3.0 mM) or BTQ4+ (0.5-1.0 mM) with variable amounts of dithionite solution (final concentrations, 5-30 pM). Some of this dithionite was invariably consumed by traces of 02(-5-10 pM). The amounts of bis(vio1ogens) reduced could be estimated from the absorptions at the isosbestic points for disproportionation of FTQ3+to PTQ2+- (560 nm, PW+. = 6.8 X lo3 M-I cm-' ) or BTQ". to BT@+(564 nm, e B T p . = 7.5 X lo3 M-I cm-I ). The amounts of FTQ3+.and FTQ2+- or BTQ3+. and BTQ2+- could then be estimated from absorbances at other wavelengths, knowing e values for the mono- and diradicals obtained from the pulse radiolysis and dithionite reduction work. In a few experiments, reduced radicals of FTQ were produced by light irradiation of R Q 4 + (8 mM) for 5-15 s in a mixture of 3,lO-dimethyl-5-deazaisoalloxazine(-5 pM) and NazHledta (-0.5 mM).,O The equilibrium constant for reaction of BTQ4+ with DQ'. was deterand B T V ((a) mined by adding 26.9 pM dithionite to mixtures of 40 and 40 pM, (b) 50 and 30 pM, respectively). The concentration of BT@+- was determined at 536 nm (em@+-= 2.3 X lo', em+.= 300 M-I cm-l). The equilibrium constant for reaction of F T V with ET@+- was determined by adding 30 pM dithionite to mixtures of FTQ4+ and ETQ2+- (total concentration 50 pM). The total reduced species was determined at 550 nm (e ETQZ+- = e q + : = 1.9 X lo4 M-l cm-I). The concentrations of ETQZ+- and PTQ were determined at 500 nm ( P ~ T Q ' + - 6.7 X lo', em@+-= 1.9 X lo4 M-I cm-' ), 534 nm (eprQz+. 1.4 X lo', em@+-= 2.3 X lo4 M-' cm-I), and 600 nm (tETQZ+. = 2.2 X lo', em$+- = 2.7 X 103 M-' c m - I ) . The equilibrium constant for reaction of ETQ4+ with DQ'. was determined by electron-pulsing mixtures of ETQ" (10-50 pM),DQ2+(50-200 pM), NaHCO, (0.1 M), and N 2 0 (saturated). The reaction was studied at 600 nm where absorbance of Q'., and production of ETQ3+.can be ETQ3+-greatly exceeds that of D monitored as it is produced in the reduction of ETQ4+ by DQ'.. In the stopped-flow kinetic experiments, the following conditions were used: 10-20 pM X4+ mixed with 2-20 mM dithionite; 5-10 pM Xz+mixed with 5-15 mM Co(en),)+, 20-60 pM Co(edta)-, or Fe(CN)& 1-5 pM Xz+-mixed with -5 (degassed buffer)-130 pM 0,.The reaction between PTQz'- (13-25 pM) and FTQ" (-500 pM) was studied in the presence of -5-25 pM 0,.In the pulse-radiolysis experiments, -5 pM erq- or COz- was reacted with 50-100 pM X4+ to determine reduction rates. The concentration of e, was determined from the absorbance change associated with its loss at 650 nm (e = 1.6 X lo' M-I cm-I). In the disproportionation experiments, 10 pM X3+. generated by COT or e,,;/COT reduction of 50-250 pM X4+ was monitored spectrally for about 4 ms (light instability prevented longer time observations). In the study of other reactions, 10-15 pM X3+. in the presence of 0.25-1.5 mM X4+ reacted with 0.26-1.3 mM 0,or 1-2 mM Co(edta)- present in the pulsed solution. All reactions of X3+.with added oxidant were much faster than disproportionation of X3+. Most of the reactions were monitored at the absorption peaks of the radicals (Table I). Excellent first-order kinetic traces were obtained. A Gibson-Dionex stopped-flow apparatus interfaced with an OLIS data-collecting system was used. A CN Van de Graaf electron accelerator at the Center for Fast Kinetics Research (The University of Texas, Austin) was used as an electron source. Pulses of 100-ns duration are delivered to samples in a quartz cell with a 1-cm optical path length. When the only reducing radical was eW-,irradiation was carried out in 5% (CH,),COH which removes OH and H radicals, also generated in the pulse (OH(H) + RH R H20(H2)). When COz- alone was used, the irradiated solution contained N,O and 0.1 M HC02- (eaq-+ N 2 0 H 2 0 Nz + OH + OH-; OH(H) HCOz- CO2- + HzO(H,)). For a reducing mixture of eq- and CO,-, 0.1 M HC02- and an inert atmosphere of N2 were used.21 All manipulations involving radicals were carried out with scrupulous exclusion of 0,. A variety of conditions (ionic strength, buffer constituents) were used at 25 OC dictated by the type of experiment. Different conditions had little effect on the rates.

m+

-

-

+

+

-

+

-

Results It was found that 5-10 p M 0,was always present in freshly prepared buffer solutions. This did not interfere in the equilibrium measurements since the concentrations of radicals were determined spectrally. The traces of O2showed up as a rapid loss of X2+(with a small absorbance change) when it was mixed with the oxidant or the buffer system alone in the stopped-flow apparatus. It was usually separable from the reaction under investigation. In the pulse radiolysis experiments, traces of O2were sometimes removed by reduction to 0,in the first pulse. (20) Massey, V.; Hemmerich, P. Biochemisrry 1978, 17, 8. (21) (a) Wilkins, R. G. Adu. Inorg. Bioinorg. Mech. 1983, 2, 139. (b) Buxton, G. V. Ibid. 1984, 3, 131.

Table II. huilibria Data for Reactions Involvina Bis(vio1oaenst Elo,

-

reaction ETQ4++ 2e- ETQ2+-

+ +

PTQ4+ 2eBTQ4+ 2e-

-. -

PTQ2+.. BTQ2+-

2ETQ3+.ii ETQ4++ ETQ2+2PTQ3++. + PTQ4++ pTQ2+..

v

EzO,

v EIZ', V -0.28" -0.28b -0.30,' -0.28,"

K

-0.31'

-0.33' -0.19' -0.26,' -0.2gC -0.39' -0.29' -0.34,' -0.35.8 -0.37e

-

lb

260 h 60h

-

220 f 3oi 49 f 18h

2BTQ". BTQ4++ BTQ2+PTQ"' ETQ2+- F! PTQ2+- + ETQ"' 2DQ+*ii BTQ" BTQ2+-+ 2DQ2+ ETQ4++ DQ+*+ ETQ3+*+ DQ2+

+ +

11fY 0.8 f 0.2) 13 f 6k 11 f 3'

"Pulse radiolysis, from ETQ'+, D Q'. cross reaction using Eo(DQZ+/+)= -0.35 V at I = 0.1 M (NaHCO,), pH 7.3. bEstimated from E l o and EIz0 values. cThis work; cyclic voltammetry at I = 0.01 M, 10 mM Tris, pH 7.2. dFrom PTQ4+, ETQ2+- cross reaction and E I z 0 for PTQ4+/2+system. 'Reference 16, cyclic voltammetry at I = 0.1 M (KCI). 'Estimated from EIz0 and Kbp values. %FromBTQ", DQ'. cross reaction using EO(DQ2+/+)= -0.35 V, ref 18. * I = 0.04 M, pH 8.2. 'Reduced species generated by 5-1 5-s irradiation of 5 aM 3,10-dimethyl-5-deazaisoalloxazine and 0.1 mM EDTA solutions containing PTQ4+at I = 0.09 M and pH 8.2. ' I = 0.1 M,pH 7.2. ' K 5 , from spectral analysis of equilibrated solution, I = 0.1 M (NaHCO,), pH 7.3. ' K s , from kinetics ( k 5 / k 5 )by pulse radiolysis. I = 0.1 M (NaHCO,).

-

Spectra. The three bis(vio1ogen) species X4+are reduced by the pulse radiolytically generated radicals eaq- and C 0 2 - to the radical trications X3+.. The same spectra are obtained with both reductants. The concentration of the X3+. radical was equated to that of the e,; used in generating the radical and knowing this value, the molar absorptivity of X3++could be determined. These are recorded in Table I. PTQ3+- is produced in methanol by short-time white-light irradiation (500-W lamp) of a mixture of PTQ4+C14(8 mM), 3,10-dimethyl-5-deazaisoalloxazine(5 pM), and NazH2edta (0.1 mM).,O The absorbance ratio Aso5/ASs7 decreases with increasing irradiation time (increasing radical production). With 154 irradiation, -90% PTQ3+.is produced. When 50 p M PTQ4+is irradiated for even short times (10 s), substantial amounts of PTQ2+- are formed. Dithionite reduction produced the diradical dication Xz+-, shown by spectral titration a t 534 and 560 nm (2.2 f 0.2 electrons per mole of X4+). The fully reduced X2+- could also be produced by 1 5 s white-light irradiation (500-W lamp) of a mixture of X4+ (20 pM), 3,10-dimethyl-5-deazaisoalloxazine( 5 pM), and Na2H2edta(0.1 mM).ZO The spectral characteristics of X4+, X3+-, and Xz+-are shown in Table I. Those of ETQ2+- and PTQ2+were concentration-independent over 10-50 and 6-200 p M , respectively. Reduction Potentials. Overall potentials E I z 0 for the X4+ 2e- 2 X2+- couple were determined by cyclic voltammetry. The results are shown in Table 11. With X = PTQ, a reversible two-electron reduction wave was obtained (AE, = 37 mV a t a scan rate of 25 mV s-'). With X = ETQ and BTQ, adsorption of the reduced radicals on the electrodes occurred and the overall reduction potential was estimated from the cathodic wave a t a slow scan rate. The constants for the equilibria (1) and (2) were determined spectrally (DQ2+ = 1,l'-ethylene-2,2'-bipyridinium ion). Isosbestic points were observed at 404 and 452 nm for (1)

+

+ 2DQ+* s BTQ2+- + 2DQZ+ PTQ4++ ETQZ+- s PTQ2+- + ETQ4+ BTQ4+

Kl

(1)

Kz

(2)

and a t 440 and 550 nm for (2). The reduction potential for DQ2+l+is well characterized with our conditions (Eo = -0.35 V)'*

3382 J . Am. Chem. SOC.,Vol. 108, No. 12, 1986

Atherton et al.

and combined with K , leads to a value for for PTQ4+similar to that estimated by cyclic voltammetry (Table 11). The value of K2 was likewise consistent with the E12' values for the FTQ4+/2+ and ETQ4+l2+couples (Table 11). Production of small concentrations of X*+- (by dithionite or photochemically) in the presence of large concentrations of X4+ leads to a series of spectra, with an isosbestic point near 560 nm, which can be quantitatively related to the establishment of equilibria (3) and (4). Although

+ FTQ2+.. z 2PTQ3+- k3, k-3, K3 BTQ4+ + BTQ2+- z 2BTQ3+* k4, k-4, PTQ4+

K4

ETQZ+-

(4)

PTQ4+

+

z ETQ3+*+ DQ2+ kk, k-5, KS

(5)

mixture of ETQ4+and DQ2+ was reacted with a small concentration of C02-radicals. The very rapid absorbance increase at 600 nm due to generation of ETQ3+. and DQ+. was followed by a further absorbance increase a t 600 nm as DQ'. (em = 1 X lo3 M-' cm-I) is replaced by ETQ3+. ( E = 1.1 X lo4 M-' cm-' ) as reaction 5 ensues. Spectral analysis of the equilibrated solution allows determination of K5 and combined with Eo for DP+/DQ+. leads to an E,' value for ETQ4+/ETQ3+.. Values of E,' and E,' (2Elz0 - E,') and Kdispare shown in Table 11. Kinetics. One-Electron Reduction of X4+. The strongly reducing radicals eq- and C 0 2 -produced by pulse radiolysis in micromolar concentrations effect only one-electron reduction of X4+, when the latter is used in excess. The observed rate constant koM is directly dependent on [X4'], and the second-order rate constants koM/[X4+] are given in Table 111. The kinetics of (5) reduce to those for a first-order reversible reaction (rate constant = kM) since both ETQ4+ and DQ2+ are both present in excess over ETQ3+. and DQ+.. The plot of kobsd/[DQ2+I0vs. [ETQ4+Io/ [DQ2+lo*was linear with a slope value k5 and an intercept k+ The values are shown in Table 111. The singly reduced species react with .CH2C(CH3)20Hradicals also produced when (CH3)3COHwas included in the solution (to scavenge H and OH). This former reaction occupied several hundred microseconds and interfered with the observation of disproportionation. The problem was avoided when (CH3),COH was excluded and the trication radicals generated by COT or by a mixture of COT and eq-. With these conditions, PTQ3+. and BTQ3+-underwent spectral changes with isosbestic points at 460 and 560 nm (PTQ3+.) and 370, 440, and 560 nm (BTQ3+-). ETQ3+. did not show any spectral changes. Those with FTQ3+.and BTQ3+.corresponded to 100% and -90% disproportionation, and data on the time-dependent absorbance changes (Figure 1) allow calculation of k-3 and k-4. In terms of the general equilibrium (6),eq 7 and 8 come about. k

2x3+.

p+.. + x4+

1 d[X3+*] d[X2+-] ------ dt - k [X3+*] 2 dt

(5.8 f 0.5) X lolob (1.4 f 0.1) X lo8' (3.4 f 0.3) X loBc (2.5 f 0.2) X lo8' (3.2 0.3) x 107' 28 X lo7' 61 f4' (5.9 EL: 0.5) X lolob (6.3 i 0.4) x 107* (7.0 f 0.6) X lo7? (1.2 f 0.3) X IO8" (1.4 f 0.2) X lo8" (1.6 f 0.1) X loSb (3.2 f 0.2) x 107'

ETQ'+*

(3)

the equilibria lie to the left, 3 mM BTQ4+ reduced with 10 p M S2O4'- produces 190% BTQ3+.. Values of disproportionation constants Kdisp(=K3-' and K4-') are shown in Table 11. These are, within a fairly large experimental error, independent of wavelength of observation and concentration of radical. Combination of E120with K3 (K4) leads to values for E,' and Ez0 (for X4+ + e- + X3+5X3+- e- z X2+-, respectively, X = FTQ and BTQ) shown in Table 11. The same procedure could not be used with ET@+ since the absorbance coefficients of ETQ3+. were half those of ET@'.. at all wavelengths, and comproportionation would be unaccompanied by spectral change. Instead the equilibrium ( 5 ) was examined spectrally by a pulse radiolysis method. A ETQ4+ + DQ'.

Table 111. Rate Constants for Reactions of Bis(vio1ogen) Species at 25 OC

PTQ'+*

(7 f 3) x 105' (1.9 f 0.2) X lo6' (1.8 f 0.2) X lo6' (7.8 f 0.8) X

pTQ2+..

2.1 x 108d 3.4 x 108" 3.1 x 107f 3 x 1070 3.1 x 107f 4.1 X 102g,k

7.5 x 6.6

X

7.7 x 4.9 4.6 x 5.6 X

lo7/ 10,5j

x 106k 105' IO6/

1.0 x 10,5j 1.0 x 106k 15.4 f 0.8' 63g*k 4.8 x 109" >5 x 107c (1.5 f 0.2) X 10Ioe 1.7 X 10'O" BTQ4+ (5.1 f 0.5) x 107' 1.9 x 107d BTQ". (2.5 f 0.6) X 108',i (3.7 f 0.3) X loae 1.7 f IO8! -4 x 107c BTQ * 3.4 x 107f (1.3 f 0.1) X lo2' 4.2 X 1O2g