Water Photo-oxidation Initiated by Surface-Bound Organic

Subscriber access provided by LAURENTIAN UNIV

Article

Water Photo-Oxidation Initiated by Surface-Bound Organic Chromophores Michael S. Eberhart, Degao Wang, Renato N. Sampaio, Seth L. Marquard, Bing Shan, M. Kyle Brennaman, Gerald J. Meyer, Christopher Dares, and Thomas J. Meyer J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08317 • Publication Date (Web): 12 Oct 2017 Downloaded from http://pubs.acs.org on October 12, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of the American Chemical Society

Water Photo-Oxidation Initiated by Surface-Bound Organic Chromophores Michael S. Eberhart†,1, Degao Wang†,1, Renato N. Sampaio1, Seth L. Marquard1, Bing Shan1, M. Kyle Brennaman1, Gerald J. Meyer1, Christopher Dares2, and Thomas J. Meyer*,1 1

Department of Chemistry, University of North Carolina at Chapel Hill, CB 3290, Chapel Hill, North Carolina 27599 2 Department of Chemistry and Biochemistry, Florida International University, MMC 11200 SW 8th Street (CP-3044), Miami, Florida 33199 Abstract Organic chromophores can be synthesized by established methods and offer an opportunity to expand overall solar spectrum utilization for dye sensitized photoelectrosynthesis cells (DSPECs). However, there are complications in the use of organic chromophores arising from the instability of their oxidized forms, the inability of their oxidized forms to activate a water oxidation catalyst, or the absence of a sufficiently reducing excited state for electron injection into appropriate semiconductors. Three new triarylamine donor-acceptor organic dyes have been investigated here for visible light driven water oxidation. They offer highly oxidizing potentials (>1 V vs NHE in aqueous solution) that are sufficient to drive a water oxidation catalyst and excited state potentials (~-1.2 V vs NHE) sufficient to inject into TiO2. The oxidized form of one of the chromophores is sufficiently stable to exhibit reversible electrochemistry in aqueous solution. The chromophores also have favorable photophysics. Visible light-driven oxygen production by an organic chromophore for up to one hour of operation has been demonstrated with reasonable faradaic efficiencies for measured O2 production. The properties of organic chromophores necessary for successfully driving water oxidation in a light driven system are explored along with strategies for improving device performance. Introduction

ACS Paragon Plus Environment

1


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In a dye sensitized photoelectrosynthesis cell (DSPEC), semiconductors and molecular assemblies are integrated for photoelectrochemical water splitting or CO2 reduction. Molecular assemblies provide the chromophore and water oxidation catalyst while the semiconductors provide the electronic and physical properties to accept and transport charge. In this application, organic dyes could play important roles as chromophores given the extensive background synthetic chemistry, their relative ease of synthesis and the corresponding ease with which the dye properties can be modified through synthesis. Previous examples of artificial photosynthesis devices with organic chromophores have included high potential porphyrins,1-6 a subporphyrin,7 a donor-acceptor dye,8 donor-acceptor dyes combined with a WO3 photocatalyst,9 and PDI dyes.10-13 A review article highlighting the challenges in using organic dyes for water splitting has also recently appeared on the subject.14 Organic dyes that offer the required performance in light absorption and the ability to drive water oxidation for extended periods that are competitive with the metal-to-ligand charge transfer (MLCT) excited states of Ru(bpy)32+ and other polypyridyl derivatives, are not available in the literature. A chromophore for a photoanode should have redox stability, especially in aqueous environments, an excited state potential that is sufficiently reducing to inject an electron into a semiconductor substrate, and a ground state redox potential that is sufficiently oxidizing to activate a water oxidation catalyst. We describe here, three new donor-acceptor organic dyes. Each of the dyes can function as a high-energy sensitizer in Dye Sensitized Photoelectrosynthesis Cells (DSPECs). All have redox potentials sufficient to drive water oxidation by the molecular water oxidation catalyst, Ru(bda)(pyP)2 (4) (bda is 2,2’-bipyridine-6,6’-dicarboxylate; pyP is (3-(pyridin-4yloxy)propyl)phosphonic acid).15-17 One of the oxidized dyes is stable in water on the cyclic


2

Page 2 of 28

Page 3 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


voltammetry timescale and all three dyes are stable enough to be incorporated into DSPEC photoanodes for the production of oxygen. Our hypothesis is that organic chromophores can be used in water oxidizing devices, with performance and longevity mostly depending on the redox stability of the chromophores in aqueous solution, the thermodynamics of electron transfer reactions that are required to activate a water oxidation catalyst, and the thermodynamics for electron injection into a semiconductor substrate from the chromophore’s excited state. By removing functional groups that may be vulnerable to oxidative degradation, we show that there is a clear route to increasing the oxidative stability of organic chromophores. The results provide guidance on the characteristics that are necessary for organic chromophores in a water oxidizing device and factors that can lead to stability or instability in water oxidizing devices. Results and Discussion O

O N

N

N

PO3H2

O NC

NC

NC

PO3H2/PO3Et2

PO3H2/PO3Et2

PO3H2/PO3Et2

2/2-PO3Et2

1/1-PO3Et2

O

N RuII

O 4

N

H 2O 3P CO2H 1b

O O

N N

N

Ru N

N

N

N

N

O 2+

NC

N

3/3-PO3Et2

PO3H2 RuP2+

PO3H2

Figure 1. (1) (1-cyano-2-(4-(diphenylamino)phenyl)vinyl)phosphonic acid, (2) (1-cyano-2-(4(di-p-tolylamino)phenyl)vinyl)phosphonic acid, (3) (2-(4-(bis(4-methoxyphenyl)amino)phenyl)1-cyanovinyl)phosphonic acid, (1b) 2-cyano-3-(4-(diphenylamino)phenyl)acrylic acid, and


3


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Ru(4,4’-(PO3H2)2bpy)(bpy)22+ (RuP2+) and the water oxidation catalyst Ru(bda)(pyP)2 (4). Note that for 1, 2, and 3, the corresponding esters (1-PO3Et2, 2-PO3Et2, and 3-PO3Et2) were also studied.

The dyes investigated, 1, 2, and 3, are shown in Figure 1. They were bound to oxide surfaces by phosphonate-to-surface oxide binding which is stable in water at pH < 5.18-19 The related carboxylate derivative, 1b, has been incorporated in a tandem device and can produce photocurrents up to 300 µA·cm-2 using a different catalyst and under conditions different from those of this study, although the photocurrents decay rapidly in the first 5-10 minutes of operation.2 In aqueous solutions, phosphonate-surface binding is far more stable and, when combined with atomic layer deposition (ALD), provides a method to create more stable surface structures.20-22 The synthesis of the dyes was simple and economical. Chromophores 1 and 2 were synthesized in two steps from commercially available precursors and the synthesis of chromophore 3 was accomplished in three steps from commercially available materials. The synthesis was conducted on a relatively small scale, obtaining 1, 2, and 3 in 78 mg, 85 mg, and 35 mg yields respectively. The new dyes were characterized by NMR and HRMS. Construction of the photoanodes was accomplished by loading chromophores 1, 2, or 3 onto 15 nm, SnO2/TiO2(45 Å) core/shell 4 μm electrodes on fluorine-doped tin oxide (FTO) by using a procedure described in the literature.23 The core/shell structure of the electrodes exploits the potential difference between the TiO2 conduction band (-0.55 V vs NHE at pH 7) and the SnO2 conduction band (0.00 V vs NHE at pH 7) to transfer electrons away from oxidized


4

Page 4 of 28

Page 5 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


chromophores on the external surface, slowing back electron transfer and enhancing photocurrent.23-26 After loading the chromophore on the surface, an overlayer of alumina (3 Å-17 Å) was added by atomic layer deposition (ALD). The ALD overlayer protects the phosphonated metal complex surface from surface hydrolysis20-22 and provides an external metal oxide surface for catalyst binding, Ru(bda)(pyP)2 (4) as shown in Figure 2. PO3H2

PO3H2

O

O

N O

Ru

O

O

N

N N

N

PO3H2

N

O

O

Ru

O

O

N

O

N

N

O

O

O

N

N

O

H 2O 3P Al2O3

Al2O3

Al2O3

Al2O3 Al2O3

N Al2O3

CN

Al2O3

Al2O3

R R

O

O

H 2O 3P Al2O3 Al2O3

R

Al2O3

Al2O3

Al2O3

N

O

O

PO3H2

Ru

N

R

Al2O3

Al2O3

PO3H2

CN H 2O 3P

TiO2 Shell SnO2 Core

Figure 2. Surface structure of the SnO2/TiO2|dye(Al2O3)-Ru(bda)(PyP)2 chromophorecatalyst electrode assembly with R = -H, -Me, -OMe. Chromophore 1 has an absorption maximum at λmax = 377 nm when loaded on nanoITO (in air). The electron donating substituents –Me and –OMe in 2 and 3 cause red-shifts in the UV-vis spectrum, to 384 nm for 2 and to 390 nm for 3, Figure 3. The molar absorptivity of 1PO3Et2 is 3.52(7) × 105 M-1 cm-1 in CH2Cl2 (λmax = 412 nm), about twice that for RuP2+ at its λmax of 460 nm. A combination of 1, 2, or 3, with RuP2+ may be useful for photoanodes with increased total absorbance and photon utilization to achieve higher photocurrents.


5


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3. Normalized Absorption spectra for 1, 2, 3, and RuP2+ on nanoITO in air. The spectra were normalized so that the local absorption maxima near 400 nm were 1. Although visible light absorption is limited, relatively high photocurrents are still possible. Under ~1 sun illumination (100 mW cm-2) with a Thorlabs Inc. HPLS-30-04 light source, 1 produces photocurrents of > 400 µA/cm2 (Figure 4) with similar photocurrent magnitudes for 2 and 3 as shown in Figures S1 and S2. The results were reproducible (±20 %) for additional electrodes produced by identical methods. The role of the added alumina protection layer was also explored. Surface stabilization of phosphonate-bound assemblies has been demonstrated previously with ALD-added alumina21-22 or with polymer overlayers.27-28 A previous study has examined the effects of alumina overlayers on PDI dyes on the ultrafast time scale.13 A minimum external layer of alumina (at least 3 ALD cycles, 3 Å) was required to obtain the chromophore-catalyst structures shown in Figure 1. Without added alumina, poor subsequent loading of water oxidation catalyst 4 (Figure S11) resulted in poor performance. The hydrophobic nature of 1, 2, and 3 inhibits binding of the phosphonic acid groups from 4 directly on the TiO2 surface necessitating the layer of alumina.


6

Page 6 of 28

Page 7 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


In the photoanode structures that were investigated, thinner layers of alumina, ~7-10 Å, gave higher photocurrents. Thicker layers of alumina gave lower photocurrents that result from poor electron transfer from the water oxidation catalyst to the oxidized chromophore. Although thicker alumina layers assist with retention of the chromophores on the surface20-22 and the loading of 4, there was a net loss in reactivity attributed to an insulating effect between the chromophores and catalyst.


7


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4. (top) Photocurrent vs time responses for SnO2/TiO2|(1)(xÅ Al2O3)-4 under ~1 sun illumination at the indicated Al2O3 overlayer thickness. (bottom) Incident Photon to Current Efficiencies (IPCE) in the electrodes: SnO2/TiO2|(1,2, or 3)(7 Å Al2O3)-4, as in Figure 2. The electrolyte consisted of 0.1 M H3COOH/H3COONa, 0.4 M NaClO4 and an external bias of 0.4 V vs NHE. The photocurrents obtained under ~1 sun illumination with 1 were approximately the same as those obtained for RuP2+ under the same conditions (Figure S3) despite the limited light absorption. Those of 2 and 3 were somewhat smaller, about 60% as high (Figures S1 and S2). Figure S12 shows photocurrents as a function of applied potential. The photocurrent action spectra in Figure 4 reveal that up to 30% of 400 nm photons are converted to current. The IPCE values proved reproducible within 20% on repeated trials of identical electrodes. Film Characterization. The high photocurrents demonstrate that the excited states of 1, 2, and 3 are capable of injecting an electron into TiO2 in acetate buffer with ECB ~ -0.4 V vs NHE.29 Given the substituents on the dye, the oxidized forms of chromophores 2 and 3 were expected to be more stable than 1. Under non-aqueous conditions, unsubstituted triphenylamine derivatives are known to undergo follow-up reactions upon oxidation to give the corresponding substituted benzidine derivatives.30-32 Substituents in the para position have been shown to block this reactivity. To probe the redox stability of the dyes, their electrochemistry was studied on nanoITO surfaces in acetonitrile and aqueous solution. The electrode material, nanoITO, was used because it is a conductor suitable for cyclic voltammetry unlike the semiconducting SnO2/TiO2 core shells used in the photoelectrochemical studies while still providing a metal oxide surface on which to anchor the dyes. In acetonitrile, 0.1M [nBu4N][PF6], ipa/ipc = 0.48 for 1 while for 2 and 3, the waves are fully reversible, ipa/ipc = 1.0 ± 0.05 (Figure 5). Dimerization of 1


8

Page 8 of 28

Page 9 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


is a possible pathway for its degradation as this reactivity is well established in the literature for triphenylamines. 30-32

Figure 5. Cyclic voltammograms for 1, 2, and 3 on nanoITO in 0.1M [nBu4N][PF6] acetonitrile solutions at 20 mV/sec.

The non-aqueous E1/2 values for the 1•+/0, 2•+/0, and 3•+/0 couples were found to be 0.59 V, 0.57 V, and 0.38 V vs Fc+/0 respectively. Peak separations, Epa - Epc, for the 1•+/0, 2•+/0, and 3•+/0, couples were 147 mV, 96 mV, and 87 mV, respectively, compared to 0 V for an ideal, reversible surface-bound couple. For comparison, for the 2-PO3Et2•+/0 couple in solution at 100 mV/s, thepeak-to-peak separation was nearly reversible with Epa-Epc = 63 mV at a Pt disk working electrode (Figure S7), the solution E1/2 was 0.58 V vs Fc+/0, essentially the same as the corresponding surface couple. The slight excess peak separation of 2-PO3Et2 is likely due to solution resistance which is common in non-aqueous solvents and demonstrates that the redox process is electrochemically reversible. The peak-to-peak separations observed for the surface-


9


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

bound 1, 2, and 3 couples are likely a result of surface resistance effects in the FTO conductive glass substrate and nanoITO.33 Under aqueous conditions, the electrochemistry of 1, 2, and 3 is more complex. Cyclic voltammograms are shown in Figure 6 in 0.1 M KPF6.

Figure 6. Cyclic voltammograms for 1 (left), 2 (middle), and 3 (right) on nanoITO in 0.1 M aqueous 0.1 M KPF6 at 20 mV/s. The second scan of 2 was unchanged from the first scan.

As in acetonitrile, the aqueous electrochemistry of 1 is only partially reversible with E1/2 = 1.04 V vs NHE for the 1+•/0 couple. On subsequent voltammetric scans at 20 mV/sec, a new species is observed with E1/2 = 0.5 V vs NHE, this may be due to dimerization which would yield a benzidine species that is oxidized at lower potential.30-32 Upon oxidation, the appearance of a reversible wave at 0.50 V vs NHE in subsequent scans may be attributable to dimerization. In contrast to 1, 2 (E1/2 = 1.03 V vs NHE) was found to be considerably more stable with no signs of degradation. Surface desorption was observed, but the absence of new waves in voltammograms was consistent with slow loss of 2 from the surface rather than decomposition. In CV scans of 3, a high potential anodic wave at Ep,a = 1.13 V vs NHE was observed, consistent with oxidation to the 3+•, but in scans at 20 mV/s, there was no evidence for an associated rereduction. The 3+•/0 couple was irreversible up to 100 mV/s. A partially reversible wave


10

Page 10 of 28

Page 11 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


appearing at E1/2 = 0.53 V vs NHE may be attributable to hydrolysis of the –OMe substituents upon oxidation (Figure 6, right). The UV-visible spectrum of 1+• was obtained by spectroelectrochemical oxidation of 1 on nanoITO in a pH 4.8 acetate buffer solution (Figure 7). The oxidized dye displays a characteristic absorption feature at ~740 nm.

Figure 7. Spectroelectrochemical oxidation of 1 on nanoITO in an acetate buffer (0.1 M H3COOH/H3COONa, 0.4 M NaClO4).

Photophysics. Excited state potentials for the redox couples, 1•+/0* and 2•+/0*, were evaluated from the Eo’ values for the ground state couples, 1•+/0, and the 0-0 energy for the excited state (equation 1). The latter were obtained by room temperature emission spectral fitting,34-36 Figures S4-S6. Based on the electrochemical data, and E° values of 2.92 and 2.75 eV from the spectral fits, the excited state potentials were found to be -1.88 V for 1 and -1.72 V for 2 vs. NHE in 0.1 M KPF6 as calculated using equation 1.34-36 The values for E° and Δν 0,1/2 were obtained from the spectral fits shown in the SI.


11


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

•+ /0*

•+ /0

(1)

E= (1 ) E (1

(2)

 ∆GES =E  + λ

(3)



λ=



( ∆ν )

 ∆GES )− F

2

0,1/2

16k BT ln(2)

Transient absorption difference spectra were obtained following 460 nm excitation of films of SnO2/TiO2|1 with results shown in Figure 8. The absorption feature at 700 nm is characteristic of 1•+, consistent with excitation and injection by 1* to give 1•+. The observed λmax differs slightly from that observed upon oxidation on nanoITO due to the different environment of the SnO2/TiO2 core/shell electrode with an ALD alumina overlayer compared to the nanoITO electrode used for spectroelectrochemistry with no ALD overlayer.


12

Page 12 of 28

Page 13 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8. Transient absorption difference spectra following 460 nm pulsed laser excitation of SnO2/TiO2|-1(7 Å Al2O3) in pH 4.8 acetate buffer (0.1 M H3COOH/H3COONa, 0.4 M NaClO4) from 30 nsec to 80 μsec.

Figure 9. Single wavelength decay kinetics for SnO2/TiO2|-1(xÅ Al2O3) monitored at 440 nm at the indicated Al2O3 overlayer thicknesses from 0 to 17 Å. Data were fit to the stretched exponential, Kohlrausch-Williams-Watts (KWW) function.

Given the impact of the added Al2O3 layer on photocurrent response in Figure 4, transient spectroscopy was used to explore the effect of the layer thickness on light driven electron transfer. Charge recombination kinetics of the injected electrons with the oxidized dyes were not well modeled by a single exponential, but well modeled by the Kohlrausch-William-Watts function (eq 4).37 The value of β is inversely related to the breadth of the underlying Lévy distribution of the rate constants, 0 < β < 1, A0 is the initial absorbance, and k is the characteristic


13


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

observed rate constant. An averaged charge recombination rate constant, kcr, was calculated with eq. 5 where Γ is the Gamma function and a value of β = 0.2 provided the best fit. The stretched exponential kinetics may result from a dispersive transport behavior of charge carriers where transport is rate-limited by thermally activated trapping-detrapping events.38-39 (4)

A(t ) A0 exp(− kt ) β =

(5)

 1  1  kcr  = Γ    k β  β 

−1

A comparison of relative injection efficiencies for 1* core/shell electrodes as the thickness of the Al2O3 layer was increased showed that the injection efficiency was essentially constant over the range, 0 to ~17 Å (figure S8). Also, shown in the figure, are data for the role of Al2O3 coverage with RuP2+ as the dye. For the latter, added Al2O3 decreases the injection yield by more than a factor of 2. As shown in Figures 9 and S9, a similar effect was observed for back electron transfer to the electrode for 1 compared to RuP3+. The injection data do give insight into the effect of Al2O3 on photocurrents in Figure 4. A reaction scheme for activation of the assembly is shown in equations 6-8. The first step after light excitation is injection of an electron into TiO2 (eq 6) followed by activation of the catalyst (eq 7) which is in competition with back electron transfer (eq 8). Eq 6 dictates the yield of electrons injected into TiO2 in terms of photons absorbed by the dye. Photocurrent is maximized when the relative rates of eq 6 and eq 7 are maximized. The observed photocurrent is maximized at ~7-10 Å Al2O3. With thinner layers of Al2O3, the rate of catalyst activation is slowed due to poor catalyst loading. At thicker layers of Al2O3, the rate of catalyst activation is also slowed due to the insulating effect of Al2O3 slowing the rate of electron transfer from the catalyst to the


14

Page 14 of 28

Page 15 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


oxidized chromophore. Similar behavior has been observed previously on the ultrafast time scale for PDI dyes.13

(6) FTO|SnO2/TiO2|1*(xÅ Al2O3)-Ru(bda)(pyP)2 → FTO|SnO2(e-)/TiO2|1+•(xÅ Al2O3)-Ru(bda)(pyP)2

(7) FTO|SnO2(e-)/TiO2|1+•(xÅ Al2O3)-Ru(bda)(pyP)2 → FTO|SnO2(e-)/TiO2|1(xÅ Al2O3)-Ru(bda)(pyP)2+

(8) FTO|SnO2(e-)/TiO2|1+•(xÅ Al2O3)-Ru(bda)(pyP)2 → FTO|SnO2/TiO2|1(xÅ Al2O3)-Ru(bda)(pyP)2

Water Oxidation Oxygen Measurements. The stabilities of the DSPEC electrodes and their ability to generate oxygen for extended periods were explored by use of a collector-generator cell. In the collector-generator cell, oxygen is generated at the photoanode, the “generator” electrode. A closely spaced, transparent conductive oxide (FTO glass) electrode was positioned 1 mm from the collector for the evaluation of O2 at a potential suitable for oxygen reduction (-0.65 V vs NHE).40


15


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 10. Current and charge passed from a collector-generator cell, SnO2/TiO2|-2(7 Å Al2O3)4, in acetate buffer (0.1 M H3COOH/H3COONa, 0.4 M NaClO4) under ~ 1 sun white light illumination for 3600 seconds with an applied bias of 0.4 V vs NHE. The solid traces show photocurrents and the dashed lines show the charge passed for the generator (red) and collector (blue) electrodes, respectively.

Table 1. Faradaic efficiencies from 1 sun collector-generator cell measurements for photoanodes with the chromophores 1, 2, and 3 and the catalyst 4, SnO2/TiO2|-1(xÅ Al2O3)-4 with the indicated amounts of added Al2O3 in 0.1 M H3COOH/H3COONa in 0.4 M NaClO4.

7 Å Al2O3 10 Å Al2O3 13 Å Al2O3 17 Å Al2O3

1 79% 82% 50% 24%

2 63% 51% 45% 29%

3 47% 55% 32% -

The maximum efficiency for the appearance of O2 reached ~80% for 1. Efficiencies for dyes 2 and 3 were comparable but lower in magnitude. With thicker layers of added Al2O3, the catalyst is farther from the dyes and faradaic efficiency for O2 production decreased considerably


16

Page 16 of 28

Page 17 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(

Water Photo-oxidation Initiated by Surface-Bound Organic

Recommend Documents