J. Am. Chem. SOC.1980, 102, 4636-4643
Electron Transfer in the Quenching of Triplet Methylene Blue by Complexes of Iron(I1) Takeshi Ohno and Norman N. Lichtin* Contribution from the Department of Chemistry, Boston University, Boston, Massachusetts 02215. Received January 14, 1980
Abstract: A Q-switched pulsed ruby laser emitting 1.0-J flashes at 693.4 nm was used in an investigation by means of flash photolysis-kinetic spectrophotometry of the mechanism of quenching of 2-10 pM triplet methylene blue by substitution-inert complexes of Fe(I1). Complexes included Fe(H20)2+, H2Fe11(CN)2-,HFe11(CN)63-,Fe11(CN)6d,Fe11(CN)4bpy2-,Fe"(CN)2(bpy)?, ferrocene0,and Fe11(bpy)32+.System variables included solvent (H20,aqueous EtOH, aqueous CH3CN,and aqueous DMF), pH (2.1-8.2) and ionic strength (0.01-1.6 M). Energy transfer from triplet methylene blue to quencher is not possible in any of the cases investigated,and quenching proceeds via partial or complete electron transfer. Rate constants for quenching, k,, were diffusion-controlledwith all quenchers except Fe1I(H20)? and, possibly, Fe"(bpy)?+. Efficiency of net electron transfer in the quenching process, F1= k , / k , , was measured directly and found to vary with quencher, solvent, and state of protonation of triplet methylene blue and quencher. Values of FI from 0.00 to 0.86 were observed. F1does not vary systematically with k,. It was found that RT In [(l/Fl) - 11 for the diffusion-controlled quenching of 3MBH2+by Fe11(CN)4bpy2-varies linearly, with slope -+O. 16, with Kosower's solvent polarity parameter Z in water, aqueous EtOH, aqueous CH3CN,and aqueous DMF with Z values ranging from 81.5 to 94.6 kcal/mol. The kinetics of the reverse electron-transfer reaction between complexes of Fe(II1) and semimethylene blue, MBH+., was studied independently by generating MBH+. by quenching triplet methylene blue with diphenylamine in the presence of a complex of Fe(II1). Values of krcl,the second-order specific rate of reverse electron transfer, were measured directly. Values of F2,the efficiency of net electron transfer in an encounter between MBH'. and a complex of Fe(III), were taken as equal to k,/kD, where kDis the diffusion-controlled specific rate of encounter of the reactants. Values of kDwere obtained by assuming equality between kD for a given complex of Fe(II1) and k, for quenching of 3MBH2+by the correspondingcomplex of Fe(II), an assumption which is accurate only at high ionic strength. For the four quenchers and five solvents for which both F1and F2 were determined, FI+ F2 = 1. For the two quenchers and three solvents in which kD (and, therefore, F2) was determined accurately, F1+ F2 = 0.99 i 0.06. The data reported in this study are consistent with the following mechanistic features of diffusion-controlled quenching of 'MBH2+ by complexes of Fe(II), FeI1Lxm:(1) All quenching encounters produce a geminate pair, [(MBH+.)...Fel"L,"+'], made up of semimethylene blue and the ferric complex. This pair can dissociate before or after undergoing reverse electron transfer. Quenching to the ground state proceeds via dissociation of this geminate pair subsequent to reverse electron transfer. (2) The geminate pair, [(MBH+.)-.FelllL,m+l],is common to quenching of 3MBH2+by complexes of Fe(I1) and to reverse electron transfer between MBH'. and complexes of Fe(II1). (3) Intersystem crossing is rapid in both quenching of the triplet and in the reverse reaction of MBH'. with ferric complexes. (4) Solvent effects on Fl, the efficienc of net electron transfer during quenching, are determined primarily by solvation effects upon the return from [(MBH+.)-Fe' YILXm+l]to IMB+ and FeI1Lxm.The kinetic order, rate, and products of decay of MBH+. and ferric complex resulting from quenching of 3MBH2+by ferrous complexes depend on the identity of the quencher and on the medium. Generally, decay was first order, and only MB+ was formed when the ferric complex was negatively charged. With positively charged ferric complexes, decay was second order and involved both reverse electron transfer to give MB+ and disproportionation to MB+ and leucomethylene blue. Disproportionation predominated only in the case of Fe111(H20)2+.
Introduction Net electron transfer between an excited molecule and a quencher to produce charge carriers is a necessary step in photogalvanic conversion of light into electricity:*** Q
G-*E-*R *R Q' R- or Q-
(2) *E is the excited state produced directly by absorption of light while *R is an excited state which is sufficiently long-lived to engage in collisions with quencher. A more detailed picture would delineate production of a Franck-Condon state by the absorption event followed by one or more rapid vibronic relaxation steps and, possibly, a slower change in multiplicity. Such a more detailed picture would also explicitly recognize that, depending on the concentration of Q, more than one energy state might play the role of *R so that eq 2 would represent a family of reactions; e.g., it has recently been shown3that 0.05 M Fe1'(H20)2+ significantly reduces both the SI and TI states of methylene blue in 0.01 M solutions of HC1 in 50% v/v aqueous CH3CN. Alternatively, *R may be produced indirectly by sensitization. Regardless of whether (1) N. N. Lichtin in "Solar Power and Fuels", J. R. Bolton, Ed., Academic Press, New York, 1977, pp 119-142. (2) M. 2.Hoffman and N. N. Lichtin in "Solar Energy: Chemical Conversion and Storage", R. R. Hautala, R. B. King, and C. Kutal, Eds., The Humana Press, Clifton, N.J., 1979, pp 153-187. (3) T. L. Osif, N. N. Lichtin, M. Z. Hoffman, and S.Ray, J . Phys. Chem., 84, 410 (1980).
or not sensitization is involved in production of *R, or whether *R represents one or more than one excited states, quenching via pracesses represented by eq 2 must dominate over other quenching mechanisms if production of charge carriers is to be efficient. Quenching processes which are alternative to net electron transfer include energy transfer to quencher, collisionally induced intersystem crossing to the ground state if the multiplicities of *R and G differ, and collisionally induced internal conversion of the ground state if the multiplicities of *R and G are identical. Energy transfer to the quencher may provide a pathway to production of charge carriers through reaction of the excited quencher with G, eq 3 and 4. Conversely, collisionally induced intersystem Q * R + *Q + G (3)
+ R- or Q- + R+
(4) crossing or internal conversion may proceed via partial electron transfer (Le., formation of a charge-transfer complex) or by complete electron transfer followed by complete reversal of electron transfer, both within the lifetime of the encounter complex, as well as by other mechanisms. This paper describes a study by means of laser flash photolysis-kinetic spectrometry of the dependence of quenching rates and the extent of net electron transfer in the quenching of triplet methylene blue by substitution-inert coordination complexes of Fe(I1) upon the nature of ligands and medium. In all the cases reported here the quencher has no known excited state with an energy low enough to allow significant energy transfer from the
0002-7863/80/1502-4636901.OO/O 0 1980 American Chemical Society
Quenching of Methylene Blue by Complexes of Fe(II)
J . Am. Chem. Soc., Vol. 102, No. 14, 1980 4637
Table I. Relevant Properties of Methylene Blue and Quenchers
M B+ MBH" MBH," IMB+ IMBH2+ MBH+. MBH,".
11 640b 8 22OC
Fe11(H10)6'+ HFeII(CN), '-
1 9 800' 23 700"
MB. 1W67, s -0.23 (MB+/MB.)~ + o x (MBH~+/MBH+.)~
+ 1.21 ('MB+/MB.)g
+1.33 ('MBH,'+/MBH+*)g 0.5 1 (pH 2)'
-0.770"' -0.56 (/.I = 0.1M)O
PK, 0.oe -5.1e 7,2h -9 1.9 (5% aqueous EtOH)k 1.2 (50% v/v aqueous AN)k 4.2 (pK, 2.2 (pK, 1.7 (pK, -0.3 (pK,
of H,Fe(CN),)O of H,Fe(CN),)O of HIFe(CN),bpy)q of H,Fe(CN),bpy)q
Fell(CN),(bpy),o -12 500' -0.78' ferrocene( 0) -15 000' -0.54 (95% EtOH, pH 0.65)' -1.07' Fe"(bpy), 11 500" a VS.std hydrogen electrode. b Reference 38. c Calculated from h E 3 M B H z + - AE'MB+ = 2.3RT[pKa(ground) - pK,(excited)]. d Galculated from Eredo' at pH 2 = +0.19 V (ref 42) with Eredo =Eredo'(pH 2) t 0.06(2) :E M B H ~ + ' .EMB+' calculated from EMB+' + (2.3RTl of MBH+. = E ~ ~ + (~2 . 3~R T /+q K ao y H 1 + , e Reference 40. f Reference 41. Eredo of 'MB+ = Eredo of MB' + E('MB+); E,, = Eredo of MBH'* + E(IMBH +). Reference 4. Reference 42. J References 9 and 43. Reference 8. Reference 44. iReference 45. " Reference 46. O Reference 35. P Assumed to be between 23 700 and 1 2 500 cm-I. Q Reference 36. ' Estimated from the triplet energy of 'Fe(CN),(phen),O, ref 47. Reference 48. Reference 49. ' Reference 50. " Reference 51.
triplet dye. Energies of t h e TI states of methylene blue and quenchers and redox potentials of the So and TIstates of methylene b l u e and of t h e So s t a t e of quenchers are summarized in T a b l e I along with pK, values a n d intrinsic lifetimes of a number of species involved in this study. The kinetics of t h e reaction of semimethylene blue with complexes of Fe(II1) h a s been studied under t w o circumstances. In one, t h e back-reaction of t h e products of photoreduction of triplet methylene blue was followed directly. In t h e other, semimethylene blue was generated in situ by reduction of triplet d y e by diphenylamine and its reaction with excess complex of Fe(II1) was followed. Results appear to be fully accommodated by a mechanism involving complete electron transfer from quencher to dye in t h e quenching encounter complex and control of efficiency of n e t electron transfer by competition between dissociation of t h e products of this transfer t o semimethylene blue and t h e ferric complex and reversal of electron transfer t o give t h e ground-state d y e and the ferrous complex.
Experimental Section Apparatus. A Holobeam Series 630 Laser System with a Q-switched ruby laser capable of providing up to 3.6 J per flash at 694.3 nm with a nominal pulse width of 19 ns was used. Details of the laser and monitoring apparatus are reported elsewhere.' Flash intensity of the laser was 1.O J per pulse in the work reported here and a preamplifier was not employed in the monitoring circuit. Materials. Methylene blue chloride trihydrate (mol. wt. 374) was Fluka puriss grade. Aldrich diphenylamine was recrystallized twice from aqueous methanol and then once from methanol. K,[Fe"(CN),bpy]* 3H20, Fe11(CN)2(bpy)2.3H20and [Fe11(bpy)3](C104)zwere prepared by the method of Schilt,J and their purity was verified by UV-vis spectrometry. H[Fe11'(CN)4bpy].2H20 was prepared by oxidation of K2[ Fe11(CN)4bpy]in aqueous solution followed by addition of hydrochloric acid. The product was recrystallized from water and its purity verified by UV-vis spectrophotometry.s Alfa ferrocene was recrystallized from cyclohexane. Ferrocinium perchlorate, which is unstable6 both as pure solid and in solution, was prepared immediately before use by oxidation of ferrocene with Fe(C104)3 in 0.01 M aqueous HClO,. Purity was (4) T. Ohno, T. L. Osif, and N. N. Lichtin, Phorochem. Phorobiol., 30, 541-546 (1979). ( 5 ) A. A. Schilt, J . Am. Chem. Soc., 82, 30012-3005 (1960). (6) J. R. Pladziewicz and J. H. Espenson, J . Am. Chem. Soc., 95, 56-63 (1973).
verified by UV-vis spectrophotometry.6 K4[Fe11(CN),] and K3[Fe111(CN),] were Baker Analyzed Reagent grade. KHzP04,HCI, and H2SO4 were Baker Analyzed grade. HClO, and Na2B4O7.10HZ0were Fisher Certified ACS grade. Potassium hydrogen phthalate was Mallinckrodt Analytical grade. MgClz was Matheson Coleman and Bell ACS Reagent grade. Laboratory distilled water was purified by passage through a Millipore deionizer and filter. Aldrich dimethylformamide was distilled in vacuo. Acetonitrile was Burdick and Jackson UV grade. Ethanol was US1 Reagent grade. Test solutions were deaerated by purging for 15-20 min with deoxygenated nitrogen which had first been bubbled through the same solvent used in the test solution. The nitrogen was deoxygenated by bubbling through chromous solution which was stored over zinc amalgam. Measurements. Decay of protonated triplet methylene blue ("BHZt) was monitored at 370 and 710 nm.' Measured rates of recovery of ground-state methylene blue (MB'), monitored at 635-650 nm, and of decay of )MBHZt were in good agreement in the absence of added quencher. Decay of 3MBt was monitored at 415 and 825 Yields of semimethylene blue, MBH'., were measured by decrease of absorbance of MB+ at 635-650 nm immediately after decay of absorbance by 'MBHZt or 3MBt. Decay of MBHt- was monitored directly at 840 or 880 13131' and indirectly by increase in absorbance due to MBt. It was verified that MBH'. does not absorb significantly at 570-700 nm, where MBt absorbs strongly, by observing that the decrease in absorbance in this region (difference spectrum) caused by conversion of MBt to MBHt. agreed very well with the spectrum of MB+. In studies of the kinetics of oxidation by the ferric complex of MBHt. which had been generated by reduction of 'MBH2+ or 3MBt by 0.3-1 mM diphenylamine (DPA) both MBt and MBH'. were monitored. The rate constant for quenching of triplet methylene blue by DPA was 1 X lo9 M-I s-l in all the media investigated. Concentrations of ferric complexes were less than 0.08 mM in all cases.
Data Quenching Constants, k,. Specific rates of quenching of triplet methylene blue, 3MBH2t or 3MBt depending on pH (pK, = 7.2 in aqueous solution)! by ferrous complexes were determined from t h e slopes of plots of pseudo-first-order specific rates of its deca vs. concentration of quencher, eq 5 . Quenchers included Fete
kd = ko + k,[Fe"] (H20)6*+,Fe"(bpy);'
(where bpy = 2,2'-bipyridyl), ferrocene(O),
(7) J. Faure, R. Bonneau, and J. Joussot-Dubien, J . Chim. Phys., 65, 369-370 (1968).
Ohno and Lichtin
4638 J . Am. Chem. SOC.,Vol. 102, No. 14, 1980 Table 11. Values of k , and F,( 3cet/kq) in Quenching of Triplet Methylene Bluea by Complexes of Fe(I1) quencher FeII(H ,O)
H2 0 50% v/v AN H2O 5% EtOH 5% EtOH 5% EtOH 45% EtOH 55% EtOH 75% EtOH 60% DMF H,OC H,OC H20C 5% EtOH 5% EtOH 30% EtOH 30% EtOH 30% EtOH 60% EtOH 60% EtOH 78% EtOH 88% EtOH 30% DMF 70% DMF 30% v/v AN 50% v/v AN 75% v/v AN 20% EtOH 60% EtOH 78% EtOH 95% EtOH 5% EtOH H,O
FeII(CN), 4 H,Fe"(CN),'Fe11(CN)64H, Fe"(CN), 'I:e11(CN)64H,Fe"(CN), 'FelI(CN), bpy2-
Fe1I(CN),(bpy),'' ferrocene(0) FeII(bpy),'+
0.01 M H,SO,d 0.01 M H,SO,d
0.04 M KH,PO,e 0.01 N H,SO,f 0.04 M potassium hydrogen phthalateg 0.006 M Na,HPO,, 0.04 M KH,PO,h 0.01 N H2S0, 0.04 M KH,PO, 0.01 N H,SO, 0.01 N H 2 S 0 , 0.04 M KH,PO,e 0.04 M KH2P0,e 0.025 M Na,B O,i 0.01 N H,S0,4
0.008 M HC1 0.01 N H,SO, 0.04 M KH,PO, 0.025 M Na,B,O, 0.01 N H,SO, 0.04 M KH,PO, 0.01 N H,SO, 0.01 N H,SO, 0.01 N H,SO, 0.01 N H,SO, 0.01 N H,SO, 0.01 N H,SO, 0.01 N H,SO, 0.04 M KH,PO, 0.04 M KH,PO,m 0.01 M HC10, 0.01 M HClO, 0.008 M HCl 0.04 M KH,POde
0.2! 0.2J -0.08 -0.02 -0.08 1.6k -0.02 -0.08
-0.02 -0.02 -0.08 -0.14j -0.06 -0.02 1 9 -0.02 -0.14j -0.06 -0.02 -0.08 -0.02 -0.02 -0.02 -0.02 1.2k 0.5k 0. l k
109kq, M - I S-'
0.085 14.3 11.4 9.0 4.0 5.6 4.3 4.2 2.8 tl.0 7.7 4.0 8.5 3.1 3.8 4.1 2.8 2.8 4.2 2.7 2.2 4.3 2.5 3.6 4.4 7.1 3.3 1.7 3.4 4.9 1.12 0.42
0.84 0.16 0.23 0.36 0.27 0.5 2
0.58 0.53 0.70 0.088 0.095 0.00
0.10 0.095 0.24 0.22 0.07 0.32 0.31 0.58 0.69
0.28 0.86 0.30 0.47 0.66 0.078 0.076 0.68 0.68 0.00 0.00
All binary solvents have H,O as a ['MBH''] at pH < 5.5, [ ,MB'] at pH 8.8. [Triplet] = [MB+], = 10 pM unless otherwise indicated. second component. Compositions designated a s n % EtOH or DMF signify n volume of solvent made up to a 100 volume with water. Compositions designated as n% v/v AN signify n volume of AN mixed with 100 - n volume of water. Principal states of protonation of quenchers are shown as they would be ,in water at the indicated concentrations of buffers. [ MB'] = 2 pM. pH 2.0. e pH 4.4. pH 2.1. g pH 5.5. pH 6.0. pH 8.2. J Made up with Na'SO,. Made up with MgCl,. Taken from ref 10. Buffering of triplet dye incomplete. Quenching of 'MBH2* by [FdiCNi,bpy]*I
Figure 1. Dependence of pseudo-first-order specific rate of decay of 3MBH2' on [Fe*1(CN)4bpy]2-in water and aqueous ethanol: 10 FM MB'; 0.01N H 2 S 0 4in 60% and 78% aqueous EtOH; 0.04 M K H z P 0 4 and 0.02 M N a 2 S 0 4in 30% aqueous EtOH and water.
Fell( CN),( bpy) 20, Fell( CN) 4( bpy) ,-, a n d Fell( CN)64-. Typical d a t a a r e illustrated in Figure 1. T h e resulting values of k, a r e total specific rates of quenching summed over whatever quenching mechanisms a r e operative. Dependence of k, o n n a t u r e of quencher, solvent, a n d anions a s well a s on p H a n d ionic strength was explored. Values of k a r e presented in T a b l e I1 along with values of F1,t h e fraction of quenching events resulting in net reduction t o semimethylene blue, MBH?+,, MBH'., o r M B o depending on p H (pK,, = 1.9 and pKa2 = 9.0 in aqueous s 0 1 u t i o n ~ ~ ~ T) .h e determination of F', values (8) P. V. Kamat and N. N. Lichtin, Photochem. PhotobioL, 31, in press.
Figure 2. Determination of k,, in quenching of 'MBH2+ by Fell(CN)4bpy2-under the same conditions as data of Figure 1.
is described in t h e next section. Quenching by Fe11(Hzo)62+ is a t least two orders of magnitude slower than t h e diffusion-controlled rate. All t h e other complexes quench a t or n e a r t h e diffusion-controlled limit.
Rate Constants for Net Electron Transfer from Ferrous Complexes Acting as Quenchers to Triplet Methylene Blue, k,,, and Values of F , = k,,/kg. Values of k,, were determined with t h e aid of eq 6 where [3Dye] = [3MBH2+]or [3MB'], depending upon [MBH+.]/[3Dye] = k,,[Fe"] / k d
t h e p H , was shown t o be equal under t h e conditions employed (9) R. Bonneau, J. Faure, and J. Joussot-Dubien, Ber. Bunsenges. Phys. Chem., 12, 263-266 (1968).
Quenching of Methylene Blue by Complexes of Fe(II)
J. Am. Chem. Soc., Vol. 102, No. 14, 1980 4639
Table 111. Values of kret and F , (=kret/kD)' in Oxidation of Semimethylene Blue by Complexes of Fe(II1)
solventb 5% EtOH 5% EtOH
buffer P, M -0.02 0.01 N H,SO,C 0.04 M potassium hydrogen -0.08 phthalated 1.6f 5% EtOH 0.b4 M KH,PO,, 0.006 M Na,HPO,e -0.02 75% EtOH 0.01 N H,SO, FelI1(CN),bpy5% EtOH 0.01 N H,SO,C -0.02 5% EtOH 1.5f 0.008 M HCIC 1.2f 30% v/v AN 0.01 N H,SO, 0.5f 0.01 N H,SO, 50% v/v AN O.lf 75% v/v AN 0.01 N H,SO, 0.01 ferrocenium(l+) 75% EtOH 0.01 M HClO, -0.05g Fe"'(H,O). 3+ 50% v/v AN 0.01 M H,SO. kD taken as equal to k, for the corresponding ferrous complex under the indicated conditions. tion of compositions. pH 2.1 in aqueous solution. pH 5.5 in aqueous solution. e pH -6.0. Na,SO,. !I From ref 13. ' Estimated by assuming kD = 2.0 X 10'. to the total initial concentration of methylene blue and kd is defined in eq 5 above. Typical data for quenching by Fe11(CN)4bpy2-in 0.01 N aqueous and aqueous alcoholic solution are shown in Figure 2. Values of F , = k,,/k, are summarized in Table 11. Values of k,, can be obtained by multiplying F , values by corresponding values of k, given in Table 11. Casual perusal of Table I1 reveals that F I varies systematically with medium in a way which is different from the dependence of k, on medium. This systematic variation is part of the evidence for a quenching mechanism which is presented a t length in the Discussion below. Rate Constants for Electron Transfer to Ferric Complexes from Semimethylene Blue, k,,,, and Values of F2 = krqt/kencounter* An important aspect of the quenching mechanism which is presented below in the Discussion is the partitioning of a caged pair consisting of a molecule of semimethylene blue and a molecule of ferric complex between, on the one hand, free ground state methylene blue and ferrous complex and, on the other, free semimethylene blue and ferric complex. Equation 7 illustrates the process. In MBH'.
chis connecJion, there is a requirement for knowledge of the ratio k30[k34 k30]s F2. Second-order rate constants for oxidation of semimethylene blue by ferric complexes, kret,are equal to F, multiplied by k,, the diffusion-controlled specific rate of formation of the geminate encounter pair in the latter reaction. Values of k,,, have been measured directly. W e have chosen to estimate values of kD by equating them to values of k,, the specific rate of quenching of triplet methylene blue by ferrous complexes, under identical conditions. The validity of our conclusion that most values of k, are diffusion-controlled is taken up in the Discussion below. Quenching of triplet methylene blue by theferric complexes is slower than diffusion-controlled for all the cases we investigated.loaSb The assumption that k, = kD cannot be rigorously correct for all cases because of differences in ionic charge; e.g., at 1.9 < pH < 9 in aqueous s ~ l u t i o n , the * ~ ~principal species of semimethylene blue is MBH'. while, at p H < 7.2 in aqueous s ~ l u t i o nthe ,~ principal species of triplet dye is 3MBH2+. If the ferrous complex bore a 1+ charge and the ferric complex bore a 2+ charge the ideal situation shown in eq 8 could result. However, the quenchers
F, + F ,
0.4 1 0.43
2.8 0.73 4.4 2.4 2.9 2.3
0.70 0.17 0.5 2 0.79 0.80 0.52 0.21
0.97 0.70 0.62 0.89
1.10 0.99 0.87 0.80 0.84
See footnote b of Table I1 for explanaMade up with MgC1,. g Made up with
[Fe"l(CNl,(bpy)-], p M
Figure 3. Determination of k,,, for oxidation of semimethylene blue by Fe"'(CN),bpy- in 5% EtOH.
strength. Thus, Holzwarth and Jiirgensen have shown that the specific rates of a group of eight diffusion-controlled electrontransfer reactions of coordination complexes ranging in charge type from A3+ B4- through Ao + Bo to A3+ + B2+ approach each other with increasing ionic strength in aqueous solution and M-' S-I at 23 OC with have a common value of (3.2 f 0.1) X p I 1 M.ll With p = 0.1 M the rate constants vary from 2.0 X lo9 to 6.3 X lo9 over the same range of charge types while with p = 0.01 M the range of variation of rate constants is a factor of -16." Semimethylene blue was prepared in the presence of complexes of Fe(II1) through photoreduction of methylene blue by an excess of unprotonated diphenylamine.12 Values of k,,, were derived from the dependence of pseudo-first-order specific rates of conversion of semimethylene blue to methylene blue upon concentration of ferric complex as illustrated in Figure 3. Table 111 summarizes values of k,, and F2= k,,/k,, where k, is the specific rate of quenching of triplet methylene blue by the corresponding ferrous complex in the same medium, as given in Table 11. Also included in Table I11 are values of F, + F2 which are compared in the Discussion below with predictions of the mechanistic model. Kinetics of Decay of Semimethylene Blue after Its Formation by Quenching Triplet Methylene Blue with Complexes of Fe". As has been pointed out,2 the attractiveness of the Fe"(H,O)JMB+.-H+] and respectively Fe"'Lx'+.-MBH1'. Any difference in energy of solvation of these species must be very small. The small value of Fl, 0.09, reported by Kikuchi et al. for the quenching of triplet methylene blue by ferrocene in unbuffered EtOHI8 is presumably due to the preponderance of 3MB' over 3MBH2+in this medium. With hexacyanoferrous as quencher of 3MBH2+in 0.01 N H2SO4 solutions in aqueous ethanol, In [(l/Fl) - 11 does not appear to vary linearly with Z but does increase with increasing Z. Analysis of the data is complicated by the probability that the state of protonation of the quencher varies significantly with change of solvent composition, particularly since pK2 is 2.22 in water.35 However, explanation of the solvent dependence of F1 for this quencher in terms of differences in the solvation energies of X3 0 and X3 4 is still appropriate.
Acknowledgments. This work was sponsored by the US.Department of Energy under Contract EY-76-S-02-2889. We wish to express our appreciation to Professor M. Z. Hoffman for many helpful discussions.