Low-Temperature Processing of TitaniumOxide Nanoparticles Photoanodes for Dye-Sensitized Solar Cells

Using the low-temperature processing of different organofunctional silanes like TEOS, GPTS, and MPTS to incorporate within TiO2 network, dye-sensitized solar cells (DSCs) processed at low temperatures were obtained.eUV-curedMPTS-modi�ed layer exhibited better performance over the TEOS and GPTS, where better mechanical stable layer is achieved in addition to better interconnection between the TiO2 nanoparticles. e J-V characteristics of the DSC composed of silane-based layer showed that the improved cell performance was due to the high photocurrent density accompanied withmore dye adsorption and higher charge injection from TiO2 to FTO substrate resulting from the formation of an ohmic contact with the substrate. e highest conversion efficiency attained forMPTS-TiO2 layer cured with UV and followed by heating at 300 Cwas 3.75±0.07%, which is 2.8 times better than the GPTS-based layer.


Introduction
e increasing global demands of clean energy is becoming one of the major scienti�c challenges for scientists, economists, and politicians [1], as the combustion of fossil foils has produced widespread environmental damage [2].erefore, the Sun as an abundant source of energy represents the ideal source of clean energy, which has a solar �ux deposited on the surface of the earth within one hour as much as the global power usage.However, the high production cost compared with the fossil foils has retarded the widespread commercialization of this alternative energy.ere have been intensive investigations for higher efficiency and cost effective conversion of solar radiation to electricity.
Silicon-based solar cell technologies are currently the widely used commercial photovoltaic technology; however, the nonsilicon photovoltaic thin �lms are a major competitor with less cost and more �exibility than traditional solar cells [3].One of the most promising technologies is the dyesensitized solar cells (DSCs), which have attracted much attention as they offer the possibility of extremely inexpensive and efficient solar energy conversion with �exible routes of production.e maximum reported efficiency since the �rst report published by O'Regan and Grätzel in 1991 [4], is with a current record of ∼11-13% [5][6][7][8][9][10][11][12].DSC is a mimic of the photosynthesis and a physical separation between photon absorption and charge percolation process.e light is absorbed by a molecular dye, that is attached to the surface of semiconducting thick layer (10-15 m) deposited on transparent conducting oxide (TCO) electrode.e excited dye rapidly injects an electron into the semiconducting photoanode, where it diffuses through the sintered particle network to be collected through the TCO electrode, while the dye is regenerated via reduction by a redox shuttle electrolyte.e oxidized electrolyte is diffused to a counter electrode to complete the circuit.e recombination of the injected electron with the oxidized dye before it is regenerated, or the redox shuttle, can intercept an electron from the photoanode before it is collected, causing a drop on the total performance of the conversion process.
e intensive investigations of many researchers have been concentrated on three main components of the cell: the photoanode, the dye, and the electrolyte and/or the interfaces between them [12][13][14].Conventional DSCs require to heat the photoanode at a temperature of 450 ∘ C or higher [15].is annealing step is needed for ensuring good adherence of photoanode onto the substrate, eliminating the organic residues present in the paste or colloid, and improving the connection for efficient charge transport between the nanocrystallites that constitute the �lm.is step is difficult to be avoided and prevents the DSCs on �exible substrates that do not withstand high temperature annealing [16,17].Different reports have been published to avoid the high temperature annealing of the thick photoanode layer, for example, by not using organic compounds in the formulation of the paste [18] or by using alternative way of sintering the nanoparticles via utilizing high mechanical pressure or curing them by electron beams or UV irradiation [19][20][21].
e inorganic-organic binders like Tetraethoxysilane (TEOS), 3-glycidoxypropyltrimethoxysilane (GPTS) or 3-methacryloxypropyltrimethoxysilane (MPTS), have also been used to enhance the electrical and mechanical properties of layers consisting of ITO nanoparticles [22,23].e so-called organofunctional silanes are hybrid compounds that can be used as "molecular bridges" between inorganic materials.e advantages of such monomers include the fact that they can be readily polymerized at relatively low temperature, which allows them to be used for nanocomposite coatings that processed at low temperatures.
Recently, we have developed a DSC of multiporous structure by the modi�cation of the TiO 2 nanoparticles using MPTS, where a signi�cant enhancement of the total performance of the cell was obtained [24].In this study, we applied different silica coupling agents to coat TiO 2 nanoparticles to further lower the processing temperature as a photoelectrode.Such coupling agents are aimed to have a cross linkage between titania nanoparticles at relatively low processing temperatures (100-300 ∘ C). e microstructures, electrochemical properties, and the power conversion performance of the assembled DSCs composed of different silane networks, such as TEOS, GPTS, and MPTS, are compared and characterized.

Experimental Work
2.1.Photoelectrode Coating.Nanoparticles of TiO 2 used in this work for preparing the coating paste were obtained by hydrothermally treating TiO 2 sol, as previously reported [25].Typically, 0.3 g of TiO 2 powder is wetted and grinded by using 0.55 g of polyethylene Glycol solution (PEG400) in a mortar for 5 minutes.0.05 g of 25 wt% ethanolic solution of Pluronic P123 and 0.1 g of 50 wt% ethanolic solution of hydrolyzed silane like TEOS, GPTS, and MPTS were added to the formed paste and further grinding for 5 minutes was applied to form a homogeneous viscous paste.e TiO 2 paste was deposited onto FTO glass substrates using doctor blading technique to form a (12-14) ± 1 m thick �lm.e obtained �lms were cured by UV irradiation using a 4 W UV Mineralight multiband lamp (UV 254/365 nm, UVGL-25) for 10 hours, and �nally the layers are heated at 120, 200, and 300 ∘ C for 3 h in air.

Cell Assembly and Characterization
. e �lms were all sensitized by immersing them in 0.5 mM ruthenium-based N3 dye solution for 24 h, then rinsed in ethanol and dried in air at RT.A Pt-coated silicon substrate was used as the counter electrode, and an iodide-based solution, consisting of 0.6 M tetrabutylammonium iodide, 0.1 M lithium iodide, 0.1 M iodine and 0.5 M 4-tert-butylpyridine in acetonitrile, was used as the liquid electrolyte.Photovoltaic properties of each solar cell were characterized using simulated AM 1.5 sunlight illumination with an output power of 100 mW/cm 2 .An Ultraviolet Solar Simulator (Model 16S, Solar Light Co., Philadelphia, PA) with a 200WXenon Lamp Power Supply (Model XPS 200, Solar Light Co., Philadelphia, PA) was used as the light source, and a Semiconductor Parameter Analyzer (4155A, Hewlett-Packard, Japan) was used to measure the current and voltage.e IV measurements were repeated three times and the average results were reported.e electrochemical impedance spectroscopy (EIS) was carried out through the Solartron 1287A coupling with the Solartron 1260 FRA/impedance analyzer to investigate electronic and ionic processes in DSCs.
ermogravimetric analysis (TGA 7, Perkin-Elmer) was used to investigate the thermal behavior of the UV cured paste as a function of the annealing temperature.
X-ray Diffraction (XRD, Philips PW 1830 Diffractometer) was used to verify the phase and particle size of the TiO 2 nanoparticle �lms.e morphology of the photoanode surfaces were characterized by scanning electron microscopy (SEM, JEOL JSM-7000).e BET surface area and the pore size distribution of the layers were determined by using Brunauer-Emmett-Teller (BET, Quantachrome NOVA 4200e).
To measure the adsorbed N3 dye amount on the silane modi�ed TiO 2 �lms, the dye was desorbed by immersing dye-sensitized �lms in a 0.1 M NaOH solution in water and ethanol (1 : 1, v/v).An ultraviolet-visible-near infrared spectrophotometer (UV-VIS-NIR, Perkin Elmer Lambda 900) was employed to measure the dye concentration of the desorbed-dye solution.

Results and Discussions
3.1.DSCs Performance.e photovoltaic properties of the DSCs composed of silica and titania network were investigated for different types of hydrolyzed alkoxysilanes, such as TEOS, GPTS, and MPTS.e titania-TEOS layers formed at low temperatures up to 300 ∘ C were not stable and easily removed from substrates aer the dye sensitization.e titania-GPTS layers were stable when the samples heated at temperatures greater than 180 ∘ C, though the sample heated at 120 ∘ C was fragile aer the dye sensitization.e titania-MPTS layers were stable for all samples, which indicates the better cross-link gained by using the MPTS due to siloxanerich species especially aer the UV polymerization that promote the condensation of the silanol groups [26].Ukaji et al. [27] showed that the photo-catalytic activity and UVshielding ability were in�uenced by the number of surface e JV curves of the cells composed of the photoelectrodes treated at different temperatures for MPTS-and GPTS-modi�ed titania samples are shown in Figures 1(a) and 1(b) and Table 1.No data is shown for all TEOS-titania samples and the GPTS-titania sample which treated at 120 ∘ C, as part of the layer dissolved aer the immersion of the dye solution.For the GPTS-modi�ed sample (Figure 1(b)), it is exhibited that heating the photoanode layers from 200 to 300 ∘ C improved both the photocurrent density and the open circuit voltage with an increase of the efficiency from 0.93 ± 0.03 to 1.34 ± 0.06%.Also a minimal increase of the �lling factor was observed.On the other hand, the MPTS-modi�ed sample attained an efficiency of 1.17 ± 0.04%, which is obtained even for the sample tempered at low temperature (120 ∘ C). is means that MPTS may be considered as a suitable modi�er for low-temperature processing, where silica network interconnects the titania nanoparticle forming an ohmic contact with the FTO substrate.Further improvement was obtained by further thermal treatment to 200 ∘ C and up to 300 ∘ C, which is referred to the better contact between the titania nanoparticles, in addition to the reduction of the existence of the organics species that affect the dye loading.e conversion efficiency of the DSCs was improved considerably from 1.17 ± 0.04% to 2.86 ± 0.05% by raising the heating temperature from 120 to 200 ∘ C. e main improvement of the efficiency was coming from the enhancement of the photocurrent density from 2.45 ± 0.11 to 6.35 ± 0.13 mA/cm 2 , while a minimal change of the open circuit voltage was observed.Further step of improvements was obtained by raising the heating temperature of the MPTS-titania photoelectrode layer up to 300 ∘ C due to an increase of the open circuit voltage by 37 mV and a photocurrent density of 7.4 ± 0.14 mA/cm 2 with an overall conversion efficiency of 3.75 ± 0.07%.is value is 64% greater than the efficiency of the bare titania nanoparticles heated at the same temperature [28].Nguyen et al. [29] have reported that the enhanced performance of TiO 2 /SiO 2 electrode over bare TiO 2 one was ascribed to the role of SiO 2 as an energy barrier, increasing the physical separation of injected electrons and oxidized dyes/redox couple, and thereby retarding the recombination reactions in the resulting DSC.
e superior performance results of MPTS over the TEOS-and GPTS-based layers might be referred to the fact that MPTS is a UV curable monomer much likely than GPTS and TEOS.From previous studies of using such monomers to interconnect tin-doped indium oxide (ITO) nanoparticles, it was found that MPTS allows the best cross linking of the conducting ITO nanoparticles under UV curing process [22].e obtained sheet resistance of MPTS-ITO system was 8 KΩ, where those of GPTS-and TEOS-based system were 18 and 63 KΩ, respectively.e degree of polymerization as well as the cross linking of MPTS is enhanced by utilizing the UV irradiation as a result of the radical polymerization of the methacrylic moieties [30].e thermal curing alone is not effective for the polymerization and the condensation process of the coatings.On the opposite, the UV treatment eliminates the C=C band, strongly reduces the C=O band, and forms a well-de�ned Si�O�Si network linking homogeneously the particles together.e epoxy group in the GPTS could react with water to form glycol, which may terminate the polymerization of the monomer [31].
e main characteristic of the enhancement of the conversion efficiency of MPTS-titania layer is attributed to the improvements of the photocurrent density, which is mainly referred to the difference of the dye adsorption as shown in Figure 3. e removal of the organic species from the �lms allowed more dye to penetrate the bulk of the thick �lm, where the texture of the heated �lms was changed accordingly.e TGA results shown in Figure 2 exhibited a signi�cant weight loss for UV-treated MPTS-TiO 2 sample compared with that of GPTS-TiO 2 system.Larger surface area and pore size have a strong in�uence on the dye adsorption.It is thought that the silica network interconnected the TiO 2 nanoparticles, which enhances the electron path within the photoanode layer.It is noticeable that the outperformance of the hydrolyzed MPTS over the other silica coupling agents may refer to its capability of long chain polymerization aer the UV treatment, which allows direct pathway for the electron transport.2 belongs to the GPTS-and MPTS-based Titania paste aer 10 h UV curing and drying overnight at 80 ∘ C. e curve of the GPTS-titania sample showed a slower weight loss compared with MPTS-titania sample.e sample exhibited three distinct steps.e �rst stage is due to the removal of the physically adsorbed water up to 100 ∘ C. e second stage is up to 250 ∘ C, which is attributed to the release of the surfactant and the polyethylene glycol.e third stage is between 250 and 350 ∘ C, which is due to the condensation of the silica network.e degradation of the organic species in the paste of MPTS-titania exhibited a continuous weight loss of 4.5%, 21%, and 55% at the working temperatures 120, 200, and 300 ∘ C. Contrary to the GPTS-titania paste, no signi�cant weight loss for the MPTS-titania sample was observed at temperature higher than 300 ∘ C, which means that most of the hydrogel network can be removed at relatively low processing temperature.e UV irradiation of the TiO 2 paste may be resulted in a degradation of the organic groups, where it is common to use titanium oxide as a mean to degrade the organic wastes [32,33].

Dye
Loading.e amount of the dye loaded on the heattreated MPTS-titania photoelectrode �lms were determined using UV-Vis spectroscopy.e dye ��1� was �rst detached from the layers using 5 mL 0.1 mM KOH solution.shows that the sample absorption of the detached dye increases by increasing the heat treatment temperature.
Based on the experimental data shown in the UV-Vis absorption spectra, the amounts of dye loaded on each layer is estimated to be 2.75 × 10 −8 , 6.53 × 10 −8 , and 9.21 × 10 −8 mol/cm 2 for the samples treated at 120, 200, and 300 ∘ C, respectively.e amount of the dye loaded for the GPTS-titania sample heated at 300 ∘ C was 3.2 × 10 −8 mol/cm 2 .e amount of the dye loaded on the layers is an indication that the porous structure obviously improved the absorption of dye because of its large surface areas as shown in Table 2. e signi�cant improvement in dye loading could result in a better light harvesting and a higher photocurrent conversion efficiency of the device, which is obvious from the results in Table 1.
3.4.BET Surface Area.e heat treatment temperature strongly in�uences the texture properties of the MPTS-TiO 2 electrodes, which were investigated by N 2 adsorption measurements.e resulting adsorption and desorption isotherms for the different heat-treated MPTS-based samples exhibited a hysteresis loop with type IV isotherms with a sharp capillary condensation step at high relative pressures (   as shown in Figure 4. e surface area, pore volume, and average pore diameter were calculated by the BET method and the results are given in Table 2.An increase of the BET surface area from 32.6 to 56 and then to 70 m 2 /g was observed when the heating temperature was raised from 120 to 200 and then to 300 ∘ C, respectively.A relative increase of the pore size was calculated (Table 2) corresponding to the increase of the heating temperature, which indicates a removal of organic species of the binder.e increase of the surface area and the pore size might well explain the increased amount of dye loaded as revealed in Figure 3.   ese multiporous structures resulted from the decomposition of organic binder, which could be expected to have a larger surface area and more convenient dye loading as well as more electrolyte transportation.the inverse of the recombination lifetime (  ) or electron lifetime (  ) in the TiO 2 �lm [36].e increase of the recombination rate of the 300 ∘ C sample might be attributed to more and easy electrolyte penetration due to high mesoporous surface area, large porosity, and large pores, which makes the recombination centers closer to the TiO 2 surfaces.e Ohmic serial resistance ( 1 ) corresponds to the overall series resistance decreased gradually by raising the heating temperature from 120 to 300 ∘ C. Such decrease is expected due to better interconnection between the TiO 2 nanoparticles, which results in the improvement of the electron path and consequently higher photocurrent density will be observed.

Conclusions
Among three silane coupling agents investigated, MPTS shows a great promise for low-temperature processing of TiO 2 photoanodes for DSCs.e J-V characteristics of the DSC composed of MPTS-based layer showed that the improved cell performance was due to the high photocurrent density facilitated with more dye adsorption and higher charge injection from TiO 2 to FTO substrate resulting from the formation of an Ohmic contact with the substrate.e conversion efficiency attained for MPTS-TiO 2 layer cured with UV and followed by heating at 200 ∘ C was 2.86 ± 0.05%, where further heating at 300 ∘ C raised the power conversion efficiency to 3.75 ± 0.07%.

Figure 3 AbsorptionF 3 :
e UV-VIS absorption of the deattached dye from MPTS/TiO 2 layers treated with UV curing and further annealing at different temperatures.

3. 5 .
Structural Properties 3.5.1.XRD.e diffraction patterns of the MPTS-TiO 2 nanoparticulate layers deposited on FTO glass substrate, UV cured in addition heating at 120, 200, and 300 ∘ C are shown in Figure5.All peaks correspond to Anatase phase, like the original nanopowder[25].No existence for other phases were detected, and almost no change in the crystallite size was observed at the different heating temperature.

3. 6 .
Surface Morphology.e SEM surface morphologies with different magni�cation of the prepared MPTS-TiO 2 electrodes heated at 120, 200, and 300 ∘ C are shown in Figures6(a), 6(b), and 6(c), respectively.As shown in Figure6(a), the sample treated at 120 ∘ C showed homogeneous surface with obvious loosely packed aggregates dispersed in organic matrix.e 200 and 300 ∘ C treated samples showed sharper morphology and closely packed aggregates were formed; however, a closer inspection revealed that the TiO 2 �lm treated at higher temperature had indeed multiporous structure made up of nanocrystalline titanium dioxide particles.e 300 ∘ C treated sample showed larger porous structure.

F 4 :F 5 :
Adsorption-desorption isotherm using N 2 at 77 K for MPTS/TiO 2 samples cured with UV an further annealing at different temperatures.XRD patterns of MPTS-TiO 2 nanoparticulate �lms on FTO substrates aer UV curing and further heating at 120, 200 and 300 ∘ C.

F 6 : 1 F 7 :
Spectroscopy.e impedance data measured at the open-circuit voltage under dark condition with electrochemical impedance spectroscopy (EIS) was utilized to determine the charge-transfer resistance between the transparent-conducting oxide (FTO) and the TiO 2 porous SE� images of porous ��TS modi�ed TiO 2 �lms produced by curing them using U� irradiation and further annealing at (a) 120 ∘ C, (b) 200 ∘ C, and (c) 300 ∘ C.layer and the interface between TiO 2 /dye/electrolyte, as shown in Figure7.e measured frequency range was from 0.1 Hz to 100 kHz.Usually, the impedance plot of a DSC exhibits two semicircles or arcs in the Nyquist diagram [34, 35�.e �rst semicircle in the high-frequency region is attributed to the impedance of the charge transfer process at the counter electrode.e second semicircle in the middlefrequency region is related to the charge-transfer process at the TiO 2 /dye/electrolyte interface as well as the resistance to the transport of electrons to the working electrode.e impedance data is interpreted using equivalent electrical circuit elements as shown in the inset of Figure7.In this study, the charge transfer resistance ( ct ) increased from 71 ± 6 Ω to 144 ± 5 Ω when the annealing temperature increased from 120 ∘ C to 200 ∘ C and a reduction to 126 ± 9 Ω by further increase to 300 ∘ C. is reveals that the lowtemperature sample suffered from a high charge combination rate due to the existence of some organic residues that were present in the photoanode.is may be due to the decrease in pore size of the TiO 2 �lm and thus decreasing the dye adsorption as well as penetration of redox couples into the pores of TiO 2 electrode.Correspondingly, the characteristic frequency peaks shied to lower frequency when the heating temperature increased to 200 ∘ C and then shied slightly back again to higher frequency when the heating temperature was raised to 300 ∘ C. e characteristic frequency is related to EIS spectra showing Nyquist plots of the MPTS/TiO 2 DSC annealed at different temperatures aer UV irradiation.e inset shows the equivalent circuit.
T 1: e DSC performance parameters for GPTS-and MPTS-modi�ed TiO 2 photoanode treated with UV curing and further annealing at different temperatures.
T 2: BET results for the MPTS-based TiO 2 photoelectrodes annealed at the different temperatures.