Micellar Effects on Photoinduced Electron Transfer in Aqueous

Aug 3, 2016 - Department of Chemistry Tulane University New Orleans, Louisiana ... Of the five chromophores, the MC13, MC17, and DC17 complexes ...
16 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/Langmuir

Micellar Effects on Photoinduced Electron Transfer in Aqueous Solutions Revisited: Dramatic Enhancement of Cage Escape Yields in Surfactant Ru(II) Diimine Complex/[Ru(NH3)6]2+ Systems Rebecca E. Adams and Russell H. Schmehl* Department of Chemistry Tulane University New Orleans, Louisiana 70118, United States S Supporting Information *

ABSTRACT: The effect of cationic micelle incorporation on light induced electron transfer, charge separation and back electron transfer between an aqueous electron donor, [Ru(NH3)6]2+, and a series of Ru(II) diimine complex chromophores/acceptors, is presented. The chromophores have the general formula [(bpy)2Ru(LL)]2+ (LL = bpy; 4-R-4′-methyl-2,2′-bpy, R = pentyl (MC5), terdecyl (MC13), heptadecyl (MC17); 4,4′-di(heptadecyl)-2,2′-bpy (DC17)). Of the five chromophores, the MC13, MC17, and DC17 complexes associate with the added micelle forming surfactant, cetyltrimethylammonium bromide (CTAB). Quenching of the luminescence of the bpy and MC5 complexes by [Ru(NH3)6]2+ is unaffected by addition of surfactant, while rate constants for quenching of the MC13 and MC17 complexes are decreased. Cage escape yields following photoinduced electron transfer to generate [(bpy)2Ru(LL)]+ and [Ru(NH3)6]3+ are approximately 0.1 for all the water-soluble chromophores (excluding DC17) in the absence of added CTAB. In the presence of surfactant, the cage escape yields dramatically increase for the MC13 (0.4) and MC17 (0.6) complexes, while remaining unchanged for [Ru(bpy)3]2+ and the MC5 complex. Back electron transfer of the solvent separated ions is also strongly influenced by the presence of surfactant. For the MC13 and MC17 complexes, back electron transfer rate constants decrease by factors of 270 and 190, respectively. The MC5 complex exhibits two component back electron transfer, with the fast component having a rate constant close to that in the absence of surfactant and a slow component nearly 200 times smaller. Results are interpreted in terms of the partitioning of the 2+ and 1+ forms of the chromophores between aqueous and micellar phases. The extended lifetimes of the radical ions may prove useful in coupling the strong reductants formed to kinetically facile catalysts for reduction of water to hydrogen.

■

INTRODUCTION Photochemically driven catalytic systems aimed at efficiently splitting water have received much attention.1−4 In addition to electrochemical methods,5,6 light-induced electron transfer systems have been used in evaluation of both water oxidation and reduction catalysts.7−9 A typical scheme for generation of redox equivalents by visible light involves electron transfer between an excited state and a reductive or oxidative quencher10 and one well-studied example of this type utilizes the chromophore [Ru(bpy)3]2+ and the oxidative quencher methyl viologen (MV2+). The excited state of [Ru(bpy)3]2+ efficiently reduces methyl viologen, producing viologen cation radical and [Ru(bpy)3]3+.11−13 In principle, both the reactive radical ions could be used to contribute to multielectron catalytic processes that result in products that store a portion of the energy of the ions. In homogeneous photochemical systems, overall net redox chemistry typically involves use of a sacrificial reagent, yielding one or more products that are of little use along with a high value product. For instance, the recent literature is filled with articles describing systems for producing hydrogen that involve a chromophore, sacrificial electron donor and catalyst for water reduction.3 The net redox © 2016 American Chemical Society

reaction is as expressed in eq 1 and involves oxidation of triethylamine

along with reduction of water. This net reaction occurs because the energy wasting back electron transfer reaction is slow relative to reaction of the triethylamine cation radical to form products that react slowly with reduced forms of the catalyst used for water reduction to hydrogen.14,15 Hypothetically, the net reaction could be water cleavage to hydrogen and oxygen if two separate catalysts were employed, as long as the respective intermediate species were isolated from one another on a time scale longer than that required for the net redox reactions to take place. However, back electron transfer following charge separation is generally diffusion limited and severely limits the efficiency of photosystems used to produce redox equivalents; Received: June 12, 2016 Revised: August 2, 2016 Published: August 3, 2016 8598

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir

chromophore with the micelle. For the case shown in Scheme 1, the decrease in positive charge on the chromophore should result in KA(I) > KA(II). The larger the difference in the two equilibrium constants, the larger the advantage should be in using this approach for sequestration of photogenerated oxidizing and reducing equivalents. The use of micelles allows for separation of reactants by both electrostatic and hydrophobic effects.21−24 Manipulation of micellar systems to achieve high yields of photoredox products requires the use of redox reactants that are either repelled by or attracted to the micellar surface or interior. For instance, the use of ion pairing to a micelle surface to increase reaction efficiencies has received attention.25−33 Systems which make use of intramicellar confinement of reactants have also been well-studied.34−42 The close proximity of reactants facilitated by micelles results in enhanced forward reaction rates. For extension of the back reaction half-life; however, separation of the photogenerated radical ion reactants is key. In a bimolecular electron transfer, one reactant is oxidized while the other is reduced; the changes in redox states will affect hydrophobicity and ionic interactions. The above approach was implemented years ago by Gratzel and co-workers. Utilizing variants of the well-studied [Ru(bpy)3]2+/MV2+ system that employed surfactant viologen derivatives, extended lifetimes were obtained for the charge separated products in solutions containing positively charged micelles.43 In order to make the quencher amphiphilic, the viologen was modified with alkyl chains of varying lengths at one N-alkyl position. The viologen was tuned such that the forward electron transfer occurs only in the aqueous bulk phase, producing very hydrophobic one electron reduced viologens. The reduced, functionalized viologen partitions to the hydrophobic regions of the micelles on a time scale fast relative to back electron transfer in the aqueous phase.44 The oxidized chromophore, [Ru(bpy)3]3+ has increased positive charge and is effectively repelled from the cationic micelles and, for viologens with 10−14 chain hydrocarbon substituents, the halflife for recombination of the micelle encapsulated reduced viologen and [Ru(bpy)3]3+ is greater than 10 ms. This approach has received some attention over the years, but has not been directly exploited in development of catalytic systems for water cleavage, in part because available catalysts for water reduction/oxidation were kinetically slow or not readily encapsulated in the micellar phase. A further problem in using one electron reduced viologens as an electron relay with homogeneous water reduction catalysts is the fact that, for most highly reactive catalysts, the second catalyst reduction by the viologen cation radical is thermodynamically unfavorable.3 Related work on surfactant Ru(II) diimine complexes is limited and has focused on aggregation properties of the complexes themselves.45−47 The one electron reduced forms of the complexes are very strong reducing agents and would serve as excellent reductants for homogeneous water reduction catalysts. It has been demonstrated that derivatives of [Ru(bpy)3]2+, having a single bipyridine modified with a single long chain hydrocarbon of 13 or more methylene units, form micelles in aqueous solutions with very low cmc values.38 In this paper we present a time-resolved laser flash photolysis investigation of surfactant Ru(II) bipyridyl complexes of the type [(bpy)2Ru(MCx)]2+ (MCx = 4-methyl, 4′-x-2,2′-bipyridine where x = pentyl (MC5), terdecyl (MC13), or heptadecyl (MC17) or DC17 = 4,4′-diheptadecyl-2,2′-bipyridine) in the presence of the reversible electron donor [Ru(NH3)6]2+. Both

even under low excitation light intensities the half-life of the radical ions produced will only be a few milliseconds.16−20 For photoinduced redox reactions in aqueous solutions, one approach to diminishing interactions between reactive radical ions is to sequester one or both of the reactive species in a microenvironment. If the catalysts for the subsequent oxidation and reduction reactions are also separated, multielectron redox chemistry can occur in each microdomain to yield permanent products. In order for the sequestration to be successful, however, the light induced electron transfer must efficiently occur in one or both of the microdomains followed by efficient charge separation and charge sequestration. An example of such a process is shown in Scheme 1 and involves a cationic Scheme 1

surfactant chromophore, a cationic electron donor, and a cationic micelle. While the oxidized form of the chromophore (Ru(II) complex) can associate with the micelle, the equilibrium constant, KA(II), will depend on the hydrophobicity of the complex. The photoinduced electron transfer reaction can occur in the aqueous solution and also at the micelle/ solution interface; the degree to which this occurs in each of the phases is dependent on the partitioning of the chromophore into the micelle and the relative electrostatic/hydrophobic interactions of the electron donor with the micelle. The yield of radical ions formed will depend upon the fraction of excited chromophores that react with the donor and the subsequent overall efficiency of charge separation from the encounter complexes (ηcss and ηcsm). Finally, a critical factor in developing systems for which recombination of the reduced chromophore (Ru(I)) and oxidized donor (D2+) has a very long half-life, is the equilibrium constant for association of the reduced 8599

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir

followed by further degassing to minimize the introduction of oxygen into the solution.

charge separation yields and back electron transfer rate constants are influenced markedly by the presence of CTAB micelles.

■

■

RESULTS Five ruthenium(II) chromophores: [Ru(bpy)3]Cl2, [Ru(bpy)2MC5]Cl2, [Ru(bpy)2MC13]Cl2, [Ru(bpy)2MC17]Cl2, and [Ru(bpy)2DC17]Cl2, were prepared by established literature methods from cis-[Ru(bpy)2Cl2] and the alkyl modified bipyridyl ligand. The complexes were characterized by a variety of methods including combustion analysis, 1H NMR, MALDI mass spectroscopy, UV−vis absorption, cyclic voltammetry, and differential pulse voltammetry.45,48 Absorption and Luminescence Spectra of Complexes in Water and Aqueous Micellar Solutions. The ground state absorption spectra of the ruthenium(II) chromophores in aqueous solutions are quite analogous to that of [Ru(bpy)3]2+ with principal components being a ligand localized π → π* transition between 280 and 300 nm and longer wavelength transitions assigned as metal-to-ligand charge transfer (MLCT), the lowest energy having a maximum at approximately 450 nm. A modest red-shifting of the MLCT absorption maximum is seen when comparing the [Ru(bpy)2MCx]2+ complexes relative to [Ru(bpy)3]2+; the maximum in water shifts from 452 to 456 nm (Figure S1). The presence of CTAB micelles has no effect on ground state absorption spectra of any of the complexes (Figure S2). Red luminescence is observed from all of the complexes. Unlike absorption, the emission maxima of the chromophores containing MC13, MC17 or DC17 red shift in solutions containing micelles. For both [Ru(bpy) 3 ] 2+ and [Ru(bpy)2MC5]2+, the emission maximum is 616 nm in aqueous solutions containing 0 and 0.01 M CTAB. The chromophores bearing one MC13 or MC17 ligand have maximum emission intensity at 616 nm in pure water; in a 0.01 M CTAB solution, these chromophores exhibit maximum emission intensity at 628 nm. The complex [Ru(bpy)2DC17]2+ is insoluble in water in the absence of added surfactant. When solubilized in aqueous solutions containing 0.01 M CTAB, the excited state of [Ru(bpy)2DC17]2+ has an emission maximum at 628 nm. Emission spectra of the alkyl modified complexes are shown in Figure S3. Luminescence lifetimes for the complexes were measured in N2 degassed solutions and are reported in Table 1 for all the complexes. Lifetimes ranged from 475 to 600 ns for the unquenched chromophores in either pure water or CTAB containing solutions. Interestingly, the chromophores that exhibited a change in luminescence maximum upon addition of CTAB had excited state lifetimes that were invariant with the

EXPERIMENTAL DETAILS

Materials. [Ru(bpy)3]Cl2, [Ru(bpy)2MC5]Cl2, [Ru(bpy)2MC13]Cl2, [Ru(bpy)2MC17]Cl2, and [Ru(bpy)2DC17]Cl2 were prepared according to previously published methods.45,48 Cetyltrimethylammonium bromide and [Ru(NH3)6]Cl3 were obtained from Sigma-Aldrich and were used without further purification. [Ru(NH3)]2+ was prepared by Zn(Hg) reduction in aqueous solutions for 30 min under an inert atmosphere. Luminescence Quenching. The quenching of Ru complex luminescence by 9-anthracene carboxylate was examined in buffered aqueous solutions. The chromophore and quencher solutions were prepared in a 0.02 M stock phosphate buffer solution at pH 7. Stern− Volmer luminescence quenching measurements were determined in aerated solutions when the quencher was 9-anthracene carboxylate and thoroughly N2 degassed solutions when [Ru(NH3)6]2+ was quencher. Luminescence spectra and intensity quenching measurements were made using a PTI Felix 32 MD-5020 spectrofluorimeter. Quenching of solutions containing CTAB required degassing by passing a rapid stream of N2 over a solution being sonicated (Branson 1800 Ultrasonic Cleaner) for 15 min. This procedure avoided excessive bubbling of the surfactant and provided reproducible luminescence lifetimes. Spectroelectrochemistry. A UV−vis spectrum of [Ru(bpy)3]1+ was obtained on an Ocean Optics USB 2000 spectrophotometer, where a CH Instruments model 630 electrochemical workstation was used for potential control. A Pine Instruments honeycomb spectroelectrochemical electrode card provided the working and counter electrodes in a 0.2 cm path cell. Bulk electrolysis of the chromophore was done in an argon-flushed acetonitrile solution containing 0.1 M tetrabutylammonium hexafluorophosphate as the supporting electrolyte. The applied potential was referenced to a silver/silver chloride electrode. Time-Resolved Transient Absorption and Luminescence. Transient spectra and kinetic decays for electron transfer reactions were obtained by laser flash photolysis using an Applied Photophysics LKS 60 optical system and excitation with the output of an OPOTek optical parametric oscillator pumped by a Quantel Brilliant Laser equipped with doubling and tripling crystals. Excitation of the chromophores was typically at 450 nm using samples having an optical density of about 0.6. Luminescence lifetime data were acquired by averaging emission decays at 610 nm. Kinetic data for back electron transfer were collected at 510 nm over various time scales (microseconds to milliseconds). All data were collected with linear oversampling on a 600 MHz Agilent Infiniium oscilloscope. Preparation of Solutions for Transient Absorption and Luminescence Involving [Ru(NH3)6]2+. A 3 mL aqueous solution containing 40 μM chromophore and 0.01 M [Ru(NH3)6]Cl3 was deaerated with inert gas (nitrogen or argon) for 20 min. The Zn(Hg) amalgam was prepared using ∼1 g of mossy zinc. The zinc was immersed in an aqueous 1 M HCl solution for 10 min; immediately after removal from the acidic solution, the zinc was placed in an aqueous 0.05 M HgCl2 solution. The zinc was left in the HgCl2 solution for 10 min, then quickly rinsed with at least 50 mL of deionized water. The amalgamated zinc was dried with a stream of inert gas, then immediately placed into solutions containing the chromophore and [Ru(NH3)6]Cl3. The mixture was allowed to react for 30 min under a nitrogen or argon atmosphere. A 2 mL aliquot of the solution was then syringe transferred into a nitrogen purged spectrophotometric cell. Kinetic measurements were made immediately after reduction of the quencher. Solid CTAB and NaCl were added to the above solutions by quickly introducing the solid into the solution by funnel transfer, immediately followed by sonication under a rapid stream of nitrogen for 3 min. The rate of oxidation of [Ru(NH3)6]2+ by O2 is reported to be 126 M−1 s−1 at room temperature in aqueous solutions,49 so additions of solids were

Table 1. Luminescence Lifetimes of N2 Deaerated Solutions of [Ru(bpy)2L]2+ (L = bpy, MC5, MC13, MC17 and DC17) in the Presence and Absence of CTAB Micelles

8600

[Ru(bpy)2L]2+ complex

[CTAB], M

λmax, nm

τ0, ns

bpy bpy MC5 MC5 MC13 MC13 MC17 MC17 DC17

0 0.01 0 0.01 0 0.01 0 0.01 0.01

615 615 616 616 616 628 616 628 628

575 575 480 480 535 530 540 540 607

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir change of medium (Figure S4). In all cases, excited state lifetime decays were monoexponential, suggesting the chromophore exchange between micelles is fast relative to excited state decay. An alternative explanation is that the lifetime change between the two environments may be small enough that the two decays were still able to be fit as a single exponential (simulations clearly show that a 30% difference in lifetimes can be fit well with a single exponential decay with a modest level of noise masking the irregularities of the residuals). Association of Cationic Complexes with Cationic Micelles: 9-Anthracene Carboxylate (9AC) Quenching of Chromophores in the Presence of Micelles. In an attempt to gain insight into the location of the dicationic chromophores in cationic micellar solutions, luminescence quenching of the chromophores by energy transfer to anthracene-9 carboxylate (9AC) was evaluated in the absence and presence of CTAB. When buffered to pH 7, 9AC is deprotonated and the anionic anthracene interacts with the positively charged micelle surface through a combination of ion pairing and hydrophobic interactions. This was established earlier through investigation of quenching of micelle forming cationic chromophores with 9AC.50 If a solution contains no micelles, the luminescence of [Ru(bpy)3]2+ and the watersoluble MC5, MC13 and MC17 chromophores are quenched with kq ∼ 3 × 109 M−1 s−1 (Figures S5−S8). Stern−Volmer plots comparing emission intensity changes as a function of 9AC concentration were linear in solutions containing 0.02 M phosphate buffer and 0 M CTAB (Figure S9). While lifetime quenching was not measured, the linearity of quenching in the absence of CTAB was excellent and the difference between quenching by 9AC in aqueous buffer and buffer containing CTAB differed qualitatively. When cationic micelles are present at any concentration, both [Ru(bpy)3]2+ and [Ru(bpy)2MC5]2+ display no measurable luminescence quenching by 9AC, reflecting the fact that 9AC is bound to the CTAB and the Ru chromophores are in the aqueous phase, electrostatically repelled from the micelles. Conversely, photoexcited [Ru(bpy)2MC13]2+, [Ru(bpy)2MC17]2+, and [Ru(bpy)2DC17]2+ efficiently transfer energy to anthracene in the presence of low micelle concentrations. As the concentration of CTAB is increased from 0.01 to 0.05 M, the average number of energy acceptors per micelle and the quenching efficiency decreases. The results provide qualitative evidence that the surfactant complexes spend at least a portion of their excited state lifetime incorporated into the micelle where quenching by 9AC, bound to CTAB micelles, can occur. Spectroelectrochemistry of Ru Complexes. Since the remainder of this manuscript focuses on spectral changes and kinetic behavior associated with the photoinduced reduction of the Ru(II) imine complexes with [Ru(NH3)6]2+, spectra of the reduced chromophores and oxidized [Ru(NH3)6]2+ were obtained. Spectra of each complex in the relevant oxidation states are shown in Figure 1. Reduction of [Ru(bpy)3]2+ provides spectral data to serve as a model for the group of four complexes, although it is acknowledged that small spectral differences may exist for the complexes having one alkylated bipyridine. In reduction of [Ru(bpy)3]2+ the potential of the honeycomb working electrode was held at −1.6 V (vs Ag+/ AgCl), and the absorbance changes of the chromophore were monitored. With a known concentration of chromophore, the molar absorptivity of [Ru(bpy)3]1+ can be calculated, assuming complete reduction and stability of the reduced species on the

Figure 1. Spectra of (A) [Ru(NH3)6]3+/2+ in aqueous solution and (B) [Ru(bpy)3]2+/+. The spectrum of the reduced chromophore was obtained via spectroelectrochemistry in CH3CN/TBAH at −1.6 V vs Ag/AgCl. WE = Gold honeycomb; CE = Pt.

∼5 min time scale. In CH3CN, [Ru(bpy)3]1+ has an absorbance maximum of 510 nm. The reductive quencher, [Ru(NH3)6]2+, absorbs light around 278 nm. Similarly, the absorption spectrum of the oxidized quencher, [Ru(NH3)6]3+, has a maximum at 274 nm. The molar absorptivity of the quencher is small at 278 nm (650 M−1 cm−1), but at concentrations used in transient absorption measurements, nearly complete absorption of the analyzing light occurs at wavelengths below 420 nm. At visible wavelengths where significant absorption of the chromophores occurs, the absorbance of the quencher in both oxidation states is negligibly small, even at concentrations well above those used for luminescence quenching. As a result, only the spectral changes of the chromophore were monitored after photoinduced electron transfer. Transient concentrations of [Ru(bpy)3]1+ were determined using the difference in absorptivity of the two forms of the Ru complex at 510 nm (ε{[Ru(bpy)3]1+} − ε{[Ru(bpy)3]2+}), Δε = 12 000 M−1 cm−1. Δε values of [Ru(bpy)3]2+/+ were used for all chromophores in the series. Luminescence Quenching, Cage Escape, and Back Electron Transfer of Amphiphilic Chromophores with [Ru(NH3)6]2+. Luminescence quenching rate constants for photoreduction of the Ru(II) complexes with [Ru(NH3)6]2+ were evaluated using lifetimes of the Ru(II) chromophore luminescence in deoxygenated solutions and are presented in Table 2. For the complexes soluble in water, quenching rate constants, expressed as KSV, varied by less than a factor of 2 regardless of the alkyl chain length. These values were obtained using concentrations of each chromophore below its critical micelle concentration as reported in previous work. 45 Quenching in aqueous solutions containing CTAB exhibited a strong dependence on the alkyl chain length with KSV of MC5 essentially unchanged relative to water alone, but MC13 and MC17 being 30% and 20% of their respective aqueous quenching values. Addition of NaCl to the solutions containing surfactant resulted in an increase in KSV for all three alkyl derivatives. The KSV of MC5 more than doubled relative to aqueous and micellar values, while the values for MC13 and MC17 increased to effectively eliminate the micellar electrostatic repulsion effect. 8601

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir

Table 2. Excited State Lifetimes in N2 Degassed Solutions in the Presence and Absence of [Ru(NH3)6]2+ and Quenching Rate Constants Obtained [Ru(bpy)2L]2+ complex

[CTAB], M

[NaCl], M

[Ru(NH3)6]2+, M

τ, ns

ksv, M−1

ηq, 1 − (τ/τo)

(bpy)3 (bpy)3 MC5 MC5 MC5 MC13 MC13 MC13 MC17 MC17 MC17 DC17 DC17

0 0.01 0 0.01 0.01 0 0.01 0.01 0 0.01 0.01 0.01 0.01

0 0 0 0 0.2 0 0 0.2 0 0 0.2 0 0.2

0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01

110 100 135 125 60 135 280 95 135 350 125 410 155

420 475 255 285 700 300 90 455 300 54 330 50 295

0.81 0.83 0.72 0.74 0.87 0.75 0.47 0.82 0.75 0.35 0.77 0.32 0.74

Visible transient absorption spectroscopy provided an approach to examine the chromophore excited state in a more detailed way and to examine the back electron transfer reaction between the [(bpy)2Ru(MCx)]+ complex and [Ru(NH3)6]3+ formed in the photoredox reaction. The change in optical density at 360 nm after excitation by 450 nm light was recorded to estimate excited state concentrations in the absence of quencher. The change in absorption at 360 nm resulting from concomitant formation of [Ru(bpy)3]2+* and loss of [Ru(bpy)3]2+ can be used to calculate excited state concentrations formed immediately after laser excitation (Figure S16); literature values of the absorptivity change at 360 nm have been reported; here we will use Δε = 18 000 M−1 cm−1, based on the work of Ohno and Hoffman.19 An assumption is made that the excited state Δε of all the complexes is the same at the maximum of their absorbance between 350 and 400 nm. Excited state concentrations calculated by this approach were used in determining cage escape yields for the {[(bpy)2Ru(MCx)]+/[Ru(NH3)6]3+} geminate pairs formed in the photoredox reaction; a typical transient absorption spectrum obtained for the radical ions along with the spectrum calculated from spectroelectrochemistry (Δελ = ε([Ru(bpy)3]+)λ + ε([Ru(NH3)6]3+)λ − ε([Ru(bpy)3]2+)λ ∼ ε([Ru(bpy)3]+)λ − ε([Ru(bpy)3]2+)λ) is shown in Figure 2. Given the concentration of excited states, the fraction of excited states quenched and the absorptivity of [(bpy)2Ru(MCx)]+ obtained from spectroelectrochemistry, a reliable estimate of the charge separation (cage escape) yield can be made (see the Supporting Information, section E). Ground state concentrations of chromophores were kept to 40 μM to avoid self-aggregation of amphiphilic complexes. The concentration of [Ru(bpy)3]1+ formed in the quenching process is measured from the maximum optical density change at 510 nm after the excited state has completely decayed (∼100 ns). At this wavelength, the only significant chromophore in solution is the reduced Ru complex. The charge separation yields of [Ru(bpy)3]1+ and [Ru(bpy)2MC5]1+are 10% on average, and the yields of reduced species for these two complexes are unaffected by the addition of 0.01 M CTAB. Charge separation yields for [Ru(bpy)2MC13]1+ and [Ru(bpy)2MC17]1+ are dramatically increased in the presence of 0.01 M CTAB. Transient decays obtained with the same sample before and after addition of CTAB are shown in Figure 3 for the MC13 complex. With no micelles present, these chromophores behave similarly to [Ru(bpy)3]2+. Addition of

Figure 2. Transient absorption spectrum obtained following 460 nm excitation of (A) an aqueous solution of [Ru(bpy)3]2+ (10 μM) and [Ru(NH3)6]2+ (0.01 M) and (B) the difference spectrum obtained from Δε = ε([Ru(bpy)3]+) + ε([Ru(NH3)6]3+) − ε([Ru(bpy)3]2+).

0.01 M CTAB gives a cage escape yield of 37% for the MC13 chromophore, and 57% for the MC17 chromophore (Table 3). [Ru(bpy)2DC17]1+ yields were only measured in 0.01 M CTAB, where 63% charge separation was observed. Increases in ionic strength of the solution (to 0.2 M NaCl) resulted in a large increase in the observed cage escape yields for the three surfactant chromophores in the presence of CTAB (Figure 3). The large change occurs due to an increase in the fraction of excited states quenched in the 0.2 M NaCl solution (Tables 2 and 3). A factor that prevented determination of exact rate constants for recombination of the reduced chromophore and [Ru(NH3)6]3+ in this work was the fact that we were unable to prepare solutions of the divalent hexammine quenching complex that were more than 99% pure. [Ru(NH3)6]2+ is prepared by Zn(Hg) amalgam reduction of [Ru(NH3)6]3+ under an inert atmosphere. In our hands, about 1% of the [Ru(NH3)6] remains oxidized in the photoreaction cell, which, with 0.01 M concentrations of the [Ru(NH3)6]2+, results in concentrations of [Ru(NH3)6]3+ on the order of 10−4 M. In determination of rate constants for the reaction of reduced chromophore with oxidized quencher, transient decays measured at 510 nm were fit by an exponential decay function, 8602

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir

solution the measured rate is k = 4.5 × 105 s−1. The kinetics of the back reaction in a 0.01 M CTAB solution in the case of the MC5 chromophore has two first order components, with the fast component having a rate constant of 6.3 × 105 s−1. The slow component shows a 200-fold decrease in the rate constant, with k = 2.3 × 103 s−1. Decays of the reduced MC5 complex absorption in the absence and presence of CTAB are shown in Figure 4. Further increases of the CTAB concentration for this chromophore/quencher mixture result in a significant increase in the fraction of the slow component of the decay. Both the MC13 and MC17 chromophores show back reaction kinetics that differ from the kinetics of [Ru(bpy)3]1+ and [Ru(bpy)2MC5]1+ in the presence of micelles. In aqueous solutions lacking surfactant [Ru(bpy)2MC13]1+ and [Ru(bpy)2MC17]1+ back react with [Ru(NH3)6]3+ with rate constants of 8.2 × 105 and 1.5 × 106 s−1, respectively. When CTAB is added to make the 0.01 M surfactant solutions, the kinetics are single exponential with a significantly decreased rate constant for the back reaction (k = 3.0 × 103 s−1 and k = 8.1 × 103 s−1 for MC13 and MC17, respectively). No fast component is seen in these back reactions, as is seen with the MC5 chromophore; Figure 5 shows representative decays for the MC13 complex. [Ru(bpy)2DC17]1+, soluble only in solutions containing surfactant, back reacts with [Ru(NH3)6]3+ in a 0.01 M CTAB solution with k = 5.4 × 103 s−1 and thus, shows no significant difference from the MC13 and MC17 complexes. Ionic Strength Effects on Excited State Quenching, Cage Escape Yield, and Back Reaction Rate. Chromophores that showed decreased back electron transfer rates in micellar solutions were studied under high ionic strength conditions. At 0.01 M [Ru(NH3)6]2+, the ionic strength of the solution is 0.03 M; addition of 0.01 M CTAB increases the ionic strength to 0.04 M. Ionic strength values of 0.24 M were achieved through the addition of 0.2 M NaCl to the micellar solutions. Quenching efficiencies of the MC13, MC17, and DC17 chromophores under these higher ionic strength conditions increased to greater than 0.75 (Table 2), which is quite similar to values obtained in nonmicellar solutions. In

Figure 3. Kinetic traces for back reaction of [Ru(bpy)2(MC13)]2+ with [Ru(NH3)6]3+ in aqueous solution before (0 M CTAB) and after (0.01 M CTAB) addition of cationic surfactant as well as following addition of NaCl to surfactant containing solution. Time zero for the traces is offset intentionally.

assuming the back reaction follows pseudo-first-order kinetics. In order to compare the effect of CTAB on back electron transfer rates of the various chromophores under consistent conditions, the same chromophore/quencher solution was used for both 0 and 0.01 M CTAB measurements. In this way, the relative effect of CTAB on the back electron transfer rate constant can be evaluated from the ratio of the pseudo-firstorder back electron transfer rate constants obtained in the absence and presence of CTAB. Back reaction of [Ru(bpy)3]1+ with [Ru(NH3)6]3+ has a pseudo-first-order rate constant of 1.2 × 106 s−1. When 0.01 M CTAB is added to the chromophore and quencher solution, the recombination rate constant differs only slightly (Figure S10). For [Ru(bpy)2MC5]1+ and [Ru(NH3)6]3+ in a 0 M CTAB

Table 3. Charge Separation Effiiciencies, Relative Back Electron Transfer Rate Constants and the Ratios of the Back Electron Transfer Rate Constant in the Absence of CTAB to that with Added Micelles for each of the Chromophoresa [Ru(bpy)2L]2+ complex

[CTAB], M

[NaCl], M

(bpy)3 (bpy)3 MC5 MC5 MC5 MC5 MC5 MC13 MC13 MC13 MC13 MC13 MC17 MC17 MC17 DC17 DC17

0 0.01 0 0.01 0.01 0.025 0.05 0 0.01 0.01 0.025 0.05 0 0.01 0.01 0.01 0.01

0 0 0 0 0.2 0.2 0.2 0 0 0.2 0.2 0.2 0 0 0.2 0 0.2

ηq 0.81 0.83 0.72,a 0.74a 0.87,a 0.89b 0.89b 0.75,a 0.47a 0.82,a 0.80b 0.80b 0.75 0.35 0.77 0.32 0.74

0.77b 0.89b

0.76b 0.80b

ηcs 0.11 0.11 0.10,a 0.10a 0.09,a 0.12b 0.15b 0.11,a 0.37a 0.50,a 0.55b 0.58b 0.08 0.57 0.60 0.63 0.59

0.11b 0.11b

0.12b 0.53b

kback, 510 nm (×10−3 M−1 s−1) 1200 1600 450,a 400b 2a 13,a 15b 18b 21b 820,a 690b 3a a 34, 36b 42b 39b 1500 8 72 5 99

ratio of kback 0 M/0.01 M CTAB 0.75 200a 35,a 27b 22b 19b 273 24,a 19b 16b 18b 189 21

a

Two different samples, denoted by superscripts a and b, (prepared with separate Zn(Hg) amalgam reductions), were measured with all other solution conditions unchanged. 8603

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Langmuir

■

Article

DISCUSSION

The location of the chromophores in aqueous CTAB solutions can be inferred from the wavelength of the maximum emission and the degree of interaction with a quencher known to be associated with the micelles. A decrease in energy of the luminescence maximum is seen when chromophores of this type self-aggregate;50,51 observation of red-shifted emission energy when micelles are added to a solution of chromophore suggests association of the dicationic MC13, MC17, and DC17 chromophores with CTAB micelles. The amount of excited state quenching in the absence and presence of CTAB gives evidence of whether the dicationic chromophores are associated with micelles or exist mainly in the aqueous bulk. For solutions of 9AC in the presence of CTAB, excited state quenching of [Ru(bpy)3]2+ and [Ru(bpy)2MC5]2+ is ineffective because these two chromophores are repelled from the cationic 9AC/CTAB micellar aggregates (Figures S5 and S6); in these solutions (0.01 M CTAB, pH 7) essentially all the 9AC is bound to CTAB. The [Ru(bpy)2(MCX)]2+ chromophores bearing an MC13 or MC17 ligand associate with the micelles, based on the fact that these complexes are significantly quenched by 9AC bound to CTAB (Figures S7 and S8). Quantitation of the association equilibrium constant of the MC17 and MC13 complexes with CTAB was not pursued from this data. The results for 9AC quenching strongly suggest the MC5 complex resides in the aqueous phase as the 2+ ion, but the MC13 and MC17 complexes partition primarily into the CTAB micelles. While the hydrophobic chromophores associate with the micelles to some extent in their ground states, the one electron reduced chromophores are certainly more readily incorporated into the CTAB micelles due to the decreased positive charge on the metal complex. Reductive quenching of the excited states of the complexes can be achieved through a bimolecular reaction with [Ru(NH3)6]2+. While the ground state reduction potential of [Ru(bpy)3]2+ is −1.28 V vs NHE, its excited state energy is 2.12 eV, yielding an excited state reduction potential of 0.84 V vs NHE.52 The 3+/2+ potential of [Ru(NH3)6]3+ is reported as 0.10 V vs NHE.53 Thus, photoreduction of [Ru(bpy)3]2+ by [Ru(NH3)6]2+ is favorable by ∼0.7 V and the recombination is exergonic by nearly 1.2 V. Potentials for the alkyl substituted derivatives, while not identical, are similarly highly exergonic for both excited state quenching and back electron transfer. Excited state quenching of the complexes is affected by ground state association of the chromophores with added CTAB. The two complexes that exhibit no emission spectral changes and large 9AC quenching changes upon introduction of CTAB, [Ru(bpy)3]2+ and [Ru(bpy)2MC5]2+, also exhibit little change in quenching by [Ru(NH3)6]2+ following addition of 0.01 M CTAB. This is clear evidence that these complexes are essentially completely in the aqueous phase and are not associated with the micelles. The MC13 and MC17 complexes have luminescence that red shifts by ∼10 nm and 9AC quenching that remains significant following addition of CTAB, suggesting that the complexes associate in the ground state to some degree with the cationic micelles. A small change in the reduction potential of the hydrophobic complexes may accompany micelle association. A shift to a less negative reduction potential would be expected, and this would result in a higher quenching efficiency. However, the two surfactant Ru chromophores exhibit reduced quenching by [Ru(NH3)6]2+ upon addition of CTAB (Table 3). For the MC13 complex in

Figure 4. Kinetic traces at 510 nm for the back reactions of (A) [Ru(bpy)2MC5]1+ and (B) [Ru(bpy)2MC13]1+ with [Ru(NH3)6]3+ in the absence and presence of CTAB micelles (0, 0.01, 0.025, and 0.05 M); 0.2 M NaCl added to solutions with CTAB to diminish ionic strength effects. (C) Charge separation yields as a function CTAB concentration.

Figure 5. Decays of reduced chromophore ([Ru(bpy)2MC13]+) following pulsed laser excitation of solutions of the chromophore, [Ru(NH3)6]2+ (0.01 M), and CTAB (0.01 M) in the absence and presence of 0.2 M NaCl.

addition, charge separation yields remain high under high ionic strength conditions. However, the back electron transfer rate constants for the MC13, MC17, and DC17 complexes increase by an order of magnitude when compared to lower ionic strength, 0.01 M CTAB solutions. 8604

DOI: 10.1021/acs.langmuir.6b02193 Langmuir 2016, 32, 8598−8607

Article

Langmuir the presence of 10 mM [Ru(NH3)6]2+, the fraction of excited state quenching drops from 75% to 47% upon addition of 10 mM CTAB, while that of MC17 decreases from 75% to 35%. The observed change in quenching further supports the idea of ground state association of the MC13 and MC17 complexes with the cationic micelles. The electrostatic repulsion of the positively charged micelle and the cationic reductive quencher serves to reduce the quenching efficiency. It is interesting to note that the quenching rate constants remain 109 M−1 s−1 or greater, even for complexes associated with the cationic micelles. Additionally, only a small decrease in quenching efficiency is observed upon CTAB addition to solutions with higher ionic strength. The relative amount of observable [Ru(bpy)2(MCx)]+ produced in reductive quenching is dependent upon the rate of back electron transfer in the {[Ru(bpy)2(MCx)]+, [Ru(NH3)6]3+} encounter complex and the rate of escape of ions from the cage. Changes in charge separation yield with changes in the complex studied or the addition of cationic surfactant require either a decrease in the rate of back electron transfer or an increase in the cage escape rate. In the absence of cationic micelles, charge separation yields measured for all four complexes are 0.1 ± 0.01. In the presence of CTAB micelles charge separation efficiencies for [Ru(bpy)3]2+ and [Ru(bpy)2MC5]2+ remain around 0.1, while those for the MC13 and MC17 complexes are dramatically increased (0.37 for MC13 and 0.57 for MC17; Table 3; Figure 3). A further significant increase in the concentration of the reduced MC13 and MC17 complexes formed through cage escape is observed upon increasing the solution ionic strength, but this can be nearly completely explained by the much higher fraction of excited states quenched by the [Ru(NH3)6]2+ that also results from increasing the salt concentration (Table 3). The calculated charge separation yields, which take into account the fraction of excited states quenched, are not significantly increased by ionic strength increases in 0.01 M CTAB solutions. At 0.01 M CTAB, the concentration of micelles is 150 μM, greatly exceeding both the ruthenium excited state and reduced chromophore concentrations formed in pulsed laser excitation (