2216

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATI

ON

Ming Shao , Liang Yan , Haiping Pan , Ilia Ivanov , and Bin Hu *

Giant Magnetic Field Effects on Electroluminescence in Electrochemical Cells

Recently, there has been growing interest in the magnetic fi eld effects; an external magnetic fi eld can substantially change elec-troluminescence, [ 1 ] photoluminescence, [ 2 ] photocurrent, [ 3 ] and electrical current [ 4 ] in non-magnetic organic semiconducting materials with potential applications for magneto-electronics, magneto-optics, and magneto-optoelectronics. In general, three types of magnetic fi eld effects can be observed based on intercharge spin–spin interaction, spin-dependent excited proc-esses, and Lorentz force effects. First, when intercharge spin–spin interaction occurs, an external magnetic fi eld can perturb the spin–spin interaction and consequently changes singlet and triplet formation ratios in excited states [ 5 ] and carrier mobilities in charge transport. [ 6 ] Second, an external mag-netic fi eld can affect spin-dependent excited processes, such as singlet–triplet intersystem crossing, triplet–charge reaction, and triplet–triplet annihilation after the formation of excited states by involving in spin moment conservation required for those excited processes, [ 1 , 2 , 7 – 9 ] and essentially changes both singlet and triplet ratios in excited states [ 1 , 5 , 9 ] and carrier den-sities in charge transport. [ 10 ] Third, an external magnetic fi eld can introduce a Lorentz force exerted on moving charged spe-cies and changes charge transport and consequently generates magnetocurrent (MC). [ 11 , 12 ] In principle, magnetic fi eld effects can occur in both solid and liquid states. In liquid states early experimental studies have found that electrochemical reaction can show considerable magnetic fi eld effects on electrolumi-nescence intensity (MFE EL ) with the amplitude less than 30% [ 13 ] with a suggested mechanism of triplet–charge reaction and triplet–triplet annihilation. [ 14–17 ] In this paper, we report a giant MFE EL with a magnitude larger than 400% in liquid states by using conveniently controllable electrochemical co-reaction in aqueous solution based on Lorentz force effects.

Figure 1 a shows the electrogenerated chemiluminescence spectrum based on the triplet emission from the tris(2–2’-bipy-ridyl) ruthenium(II) (Ru(bpy) 3 2 + ) molecules. The spectral peak

© 2011 WILEY-VCH Verlag Gwileyonlinelibrary.com

M. Shao , L. Yan , Prof. B. Hu Department of Materials Science and EngineeringUniversity of TennesseeKnoxville, TN 37996, USA E-mail: bhu@utk.edu Dr. H. Pan Huazhong University of Science and TechnologyWu Han 430074, China Dr. I. Ivanov Oak Ridge National LaboratoryOak Ridge, TN 37831, USA

DOI: 10.1002/adma.201100193

at 610 nm is characteristic of phosphorescence from the triplet states of the Ru(bpy) 3 2 + through the electrochemical reaction [ 18 ] shown in Equation (1) .

Ru(bpy)3+3 + T Pr A• −→ Ru(bpy)2+∗

3 + products (1)

The Ru(bpy) 3 2 + metal chelate complex has been widely studied for electrogenerated chemiluminescence due to its high luminescence effi ciency and electrochemical stability in aqueous solvents. [ 19 , 20 ] The tripropylamine (TPrA) works as an effi cient co-reactant with Ru(bpy) 3 2 + upon electrochemical oxidation via a catalytic route. In general, the electrogenerated chemilumi-nescence can be divided into energy-defi cient and energy-suf-fi cient systems through triplet and singlet route emission, [ 21 ] respectively. The early studies have indicated that the oxidation of TPrA by electrogenerated Ru(bpy) 3 3 + is the dominant process in the generation of chemiluminescence. [ 22 ] The free energy released from ion annihilation is insuffi cient to generate sin-glet excited states but is enough to populate the triplet excited states of the Ru(bpy) 3 2 + ∗ , which is a so-called energy-defi cient system. [ 21 ] On contrast, when the free energy through the elec-tron transfer is available to directly generate the singlet excited states, the system is called an energy-suffi cient system. [ 23 ] It should be noted that the radical ions are effective quenchers of the generated excited states through triplet–charge reaction in the electrochemical reaction. [ 24 ] Therefore, the triplet excited states of Ru(bpy) 3 2 + ∗ can be quenched through a triplet–charge reaction due to the long lifetime required to generate non-radi-ative emission instead of the radiative emission as described in Equation (2) .

Ru(bpy)2+∗3 + ion →Quenching

Ru(bpy)2+3 (2)

The cyclic voltammograms measurements indicate two separated oxidation peaks at different scan rates with the three-electrode confi guration, as shown in Figure 1 b, which confi rms that the required electrochemical reaction occurs in the electrogenerated chemiluminescence. The fi rst peak cor-responds to the direct oxidation of TPrA at the electrode at a potential of about 0.75 V vs the reference electrode: Ag/AgCl. The second peak has a potential of 1.15 V where Ru(bpy) 3 2 + is oxidized at the electrode at a scan rate of 10 mV s − 1 . As the scan rate increases, both the oxidation current and cyclic voltammo-gram peaks increase to higher values. It has been also found that the increase in scan rate can reduce the diffusion layer thickness and subsequently increases the electrical current. [ 25 ] It can be seen from the voltage–electroluminescence char-acteristics (Figure 1 c) that the electroluminescence intensity clearly increases and then decreases with increasing electrical potential voltage. This result implies that the generation of

mbH & Co. KGaA, Weinheim Adv. Mater. 2011, 23, 2216–2220

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATIO

N

Figure 2 . Experimental setup and results for magnetic fi eld effects with a two planar electrode confi guration for a Ru(bpy) 3 -based energy-defi cient system in a magnetic fi eld of 700 mT. The experimental errors for MFE EL and MC are 5% and 1%, respectively. The electrochemical electrolyte contains Ru(bpy) 3 2 + (1m M ), TPrA (0.1 M ), and phosphate buffer solution (0.2 M ) in water. a) Experimental setup with a two-electrode confi guration for an electrochemical cell placed in magnetic fi eld; b) MFE EL at different voltages; c) MC at different voltages.

Figure 1 . Electrogenerated chemiluminescence (ECL) characteristics for triplet Ru(bpy) 3 -based energy-defi cient system with a three-electrode con-fi guration measured at room temperature. The electrochemical electrolyte contains Ru(bpy) 3 2 + (1m M ), TPrA (0.1 M ), and phosphate buffer solution (0.2 M ) in water. a) ECL spectrum; b) Cyclic voltammograms collected at different scan rates; c) ECL intensity–voltage characteristics.

550 600 650 700 750

0

10

20

30

40a

EL

inte

nsity

(a.

u.)

Wavelength (nm)

0.0 0.5 1.0 1.5-0.3

-0.2

-0.1

0.0

•+→ TPrAe-TPrA ++ ?→?− 3

32

3 Ru(bpy)eRu(bpy)

b

100 mV/s50 mV/s

10 mV/s

Cur

rent

(m

A)

E (V vs Ag/AgCl)

0.0 0.4 0.8 1.2 1.60

4

8

12

16 cECL-V

EC

L in

tens

ity

E (V vs Ag/AgCl)

electroluminescence is a mass-transport-limited process in the electrochemical reaction. In a mass-transport-limited process, the reaction species are required to diffuse to the reaction inter-face around the positive electrode to produce the precursors for the generation of electrogenerated chemiluminescence. These precursors subsequently react in a spatially restricted emission zone within the diffusion layer near the electrode. The observed electroluminescence intensity is essentially determined by mass transport of reactive species at a given reaction rate to produce the light. Therefore, the electroluminescence intensity increases when mass transport can provide enough reaction species for light generation near the electrode. However, the electroluminescence intensity decreases when the mass trans-port is limited. With the two-electrode electrochemical confi gu-ration in an external magnetic fi eld ( Figure 2 a), the Ru(bpy) 3 -based electrochemical reaction generates a giant MFE EL at dif-ferent electrical biases. The MFE EL reaches 400% at 3.3 V in a magnetic fi eld of 700 mT (Figure 2 b), which is the largest MFE EL reported so far for any electroluminescent system. No signifi cant magnetic response appears below 100 mT. We can see that the sign and magnitude of the MFE EL depend on the applied potential bias. At a 2.2 V bias, the electroluminescence intensity is monotonically quenched by an external magnetic fi eld and no clear saturation was found at the higher magnetic

© 2011 WILEY-VCH Verlag GmAdv. Mater. 2011, 23, 2216–2220

fi eld. The magnitude of the MFE EL decreases to a negative value of –17% at an external magnetic fi eld of 700 mT. We should note that the two-electrode setup, specially designed to conven-iently measure magnetic fi eld effects, requires a turn-voltage of 1.9 V to initiate the electrochemiluminescence while the three-electrode setup needs a turn-on voltage of 0.8 V. However, this difference should not affect the mechanisms of magnetic fi eld effects in the electrochemical co-reaction.

Next, we discuss the origin of the observed MFE EL in electro-generated chemiluminescence. Early studies found that an external magnetic fi eld can increase the electroluminescence intensity in an electrochemical reaction [ 14 ] and this positive MFE EL was attributed to the magnetic-fi eld-sensitive triplet–charge reaction. [ 15 , 17 ] The spin physics in solid states indicates that an external magnetic fi eld can perturb the spin interaction between a triplet excited state and a charge and consequently reduce the triplet-charge reaction rate constant. [ 2 , 8 ] In the absence of a magnetic fi eld, the triplet excited states, Ru(bpy) 3 2 + ∗ , are partially quenched by the excess of radical ions through

2217bH & Co. KGaA, Weinheim wileyonlinelibrary.com

2218

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATI

ON

Figure 3 . Schematic for Lorentz force effects and angle dependence results for magnetic fi eld effects. a) Schematic for Lorentz force effects: liquid convection and ion penetration in an electrochemical cell placed in a magnetic fi eld (700 mT). b) Angle dependence of MFE EL and MC in a triplet Ru(bpy) 3 -based electrochemical system with Ru(bpy) 3 2 + (1 m M ), TPrA, (0.1 M ), and phosphate buffer solution (0.2 M ) in water. c) MFE EL at different TPrA molar concentrations for θ = 90 ° . Inset shows ECL intensity versus TPrA molar concentration.

Planar Pt electrode

Convection

Penetration

Diffusion layer

Planar Pt electrodea

B

0 30 60 90 120150180

100

200

300

400 θIB

MC

(%

)

MF

EE

L (%

)

Angle θ (degree)

b

4

8

12

16

20

24

0.0 0.1 0.2 0.3 0.4

0

100

200

300

400

500

0.0 0.1 0.2 0.3

5101520

EC

L

CTprA

(mM)

MF

EE

L (%

)

TPrA molar concentration (mM)

c

triplet–charge reactions. [ 2 , 8 , 24 ] With an applied magnetic fi eld, this quenching process is reduced by decreasing the reaction rate constant, which subsequently increases the triplet light emission in the electrochemical reaction. As a result, a positive MFE EL can be observed in triplet energy-defi cient electrochem-ical system based on the triplet–charge reaction. However, mag-netic-fi eld-sensitive triplet-charge reaction only contributes to a few tenths of a percent MFE EL in liquid states, as reported pre-viously. [ 13 , 15 ] In addition, it has been observed in solid states that an external magnetic fi eld can only change the triplet–charge reaction by a few percents, as indicated by the studies of magnetic fi eld effects of the photocurrent. [ 26 , 27 ] Clearly, the tri-plet–charge reaction is not suffi cient to generate the giant MFE EL observed from the Ru(bpy) 3 2 + system. Here, we suggest that Lorentz force effects are the dominant process responsible for the observed giant positive and negative MFE EL . It is known that the liquid solution fl ux containing charged species can experience the Lorentz force, which is expressed as the cross product of current and magnetic fi eld: F = I × B . This Lorentz force can result in a convection for the reactive species around the diffusion layer in the liquid solution through momentum transfer between reactive ions and solvent molecules, [ 28 , 29 ] as shown in Figure 3 a. As a consequence, the Lorentz force can generate two effects through convection: increasing ion pene-tration through diffusion layer and decreasing the diffusion layer thickness in the electrochemical reaction. Increasing ion penetration can enhance the electrochemical reaction and thus increase the electroluminescence intensity, leading to a positive MFE EL , referred to as transport-based positive MFE EL . Decreasing the diffusion layer thickness can reduce the entire electrochemical reaction volume. Since the light-emitting zone occurs within the diffusion layer, [ 21 ] the reduction of the diffu-sion layer thickness can decrease the electroluminescence intensity and essentially generate a negative MFE EL , referred to as volume-based negative MFE EL . At high voltage, the high den-sity of reactive species generates a thicker diffusion layer. With a thicker diffusion layer, the reduction in diffusion layer thick-ness due to Lorentz force effects can be limited compared to the entire diffusion layer thickness, minimizing the volume-based negative MFE EL . Also at high voltage, with higher ion concentration, the increase in ion penetration caused by the Lorentz force through mass transport can be more signifi cant (as suggested by the larger MC at high voltage), which leads to a dominant transport-based positive MFE EL . As a result, a high voltage can generate an overall positive MFE EL (Figure 2 b). At low voltage, the low density of reactive species produces a thinner diffusion layer. With a thinner diffusion layer, the reduction in diffusion layer thickness due to Lorentz force effects can be signifi cant relative to entire diffusion layer thick-ness, maximizing volume-based negative MFE EL . On the other hand, at low voltage with lower ion concentration the increase in ion penetration caused by Lorentz force trough mass trans-port is less signifi cant (as suggested by lower MC at low voltage), which minimizes transport-based positive MFE EL . Therefore, a low voltage can lead to an overall giant negative MFE EL (Figure 2 b). Furthermore, it should be noted that the mass transport driven by Lorentz force can generate MC in the electrochemical reaction. This occurs because when a magnetic fi eld is applied, the Lorentz force ( I × B ) exerted on the charged

© 2011 WILEY-VCH Verlag Gwileyonlinelibrary.com

reaction species yields a momentum transfer to the solvent molecules and enhances the charge transport and the electrical current in the electrochemical reaction. It should be further noted that changing the electrical potential can affect the den-sity of reactive ions and consequently changes the total mass transport based on Lorentz force effects. [ 28 , 29 ] This can lead to a modifi cation of the MC amplitude. At high voltage, high-den-sity reactive ions can more signifi cantly increase electrical cur-rent, as compared to low-density reactive ions at low voltage, due to Lorentz force driven mass transport. It can be clearly seen in Figure 2 c that the MC reaches 16% at 3.3 V and 5% at 2.2 V in a magnetic fi eld of 700 mT. This voltage dependence of the MC further suggests that the Lorentz force effects are mainly responsible for the observed magnetic fi eld effects. To verify the Lorentz force effects in the observed MFE EL , we inves-tigated the angle dependence of MFE EL in the electrochemical reaction with two planar electrodes by changing the angle from θ = 0 ° , when I and B are parallel, to θ = 180 ° , when I and B are antiparallel. Figure 3 b shows a signifi cant angle dependence of MFE EL when the current direction is changed relevant to the

mbH & Co. KGaA, Weinheim Adv. Mater. 2011, 23, 2216–2220

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATIO

N

orientation of the applied magnetic fi eld. Clearly, the maximum positive MFE EL is observed at θ = 90 ° . The maximum MC is also shown at θ = 90 ° . In general, the angle dependence of mag-netic fi eld effects can be attributed to Lorentz force effects. It is clear that the Lorentz force can largely change its value at dif-ferent angles ( θ ) and therefore affects the ion transport in the generation of electroluminescence through convection in the electrochemical reaction cell (Figure 3 a). As a result, an applied magnetic fi eld can cause different responses in electrolumines-cence and electrical current as the angle θ changes. This phe-nomenon has been observed in the magnetic fi eld dependence of the electrical current in electrochemical reactions, as reported in early publications. [ 11 , 12 ] To further confi rm the Lorentz effects, we studied the effects of the concentration of reactive species on the MFE EL in the electrochemical co-reaction. Figure 3 c shows the MFE EL from triplet Ru(bpy) 2 + ∗ emission as a function of co-reactant TPrA concentration from 0.01 m M to 0.3 m M . We can see that the MFE EL largely increases with increasing the co-reactant TPrA concentration. In addition, increasing TPrA concentration can also enhance the electrogen-erated chemiluminescence intensity (inset in Figure 3 c). These concentration results indicate that the Lorentz force can gen-erate larger mass transport through convection and conse-quently enhances the MFE EL as the reactive mass increases. As a result, the experimental results from voltage, angle, and con-centration dependences indicate that the Lorentz force effects can generate giant MFE EL in liquid states based on the electro-chemical reaction. In addition, we should note that, if the applied magnetic fi eld is considerably non-uniform within an electrochemical cell, the magnetic fi eld gradient associated with this non-uniformity can generate magnetic fi eld effects on electrochemical reaction and consequently change the electro-chemiluminescence intensity. We have examined this issue by manually changing the non-uniformity by adjusting the dis-tances (2 mm and 5 mm) between the two platinum electrodes in the setup (two magnetic poles with diameters of 65 mm and distances of 20 mm; two platinum electrodes with sizes of 7 mm × 10 mm and adjustable distances from 2 mm to 5 mm). We observed that changing the distance between two platinum electrodes does not appreciably change the angle dependence of the magnetic fi eld effects. This means that the non-uniformity of magnetic fi eld does not have considerable contribution to the observed magnetic fi eld effects.

The remaining MFE EL at θ = 90 ° and 180 ° is discussed next. It is noted that considerable MFE EL remains when I and B are parallel or antiparallel (Figure 3 b). In principle, this remaining MFE EL can be due to two different possibilities: magnetic body force due to magnetization of paramagnetic ions or triplet–charge reaction due to interaction between triplet excited states and radicals. First, an applied magnetic fi eld can magnetize paramagnetic radicals [ 30 ] and generate magnetic body force at θ = 90 ° and 180 ° . This magnetic body force can contribute to mass transport through momentum transfer between solvent molecules and paramagnetic radicals leading to a remaining MFE EL at θ = 90 ° and 180 ° . In particular, this magnetic body force can push paramagnetic radicals away and toward the diffu-sion layer when the I and B are parallel and antiparallel, respec-tively, generating a smaller or larger remaining MFE EL at θ = 90 ° and 180 ° , as supported by the experimental results shown

© 2011 WILEY-VCH Verlag GAdv. Mater. 2011, 23, 2216–2220

in Figure 3 b. Second, triplet excited states can react with radi-cals to produce triplet–charge reaction with the consequence of quenching light emission from triplet excited states. [ 2 , 24 ] This triplet–charge reaction can lead to a positive MFE EL upon tri-plet emission at θ = 90 ° and 180 ° when the applied magnetic fi eld reduces the triplet–charge reaction. Early studies have sug-gested that the triplet–charge reaction can generate a positive MFE EL with an amplitude of less than 30%. [ 13 ] As a result, it can be suggested that the magnetic body force-based mass transport and triplet–charge reaction can generate the remaining MFE EL at θ = 90 ° and 180 ° . It should be further noted that magnetic-force-based mass transport and triplet–charge reaction can also generate the remaining MC at θ = 90 ° and 180 ° . This is because magnetic-force-driven mass transport and triplet-charge reac-tion can increase ion transport within the diffusion layer and generate a positive MC. Nevertheless, our experimental studies indicate that the Lorentz force effects function as the main mechanism to generate giant positive and negative MFE EL in electrochemical reactions. The magnetic body force and triplet–charge reaction play a secondary role in the generation of the giant MFE EL .

In summary, we experimentally demonstrate giant magnetic fi eld effects ( > 400%) in electrogenerated chemiluminescence based on a co-reaction. The angle, voltage, and concentration dependence of the magnetic fi eld effect suggest that the Lorentz force driven ion transport and the Lorentz force dependent dif-fusion layer thickness through liquid convection are mainly accountable for the observed giant MFE EL . In addition, we fi nd that the magnetic body force due to magnetization of paramag-netic radicals and the triplet–charge reaction due to interaction between triplet excited states and radicals can also contribute to giant magnetic fi eld effects as a secondary mechanism. Furthermore, our experimental results indicate that the MFE EL observed at different angles, concentrations, and voltages can be used to elucidate magnetic-fi eld-dependent mass trans-port, magnetization of paramagnetic radicals, and magnetic-fi eld-dependent triplet–charge reaction in electrochemical reactions. Moreover, rationally adjusting Lorentz force effects presents a new pathway to develop giant magnetic fi eld effects in liquid states based on electrochemical reactions.

Acknowledgements This research was supported by the National Science Foundation (ECCS-0644945) and Center for Materials Processing at the University of Tennessee. This research was partially conducted at the Center for Nanophase Materials Sciences based on user project (CNMS2009–055), which is sponsored at Oak Ridge National Laboratory by the Division of Scientifi c User Facilities, U.S. Department of Energy.

Received: January 17, 2011 Revised: February 22, 2011

Published online: April 4, 2011

[ 1 ] J. Kalinowski , M. Cocchi , D. Virgili , P. Di Marco , V. Fattori , Chem. Phys. Lett. 2003 , 380 , 710 .

[ 2 ] V. Ern , R. E. Merrifi eld , Phys. Rev. Lett. 1968 , 21 , 609 . [ 3 ] E. L. Frankevich , A. A. Lymarev , I. Sokolik , F. E. Karasz ,

S. Blumstengel , R. H. Baughman , H. H. Hörhold , Phys. Rev. B. 1992 , 46 , 9320 .

2219mbH & Co. KGaA, Weinheim wileyonlinelibrary.com

2220

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATI

ON

[ 4 ] T. L. Francis , Ö. Mermer , G. Veeraraghavan , M. Wohlgenannt , New

J. Phys. 2004 6 , 185 . [ 5 ] B. Hu , L. A. Yan , M. Shao , Adv. Mater. 2009 , 21 , 1500 . [ 6 ] P. A. Bobbert , T. D. Nguyen , F. W. A. van Oost , B. Koopmans ,

M. Wohlgenannt , Phys. Rev. Lett. 2007 , 99 , 216801 . [ 7 ] R. C. Johnson , R. E. Merrifi eld , P. Avakian , R. B. Flippen , Phys. Rev.

Lett. 1967 , 19 , 285 . [ 8 ] M. Wittmer , I. Zschokkegranacher , J. Chem. Phys. 1975 , 63 , 4187 . [ 9 ] E. Frankevich , A. Zakhidov , K. Yoshino , Y. Maruyama , K. Yakushi ,

Phys. Rev. B 1996 , 53 , 4498 . [ 10 ] Y. Wu , Z. Xu , B. Hu , J. Howe , Phys. Rev. B 2007 , 75 , 035214 . [ 11 ] J. Lee , X. Gao , L. D. A. Hardy , H. S. White , J. Electrochem. Soc. 1995 ,

142 , L90 . [ 12 ] S. R. Ragsdale , J. Lee , X. Gao , H. S. White , J. Phys. Chem. 1996 , 100 ,

5913 . [ 13 ] L. R. Faulkner , H. Tachikawa , A. J. Bard , J. Am. Chem. Soc. 1972 , 94 ,

691 . [ 14 ] L. R. Faulkner , A. J. Bard , J. Am. Chem. Soc. 1969 , 91 , 209 . [ 15 ] L. R. Faulkner , A. J. Bard , J. Am. Chem. Soc. 1969 , 91 , 6495 . [ 16 ] H. Tachikawa , A. J. Bard , Chem. Phys. Lett. 1974 , 26 , 10 .

© 2011 WILEY-VCH Verlag Gwileyonlinelibrary.com

[ 17 ] H. Tachikawa , A. J. Bard , Chem. Phys. Lett. 1974 , 26 , 246 . [ 18 ] J. K. Leland , M. J. Powell , J. Electrochem. Soc. 1990 , 137 , 3127 . [ 19 ] Rubinstein , A. J. Bard , J. Am. Chem. Soc. 1981 , 103 , 512 . [ 20 ] B. A. Gorman , P. S. Francis , N. W. Barnett , Analyst 2006 , 131 , 616 . [ 21 ] J. Bard , Electrogenerated chemiluminescence , Marcel Dekker , Inc .

New York 2004 . [ 22 ] F. Kanoufi , Y. Zu , A. J. Bard , J. Phys. Chem. B 2001 , 105 , 210 . [ 23 ] R. Y. Lai , A. J. Bard , J. Phys. Chem. A 2003 , 107 , 3335 . [ 24 ] G. J. Hoytink , Discuss. Faraday Soc. 1968 , 45 , 14 . [ 25 ] M. A. Prasad , M. V. Sangaranarayanna , Electrochim. Acta 2004 , 49 ,

445 . [ 26 ] Z. Xu , B. Hu , Adv. Funct. Mater. 2008 , 18 , 2611 . [ 27 ] V. Tolstov , A. V. Belov , M. G. Kaplunov , I. K. Yakuschenko ,

N. G. Spitsina , M. M. Triebel , E. L. Frankevich , J. Lumin. 2005 , 112 , 368 . [ 28 ] O. Aaboubi , J. P. Chopart , J. Douglade , A. Olivier , C. Gabrielli , B.

Tribollet , J. Electrochem. Soc. 1990 , 137 , 1796 . [ 29 ] O. Lioubashevski , E. Katz , I. Willner , J. Phys. Chem. B 2004 , 108 ,

5778 . [ 30 ] O. Lioubashevshki , E. Katz , I. Willner , J. Phys. Chem. C 2007 , 111 ,

6024 .

mbH & Co. KGaA, Weinheim Adv. Mater. 2011, 23, 2216–2220