当前位置：文档库 › 69 Electroless platinum counter electrodes with Pt-activated self-assembled monolayer

69 Electroless platinum counter electrodes with Pt-activated self-assembled monolayer

Electroless platinum counter electrodes with Pt-activated self-assembled monolayer on transparent conducting oxide

Jeng-Yu Lin ?,Yi-Ting Lin

Department of Chemical Engineering,Tatung University,Taipei 104,Taiwan

a b s t r a c t

a r t i c l e i n f o Article history:

Received 8March 2012

Accepted in revised form 12May 2012Available online 19May 2012Keywords:

Dye-sensitized solar cell Counter electrode Electroless deposition

Self-assembled monolayer

In this study,electroless Pt ?lms deposited on transparent conducting oxide (TCO)-coated substrates were prepared using a Pt-activated self-assembled monolayer (SAM)and employed as counter electrodes (CEs)in dye-sensitized solar cells (DSSCs).First,the mechanism for the formation of Pt-activated SAM on TCO-coated substrate was examined using X-ray photoelectron spectroscopy analyses.Second,the effects of the pH value and bath composition on the surface morphology and electrocatalytic activity were also investigated.After the optimization with the pH and composition of deposition bath,the Pt CE prepared by the optimized bath composition of electroless deposition (ELD)was found to have a lower Pt loading (15.91μg cm ?2),a higher speci ?c active surface area (209.71m 2g ?1)and lower charge-transfer resistance (1.92Ω·cm 2)than the sputtered-Pt CE.Thus,the cell ef ?ciency of the DSSC based on the ELD-Pt CE achieved 7.65%,which represented an increase of 12.33%when compared to the sputtered-Pt CE,while the Pt loading was also remarkably reduced (being 14.82%of that of the sputtered-Pt CE).

1.Introduction

In recent years,dye-sensitized solar cell (DSSC)developed by Gr?tzel et al.[1]has aroused intensive interests as a low-cost alterna-tive to conventional inorganic p –n junction photovoltaic devices.One component of DSSCs,the counter electrode (CE)plays an important and indispensable role in collecting electrons from external circuit

and reducing I 3?

to I ?for the regeneration of sensitizer after electron injection.Platinum (Pt),carbonaceous materials [2–5],conducting polymers [6–8],and cobalt sul ?des [9–12]have all been proposed to serve as electrocatalysts on CEs to speed up the reduction reaction

of I 3?/I ?.Pt is widely used as an electrocatalytic material due to its superior electrocatalytic activity and stability (compared with other electrocatalytic materials)when in contact with iodine-based electrolytes.

In practice,sputtering is commonly developed for depositing Pt thin ?lm due to its reproducibility.However,sputtering requires an ultrahigh vacuum environment and high power due to the melting point of Pt.Moreover,a large portion of such deposited Pt does not fully exhibit its electrocatalytic function as it is located inside the ?lm,thus rendering a waste of materials [13].Thermal decomposition has been proposed as another approach for fabricating CEs with low Pt loading and has demonstrated excellent electrocatalytic performance and mechanical stability [14].However,the disadvantage of thermal

decomposition is that the conversion from the precursor solution in-cluding H 2PtCl 6to metallic Pt takes place at a temperature at least 380°C,thus excluding the choice of ?exible plastic substrates [15].Elec-trochemical methods have been considered as alternative approaches for Pt deposition,as these omit the requirement for vacuum equipment and high-temperature annealing treatment [14,16–24].

Electrochemical deposition can be classi ?ed into two main methods:electrodeposition and electroless deposition (ELD).Unlike electrodeposition,ELD can be conducted without an external poten-tial,and can also easily be scaled-up for the purposes of industrial development.Chen et al.[21]prepared Pt CEs through displacement reduction on metal sheets.The DSSC assembled with the Pt CE demonstrated high cell conversion ef ?ciency of up to 7.29%.Unfortu-nately,the usage of Pt loading was also as high as ~487.5μg cm ?2.Moreover,the iodine-containing electrolyte could also corrode the metal substrates,thus posing problems in terms of reliability issue.Chen et al.[22]also prepared Pt CEs by ELD of Pt thin ?lms on indium-doped tin oxide (ITO)-coated glass substrates.They used a Pd catalyst on an ITO-coated glass substrate to initiate Pt ELD and toxic hydrazine as a reductant in the deposition bath.The ef ?ciency of the DSSC assembled with the ELD-Pt CE reached 6.46%.Neverthe-less,the loading of such Pt CE was still as high as ~487.5μg cm ?2.Previously,we developed a simpli ?ed process for fabricating ELD-Pt on a ?uorine-doped tin oxide (FTO)-coated glass substrate by grafting 3-(2-aminoethylamino)propylmethyldimethoxysilane (Me-EDA-Si)self-assembled monolayers (SAMs).The DSSC with such ELD-Pt CE can achieve a cell ef ?ciency of 6.71%with a much lower Pt loading of 34.6μg cm ?2[23].However,at that time,Pd with low electrocatalytic

Surface &Coatings Technology 206(2012)4672–4678

?Corresponding author.Tel.:+886225925252x2561119;fax:+886225861939.E-mail address:jylin@https://www.wendangku.net/doc/0d10435696.html,.tw (J.-Y.

10.1016/j.surfcoat.2012.05.044

Contents lists available at SciVerse ScienceDirect

Surface &Coatings Technology

j o u r n a l h o m e p a ge :w w w.e l s e v i e r.c o m/l o c a t e /s u r f c o a t

activity for I3?/I?reaction was still used as the catalyst to initiate the Pt ELD process.As can be seen in Fig.1,the mean Pt loading on the FTO-coated glass substrate increased with increasing the deposition time for both Pd-activated and Pt-activated ELD processes.Moreover,the mean Pt loading for Pd-activated process was much higher compared to that for Pt-activated process at the same deposition time.The mean Pt loading for Pd-activated process can reach37.4μg cm?2within the deposition time of only1min.Thus this high Pt deposition rate could limit the further reduction of Pt loading.Furthermore,the mechanism for ELD-Pt by means of the SAM modi?cation was not well understood and,as such,the bath formulation was not optimized.

In this study,Pt-activated ELD process was?rst utilized to replace Pd-activated process to further reduce the Pt loading.Next,the de-tailed mechanism for Pt CEs prepared by means of SAM modi?cation and ELD was investigated.Additionally,the effects of bath formula-tion on the surface morphology and electrocatalytic activity of the Pt deposits and the photovoltaic performance of the corresponding DSSCs were also examined.

2.Experimental

2.1.Preparation of Pt CEs

Prior to ELD,the transparent conducting oxide(TCO)-coated sub-strates,FTO-coated glass substrates(13Ω/□,Tripod Technology Co.), were ultrasonically cleaned sequentially in detergent,deionized water,and isopropyl alcohol(IPA,ECHO Chemicals)for10min,re-spectively(cleaning step).A SAM of Me-EDA-Si(Aldrich)was formed on the TCO-coated substrate by immersion in1vol.%Me-EDA-Si aqueous solution at ambient atmosphere for10min(coupling step) and then subjected to be rinsed with deionized water to remove excess silane molecules.Activation of the TCO-coated substrate covered with SAM was carried out by immersion in a HCl-acidic solu-tion containing0.01M H2PtCl6·6H2O(Merck)at ambient atmosphere for10min to immobilize[PtCl6]2?ions(catalyzation step).After immobilization,TCO-coated substrates were rinsed with deionized water and immediately immersed into the solution containing0.05M NaBH4for30min to reduce[PtCl6]2?ions to Pt catalysts.Then the Pt ELD was performed in the deposition bath consisting of0.3–1.0mM H2PtCl6·6H2O(Merck),0.3–0.75M C6H5Na3O7·2H2O(Aldrich),and 0.3M CHNaO2(Showa)at70°C for6min.The pH value of deposition bath was adjusted by using diluted NH4OH or HCl if necessary.For comparison,~100nm thick of Pt layer was sputtered on a FTO-coated glass substrate using a DC sputtering instrument(ULVAC)at a deposi-tion rate of0.28nm s?1and employed as CEs for DSSCs.2.2.Cell assembly

A bi-layer TiO2?lm composed of6μm transparent nanocrystalline TiO2?lm and6μm scattering TiO2?lm was used as a photoanode.The dense nanocrystalline TiO2?lm loaded on the cleaned FTO glass was made from a commercial TiO2paste(CCIC,18nm)by using a semi-automatic screen printer(ATMA,AT45PA),followed by gradually sintering under an air?ow at150°C for10min,325°C for5min, 375°C for5min,450°C for15min.The screen-printable paste containing mesoporous TiO2beads was further printed on the top of nanocrystalline TiO2electrode as the scattering layer.The meso-porous TiO2beads were synthesized using a combined sol–gel and solvothermal process[25,26].Firstly,amorphous precursor TiO2 beads were made by adding titanium(IV)isopropoxide(97%,Sigma-Aldrich)to an ethanol solution containing hexadecylamine(90%, Sigma-Aldrich)and0.1M KCl(Acros)aqueous solution under vigor-ous stirring at ambient atmosphere.The air-dried precursor beads were further dispersed in an ethanol water mixture(2:1by volume) containing1.0mL25%ammonia aqueous solution.Then the resulting mixture was sealed within a Te?on-lined autoclave and heated at 160°C for16h.The mesoporous TiO2beads can be obtained by?ltra-tion,washed with ethanol,and dried at ambient atmosphere.After the construction of the scattering layer,the as-prepared TiO2photoanodes were sintered at500°C for30min.After being cooled to80°C,the sintered TiO2photoanodes were immersed into an ethanol solution containing0.3mM N719dye(RuL2(SCN)2·2H2O,L=2,2′-bipyridyl-4,4′-dicarboxylic acid,Everlight Chemical Industry Co.)and kept at 40°C for4h.The dye-sensitized TiO2photoanode was assembled with a CE into a sandwich type cell and sealed with a thermoplastic hot-melt surlyn(30μm thick,Dupont).

2.3.Characterizations and measurements

The mechanism of surface modi?cation and Pt activation were further studied by an X-ray photoelectron spectroscopy(XPS)with PHI Quantera SXM(ULVAC-PHI).A?eld-effect scanning electron mi-croscopy(FESEM;JSM-7000F)was employed to study the morphol-ogies of Pt CEs.The mean Pt loadings were determined by dissolution of the same six Pt electrodes in20mL aqua regia with agitation for one week,and Pt content in the subsequent diluted solution was exam-ined by an inductively couple plasma-optical emission spectrometer (ICP-OES,PerKin Elmer Optima-2000DV).Both cyclic voltammetry (CV)and electrochemical impedance spectroscopy(EIS)were carried out using a computer-controlled electrochemical analyzer,CHI6081D (CH Instrument).The CVs for I3?/I?were conducted using a Pt wire

Fig.1.The change in Pt loading on FTO glass substrates with time for Pd-activated and

Pt-activated ELD

processes.

Fig.2.XPS N1s spectra of SAM modi?ed substrate(a)before coupling step,(b)before

catalyzation step,and(c)after catalyzation step.

4673

J.-Y.Lin,Y.-T.Lin/Surface&Coatings Technology206(2012)4672–4678

reference electrode,a 4cm 2Pt sheet as auxiliary electrode,and the as-prepared ELD-Pt or sputtered Pt CE as the working electrode with a total exposure area of 0.64cm 2in a 3-methoxypropionitrile solution con-sisting of 50mM LiI,10mM I 2,and 500mM LiClO 4.CVs were performed in the potential interval ranging from ?0.4V to 0.4V vs.Pt at a scan rate of 10mV s ?1.The CVs used to examine the active surface area of Pt CEs were made in a 0.5M H 2SO 4aqueous solution at a scan rate of 10mV s ?1.The electrocatalysis of various Pt CEs were further studied by EIS analysis on symmetric dummy cells at zero bias potential and the impedance data covered a frequency range of 0.1Hz –100kHz.A si-nusoidal AC voltage signal varying by 5mV was employed in all cases.The complex nonlinear least square (CNLS)analyses of the resulting im-pedance spectra were conducted by means of the software,ZSimpWin version 3.1.For photovoltaic testing,a solar simulator (Yamashita Denso,YSS-150A)served as a light source and its light intensity was calibrated to 100mW cm ?2(AM 1.5)using a reference Si solar cell.A

computer-controlled digital source meter (Keithely,model 2400)was used to record the photocurrent –voltage curves.The redox electrolyte consisting of 1M 1.3-dimethylimidazoliumiodine (Merck),0.5M 4-tert-butylpyridine (Aldrich),0.15M iodine (J.T.Baker),and 0.1M guanidine thiocyanate (Aldrich)in 3-methoxypropionitrile (Acros)solvent was employed for EIS and photovoltaic measurements.3.Results and discussion

3.1.Characterization of Pt 4+-activated SAMs

To characterize the Pt-activated SAMs on FTO substrates,XPS analy-ses for the different steps were carried out in which high-resolution XPS spectra of N 1s and Pt 4f peaks were recorded.Curve (a)in Fig.2presents the absence of the amino functional group (\NH 2)corresponding to Me-EDA-Si,as the FTO substrates did not proceed the coupling step.However,a peak located at a binding energy of 399.3eV was observed in curve (b)of Fig.2,indicating the existence of \NH 2[27,28].This also reveals that SAMs successfully formed on the FTO surface,which is in accordance with our previous observa-tion by means of a ?uorescent microscope [23].Curve (c)in Fig.2shows that a signi ?cant reduction of binding energy representing N 1s was observed for the Pt 4+-activated FTO surface,which may possibly have resulted from the binding of \NH 2with Pt 4+[29].Fig.3presents the Pt 4f XPS spectra with different treatments.Curves (a)and (b)in Fig.3illustrate the XPS spectra of the FTO surfaces without the catalyzation process.As expected,no Pt 4+signals were recorded on the un-catalyzed surface.After the catalyzation step,the characteristic peaks with binding energies of 71.9eV and 75.2eV corresponding to Pt 4f 7/2and Pt 4f 5/2,respectively,were observed.This suggests the presence of Pt 4+on the SAM-coated FTO substrate,which originates from the binding of \NH 2of SAMs with Pt 4+and which is in accor-dance with the binding energy shift of N 1s observed in curve (c)in Fig.3.Consequently,the mechanism of coupling and catalyzation steps can be suggested as follows:the ultrasonic cleaning of FTO substrates was made to generate the hydroxyl group (\OH)on the FTO surface,and thus the silane compounds were able to anchor \OH with a Si \O \Si bond to form SAMs and the \NH 2of SAMs

were

Fig.3.XPS Pt 4f spectra of SAM modi ?ed substrate (a)before coupling step,(b)before catalyzation step,and (c)after catalyzation

step.

Fig.4.Top-view FESEM images of (a)sputter Pt,and ELD-Pt deposits prepared at (b)pH 4,(c)pH 5,and (d)pH 6.The deposition bath which was composed of 1.0mM H 2PtCl 6·6H 2O,0.5M C 6H 5Na 3O 7·2H 2O,and 0.3M CHNaO 2was employed for the measurements.

4674J.-Y.Lin,Y.-T.Lin /Surface &Coatings Technology 206(2012)4672–4678

then able to bind further with the PtCl62?ion through electrostatic interaction.

3.2.pH effect on the characteristics of Pt CEs

Fig.4a presents the top-view FE-SEM image of a sputtered-Pt CE. The surface morphology naturally resembles that of pure FTO sub-strate,which is in accordance with our previous observations [12,20].Fig.4b–d shows the surface evolution of ELD-Pt CEs prepared at pH values ranging from4to6.When the pH of the deposition bath was controlled at4,many gray-white Pt nanoclusters were seen to form inhomogeneously and stack vertically on the FTO surface (Fig.4b).In contrast,when the pH value was increased to5,Pt nanoparticles further decreased in size and were distributed unevenly on the FTO substrate.However,parts of the surface of the FTO sub-strate were still exposed after deposition,suggesting low deposition rate at pH5.As the pH was increased to6,the Pt particles grew much larger than those at pH4and pH5due to the partial aggregation of Pt nanoparticles(Fig.4d).The FTO surface was also almost completely covered,thus indicating a signi?cant increase in the Pt deposition rate at pH6.The mean Pt loading of various Pt CEs was examined by an ICP-OES,as is shown in Table1.As can be seen from Table1,the mean Pt loading of ELD-Pt CEs was much lower than that of the sputtered-Pt CE.Of the ELD-Pt CEs,the CE prepared at pH 6had the highest Pt loading of36.7μg cm?2,while the CE prepared at pH5possessed the lowest Pt loading of~11.3μg cm?2,which is consistent with the SEM results.

The regeneration of redox species I3?/I?is carried out through the reduction of I3?on the CEs of DSSCs.To investigate the electrocatalytic activity of CEs toward I3?reduction,EIS analyses were conducted with symmetric test cells.Fig.5represents the Nyquist spectra of the sputtered-Pt and ELD-Pt CEs.The semicircle at the high-frequency range corresponds to charge-transfer resistance(R ct),representing the electron-transfer resistance between the electrolyte and the CE. The impedance parameters(obtained by?tting the arc observed at higher frequencies in Nyquist plots(leftmost semicircle)to the equiv-alent circuit model shown in the inset of Fig.5)are also summarized in Table1.In general,the lower R ct value indicates that electron trans-ferability is more facile from the CE to I3?,thus demonstrating a higher electrocatalytic activity for I3?reduction.In comparison with the sputtered-Pt CE,ELD-Pt CEs demonstrated impressive electrocatalytic activities with a much lower Pt loading.The ELD-Pt CE prepared at pH 4possessed the lowest R ct at1.92Ω·cm2,which was lower than the 2.59Ω·cm2for the sputtered-Pt CE.Moreover,CV tests were con-ducted to con?rm the?ndings from the EIS analyses.According to

Table1

The characteristics and performance of CEs employed in DSSCs.

Reductant conc. (M)Pt conc.

(mM)

Pt loading

(μg cm?2)

Speci?c active area

(m2g?1)

R ct

(Ω·cm2)

J sc

(mA cm?2)

V oc

(mV)

FFη

(%)

Sputter Pt––107.310.78 2.5913.34±0.25787±50.65±0.01 6.81±0.16 ELD-Pt(pH4)0.30 1.023.4– 2.3914.01±0.24785±60.65±0.017.26±0.14

0.500.3 3.2–15.5613.32±0.33790±30.60±0.01 6.37±0.22

0.500.5 6.4–7.5713.41±0.50788±60.63±0.01 6.73±0.17

0.50 1.015.9209.71 1.9215.07±0.38784±50.66±0.017.65±0.15

0.75 1.0 4.7–17.7413.12±0.44789±40.60±0.02 6.26±0.25 ELD-Pt(pH5)0.50 1.011.3196.97 2.4313.85±0.45789±30.64±0.017.03±0.36

ELD-Pt(pH6)0.50 1.036.745.32 2.1614.72±0.11778±60.65±0.017.42±

0.25

Fig.5.Nyquist plots of the symmetric cells assembled with sputtered-Pt CE and ELD-Pt CEs at different pH values.The inset shows the equivalent circuit model for

simulation.Fig.6.(a)CVs of sputtered-Pt CE and ELD-Pt CEs prepared at different pH values.The measurements were carried out in a3-methoxypropionitrile solution consisting of 50mM LiI,10mM I2,and500mM LiClO4.(b)CVs of sputtered-Pt CE and ELD-Pt CEs prepared at different pH values in0.5M H2SO4aqueous solution.

4675

J.-Y.Lin,Y.-T.Lin/Surface&Coatings Technology206(2012)4672–4678

previous reports [7,11],one pair of oxidation and reduction peaks for

the I 3?/I ?

redox mediator occurred between ?0.4V and 0.4V vs.Pt,which is contributed by reaction (1)and reaction (2),respectively.3I ?

→I 3?

t2e

anodic peak eTe1TI 3

?t2e ?→3I

cathodic peak eT

e2T

Fig.6a shows cyclic voltammograms for the sputtered-Pt CE and various ELD-Pt CEs.The ELD-Pt CEs show one pair of oxidation and reduction peaks similar to that for the sputtered-Pt CE,suggesting that the obtained ELD-Pt CEs demonstrate similar electrocatalytic

activity to that of the sputtered-Pt CE for I 3?

reduction.Nevertheless,the ELD-Pt CE at pH 4demonstrated the largest anodic and cathodic current densities among all Pt CEs,indicating the largest active area

for I 3?

reduction.Additionally,the cathodic peak potential of the ELD-Pt CE prepared at pH 5was more negative than those of other

CEs,indicating that the overpotential for I 3?

reduction of the ELD-Pt CE was much larger than that for other CEs.These results are in con-sistence with the observations from the EIS analyses that showed that the ELD-Pt CE prepared at pH 5had the largest R ct of 2.43Ω·cm 2.Fig.6b shows the cyclic voltammograms of the sputtered-Pt CE and ELD-Pt CEs prepared at different pH values,which were measured in a 0.5M H 2SO 4aqueous solution to elucidate the electrochemically ac-tive areas of the CEs.The electrochemically active area was evaluated from the area of H 2desorption on Pt by considering the contribution of the double-layer charge [30].It can be noted that the ELD-Pt CE prepared at pH 4had the largest total active area (the Pt loading mul-tiplies by the speci ?c active area)of all the CEs,with the calculated total active area for the ELD-Pt at pH 4of 33.34cm 2,a value ~3times higher than that for the sputtered-Pt CE,as shown in Table 1.These

results indicate that the superior electrocatalytic activity toward I 3?

reduction of the ELD-Pt CE at pH 4can be attributed to its high active surface area (as the electrocatalytic activity depends on the active surface area of the catalyst),which is a good supplement to Fig.6a [6,14,31].As further considering the effective Pt usage,the speci ?c active areas for the Pt CEs were evaluated.The ELD-Pt CE at pH 4had the largest speci ?c active surface area of all the CEs,with the calculated speci ?c active area for the ELD-Pt at pH 4of 209.71m 2g ?1,a value ~20times higher than that for the sputtered-Pt CE.

The photocurrent-voltage (J –V )curves of DSSCs assembled with the sputtered-Pt CE and ELD-Pt CEs prepared at different pH values are shown in Fig.7.The resultant photovoltaic parameters are also summarized in Table 1.When comparing the DSSCs assembled with ELD-Pt CEs prepared at different pH values,the change in pH value of bath deposition had no signi ?cant in ?uence on the open circuit voltage (V oc ).Nevertheless,the short circuit current (J sc ),?ll factor (FF ),and cell ef ?ciency (η)values were found to decrease initially and then increase as the pH value of deposition bath increased.CEs with the larger active surface areas can increase the total current of I 3?/I ?redox reaction [6,14,31].Thus the change in J sc can be

ascribed

Fig.7.Photocurrent density –voltage characteristics of DSSCs with sputtered-Pt CE and

ELD-Pt CEs prepared at different pH

values.

Fig.8.Top-view FESEM images of ELD-Pt deposits with (a)different concentrations of C 6H 5Na 3O 7·2H 2O and (b)different concentrations of H 2PtCl 6·6H 2O.The pH value of depo-sition bath was controlled at 4and the deposition time was set for 5min.

4676J.-Y.Lin,Y.-T.Lin /Surface &Coatings Technology 206(2012)4672–4678

to the increased electrocatalytic active surface area,as observed in Fig.6.Moreover,the lower R ct at CE/electrolyte interface can facilitate the electron transport from the CE to I3?ions,thus enhancing the FF for DSSCs[18].Thus the DSSC using the ELD-Pt prepared at pH4 as the CE had the best photovoltaic performance with a J sc of 15.07mA cm?2,a V oc of784mV,and a FF of0.66,thus yielding aηof7.65%.As for the DSSC assembled with the sputtered-Pt CE,the J sc,V oc and FF were13.34mA cm?2,787mV,and0.65,respectively with a resultantηof6.81%.

Although the highest Pt loading for the ELD-Pt CE was prepared at pH6,cell ef?ciency reached a maximum when the pH of the deposi-tion bath was controlled at4.Theηvalue decreased dramatically with a higher Pt loading on the FTO surface,indicating that most of the Pt became useless due to Pt aggregation.Consequently,the ELD-Pt CE with high electrocatalytic activity and low Pt loading can be obtained when the pH value of deposition bath is set at pH4.

3.3.Effect of bath composition on the characteristics of Pt CEs

Fig.8a shows the surface evolution of ELD-Pt CEs prepared at pH4 with different concentrations of C6H5Na3O7·2H2O acting as the re-ductant in the deposition bath.It can be observed that the decrease in reductant concentration to0.3M resulted in lots of Pt particles forming on the FTO surface,in which a fast deposition rate was demon-strated.In contrast,few Pt particles formed on the FTO surface when the reductant concentration increased to0.75M.Fig.9a displays cyclic voltammograms for sputtered-Pt CE and ELD-Pt CEs prepared at pH4 with different reductant concentrations.The single pair of oxidation and reduction peaks within the potential interval from?0.4V to 0.4V vs.Pt could still be found even when the reductant concentration decreased from0.5M to0.3M.In addition,the ELD-Pt CEs prepared with the reductant concentrations of0.3and0.5M demonstrated anodic and cathodic peaks of a similar shape and size,suggesting the obtained ELD-Pt CEs have comparatively similar electrocatalytic activity and active area for I3?reduction.Interestingly,a poor electrocatalytic activity toward I3?reduction was found for the ELD-Pt prepared at the reductant concentration of0.75M.The lack of Pt particles may be attributed to such poor electrocatalytic activity for I3?reduction. Fig.8b shows the surface evolution of ELD-Pt CEs prepared at pH4 with different Pt concentrations.As expected,the amount of Pt nanoparticles forming on the FTO substrate became less as the Pt concentration decreased.Fig.9b displays cyclic voltammograms for sputtered-Pt CE and ELD-Pt CEs prepared at pH4with different Pt concentrations.Although the obtained oxidation and reduction peaks of I3?/I?for the different Pt CEs were similar,decreased cathodic current density was observed when the Pt concentration was less than1mM. This suggests that the active area of ELD-Pt CE decreased while the Pt concentration decreased.It should be noted that the cathodic current density of the ELD-Pt CE prepared at a Pt concentration of only0.5mM was still comparable to that of the sputtered-Pt CE.

The J–V curves of the DSSCs assembled with the aforementioned Pt CEs are shown in Fig.10,and the corresponding photovoltaic parameters are also summarized in Table1.The J sc,FF andη

values

Fig.9.CVs of sputtered-Pt CE and ELD-Pt CEs with(a)different concentrations of

C6H5Na3O7·2H2O and(b)different concentrations of H2PtCl6·6H2O.The measurements

were carried out in a3-methoxypropionitrile solution consisting of50mM LiI,10mM I2,

and500mM LiClO4

Fig.10.Photocurrent density–voltage characteristics of DSSCs with sputtered-Pt CE

and ELD-Pt CEs prepared with(a)different concentrations of C6H5Na3O7·2H2O and

(b)different concentrations of H2PtCl6·6H2O.The pH value of deposition bath was

controlled at4and the deposition time was set for5min.

4677

J.-Y.Lin,Y.-T.Lin/Surface&Coatings Technology206(2012)4672–4678

of the DSSCs with ELD-Pt CEs increased slightly and decreased signif-icantly when the concentration of reductant increased.Also,notice-able was that the J sc ,FF and ηvalues of the DSSCs with ELD-Pt CEs increased gradually with the increase in Pt concentration.The in-creased J sc and FF in turn lead to higher conversion ef ?ciency,while the contribution from the V oc is less signi ?cant.In this study,a black stage was used during the photovoltaic measurements,thus the contributions from light scattering and re ?ection can be neglected.Therefore,the increased J sc mainly results from the increased active area,as indicated in Figs.8and 9.Additionally,Fig.11compares the resultant Nyquist spectra of the aforementioned Pt CEs using the sym-metric cells,while the corresponding average impedance parameters are also listed in Table 1.The R ct of the ELD-Pt CEs was found to decrease slightly and increased considerably as the concentration of reductant increased.As the Pt concentration decreased,the R ct of the ELD-Pt CEs also gradually decreased.These indicate that the increased

FF may possibly derive from the small R ct for I 3?

reduction.After the optimization,the DSSC with ELD-Pt CE prepared at pH 4with Pt con-centration of 1mM and reductant of 0.5M exhibited the highest cell performance of 7.65%,which represents an increase on the DSSC using the sputtered-Pt CE,with much less Pt loading (being ~14.82%of that of the sputtered-Pt CE).4.Conclusions

In summary,a Pt thin ?lm was coated on a TCO substrate to form a low Pt-loading and high-performance Pt CE by means of a SAM modi ?er and a low-temperature ELD technique.Firstly,the mecha-nism for coupling and activation steps was successfully proposed.The cleaned TCO substrate was made to generate the \OH functional group on the TCO surface,and thus the silane compounds were able to anchor the \OH with a Si \O \Si bond to form SAMs.Thus,the \NH 2

of the SAMs was able to further bind with the PtCl 62?

ion through elec-trostatic interaction.After the optimization of the bath formulation,the ELD-Pt CE prepared at pH 4with 0.01M H 2PtCl 6·6H 2O and 0.5M C 6H 5Na 3O 7·2H 2O had a lower Pt loading (15.91μg cm ?2),a higher speci ?c active surface area (209.71m 2g ?1)and a lower charge-transfer resistance (1.92Ω·cm 2)in comparison with that of the sputtered-Pt CE.The cell ef ?ciency of the DSSC based on such an ELD-Pt CE reached 7.65%at AM 1.5simulated full sunlight.The ef ?-ciency increased by 12.33%when compared to the sputtered-Pt CE,while the Pt loading was also remarkably reduced (at 14.82%of that of the sputtered-Pt CE).

Acknowledgments

The authors are very grateful to the National Science Council and Tatung University in Taiwan for their ?nancial supports under contract nos.NSC-99-2221-E-036-038and B99-C09-026,respectively.We also express our thanks to Dr.Wei in Tripod Technology Corporation for his helpful discussion and partly for material support.References

[1]M.Gr?tzel,Nature 414(2001)338.

[2]T.N.Murakami,S.Ito,Q.Wang,M.K.Nazeeruddin,T.Bessho,I.Cesar,P.Liska,

R.Humphry-Baker,https://www.wendangku.net/doc/0d10435696.html,te,P.Pechy,M.Gr?tzel,J.Electrochem.Soc.153(2006)A2255.

[3]K.Suzuki,M.Yamamoto,M.Kumagai,S.Yanagida,Chem.Lett.32(2003)28.[4]Z.Huang,X.Liu,K.Li,D.Li,Y.Luo,H.Li,W.Song,L.Chen,Q.Meng,Electrochem.

Commun.9(2007)596.

[5]W.J.Lee,E.Ramasamy,D.Y.Lee,J.S.Song,Sol.Energy Mater.Sol.Cells 92(2008)

814.

[6]Y.Saito,W.Kubo,T.Kitamura,Y.Wada,S.Yanagida,J.Photochem.Photobiol.A:

Chem.164(2004)153.

[7]J.G.Chen,H.Y.Wei,K.C.Ho,Sol.Energy Mater.Sol.Cells 91(2007)1472.[8]T.N.Murakami,M.Gr?tzel,Inorg.Chim.Acta 361(2008)572.

[9]M.Wang,A.M.Anghel,B.Marsan,N.C.Ha,N.Pootrakulchote,S.M.Zakeeruddin,

M.Gr?tzel,J.Am.Chem.Soc.131(2009)15976.

[10]J.Y.Lin,https://www.wendangku.net/doc/0d10435696.html,io,T.C.Wei,Electrochem.Solid State Lett.14(2011)D41.[11]J.Y.Lin,https://www.wendangku.net/doc/0d10435696.html,io,Z.Y.Hung,https://www.wendangku.net/doc/0d10435696.html,mun.13(2011)977.[12]J.Y.Lin,https://www.wendangku.net/doc/0d10435696.html,io,S.W.Chou,Electrochim.Acta 56(2011)8818.

[13]https://www.wendangku.net/doc/0d10435696.html,n,Y.Y.Wang,C.C.Wan,T.C.Wei,H.P.Feng,C.Peng,H.P.Cheng,Y.H.Chang,

W.C.Hsu,Curr.Appl.Phys.10(2010)S168.

[14]N.Papageorgiou,W.F.Maier,M.Gr?tzel,J.Electrochem.Soc.144(1997)876.[15]M.Toivola,J.Halme,K.Miettunen,K.Aitola,P.D.Lund,Int.J.Energy Res.33(2009)

1145.

[16]S.S.Kim,Y.C.Nah,Y.Y.Noh,J.Jo,D.Y.Kim,Electrochim.Acta 51(2006)3814.[17]G.Tsekouras,A.J.Mozer,G.G.Wallace,J.Electrochem.Soc.155(2008)K124.[18] C.H.Yoon,R.Vittal,J.Lee,W.S.Chae,K.J.Kin,Electrochim.Acta 52(2008)2890.[19]Y.J.Song,J.K.Oh,K.W.Park,Nanotechnology 19(2008)355602.

[20] C.Y.Lin,J.Y.Lin,C.C.Wan,T.C.Wei,Electrochim.Acta 56(2011)1941.[21] C.M.Chen,C.H.Chen,T.C.Wei,Electrochim.Acta 55(2010)1687.

[22] C.M.Chen,C.H.Chen,S.J.Cherng,T.C.Wei,Mater.Chem.Phys.55(2010)173.[23] C.Y.Lin,J.Y.Lin,https://www.wendangku.net/doc/0d10435696.html,n,T.C.Wei,C.C.Wan,Electrochem.Solid State Lett.13(2010)

D77.

[24]N.F.Fahim,T.Sekino,J.Nanopart.Res.13(2011)6409.

[25] D.H.Chen,L.Cao,F.Z.Huang,P.Imperia,Y.B.Cheng,R.A.Caruso,J.Am.Chem.Soc.

132(2010)4438.

[26] F.Z.Huang,D.H.Chen,X.L.Zhang,R.A.Caruso,Y.B.Cheng,Adv.Funct.Mater.

20(2010)15.

[27]L.Li,Z.H.He,W.W.Wang,Y.M.Fan,X.P.Mao,D.H.Chen,C.Chen,https://www.wendangku.net/doc/0d10435696.html,.Coat.

56(2006)126.

[28]Y.Lin,A.M.Ran,B.Sadanadan,E.A.Kenik,Y.P.Sun,J.Phys.Chem.B 106(2002)

1294.

[29]X.Cui,D.A.Hutt,D.J.Scurr,P.P.Conway,J.Electrochem.Soc.158(2011)D172.[30]S.S.Kim,K.W.Park,J.H.Yum,Y.E.Sung,J.Photochem.Photobiol.A 189(2007)

301.

[31]N.Papageorgiou,P.Liska,A.Kay,M.Gr?tzel,J.Electrochem.Soc.146(1999)

898.

Fig.11.Nyquist plots of the symmetric cells assembled with (a)sputtered-Pt CE and ELD-Pt CEs with different compositions:(b)1mM H 2PtCl 6·6H 2O+0.3M C 6H 5Na 3O 7·2H 2O,(c)1mM H 2PtCl 6·6H 2O+0.5M C 6H 5Na 3O 7·2H 2O,(d)1mM H 2PtCl 6·6H 2O+0.75M C 6H 5Na 3O 7·2H 2O,(e)0.3mM H 2PtCl 6·6H 2O+0.5M C 6H 5Na 3O 7·2H 2O,and (f)0.5mM H 2PtCl 6·6H 2O+0.5M C 6H 5Na 3O 7·2H 2O.The pH value of deposition bath was controlled at 4and the deposition time was set for 5min.

4678J.-Y.Lin,Y.-T.Lin /Surface &Coatings Technology 206(2012)4672–4678