Solvent effects on adsorption kinetics, dye monolayer, and cell performance of porphyrin-sensitized solar cells

Abstract

The effect of three dye loading solvents (THF, ethanol, and n-butanol) on the adsorption kinetics, the binding mode of the dye molecules on the TiO₂ nanoparticles, and photovoltaic performance in dye-sensitized solar cells was reported. The kinetic study indicated that solvents greatly affected the adsorption amount and adsorption rate on the surface of TiO₂ nanoparticles. The adsorption data in THF, ethanol and n-butanol were analyzed by applying an intraparticle diffusion model, pseudo-first order model, and pseudo-second order model, respectively. It was found that adsorption rate was controlled by mass transfer processes. UV-vis spectra and ATR-FTIR spectra of porphyrin-coated electrodes were measured to study the intermolecular interaction and binding mode on TiO₂ films, respectively. Finally, the cell performance was tested under different solvents and the best efficiency of 8.18% was obtained when n-butanol was used as the solvent for dye-loading.

---

Introduction

Sunlight as an abundant alternative energy source has promoted tremendous effort to find cost-effective technologies to convert it to electricity efficiently. Since the breakthrough made by O'Regan and Grätzel in 1991, dye-sensitized solar cells (DSCs) have attracted much attention as a promising alternative to Si-based solar cells.¹ A conventional DSC is comprised of three parts: dye-sensitized working electrode, electrolyte, and the Pt-coated counter electrode. Extensive research has been performed to enhance the energy conversion efficiency through systematic studies of the three components. Inspired by light-harvesting and electron transport roles in photosynthesis, porphyrin derivatives have been identified as promising dyes for DSCs.^2–6 In 2010, a device based on porphyrin dye YD2 achieved a power conversion efficiency of 11%,⁷ which is comparable to ruthenium complexes. In 2011, a conversion efficiency of 12.3% was reported when a porphyrin dye YD2-o-C8 and Co(II/III) electrolyte were used.⁸ Very recently, Grätzel's group reported a porphyrin dye SM315 with energy conversion efficiency up to 13%,⁹ indicating the promise of DSCs in converting solar energy to electricity cost-effectively.

Studies have shown that solvents have a significant effect on the cell performance of porphyrin dyes. Fox example, Lin and Diau et al. studied several alkoxyl and alkyl chains modified porphyrins having a donor–π–acceptor configuration.¹⁰ It was found the porphyrin bearing four OC₈H₁₇ groups produced quite different energy conversion efficiency (6.39% vs. 8.94%) when a THF and a MeOH/toluene solution were used for dye-loading. It was suggested the strong interaction between THF molecules may hamper the dye adsorption. We previously designed and synthesized three thiophene-functionalized porphyrin dyes and found cell performance is dependent on the solvents that were used during the dye loading process.¹¹ The study showed that the dye-loading can be controlled by tuning the volume ratio of ethanol/THF mixed solvents, and lower ethanol content is favorable for better power conversion when the porphyrin has more coplanar structure. From the view of device performance, solvents may exert influence on light harvesting and charge dynamics, which will result in different current density and voltage, respectively. Imahori et al. studied effects of solvents and loading time on charge carrier dynamics of porphyrin–TiO₂ electrodes.¹² After normalizing the efficiency to dye coverage, these authors found a direct correlation between power conversion efficiency and electron–cation recombination in devices with short loading time, while devices with long loading time exhibited inferior performances because of decreased electron injection efficiency. From the view of dye adsorption perspective, the solvent has influences on adsorption kinetics, and these differences may result in different device performance.

In this work, we report our systematic studies of solvent effects on adsorption kinetics, dye monolayer, and cell performance. A previously reported porphyrin ZZX-N7 with a thiophene-linker was chosen as sensitizer.¹¹ Its structure was shown in Scheme 1. THF, ethanol, and n-butanol were selected as loading solvents.^13,14 It was found that the dye adsorption kinetics is highly solvent-dependent. The adsorption rate was controlled by mass transfer processes. The adsorption process in THF solution can be best described by pseudo-second order model, whereas deviations from the second order at initial stage were found for adsorptions in EtOH and n-butanol.

Scheme 1 Molecular structure of ZZX-N7.

Experimental section

Preparation of TiO₂ films

To prepare a working electrode, a transparent layer (∼20/30 nm, Dyesol, Australia) was first printed on a cleaned FTO glass with a thickness of 8 μm. After being heated at 120 °C for 10 minutes, a 3 μm thick scattering layer (WER2-O, 150–400 nm Dyesol, Australia) was then printed. The obtained TiO₂-coated glass was heated in air flow consecutively at 120 °C for 10 minutes, 325 °C for 5 minutes, 375 °C for 10 minutes, 450 °C for 15 minutes, and then 500 °C for 30 minutes.

Dye adsorption profiles

To obtain the dye adsorption profiles, working electrodes were immersed into a cuvette containing dye solution, and the cuvette was sealed by a polytetrafluoroethylene cap. The variation of the absorbance at Q bands was recorded by a UV-vis spectrophotometer (Lambda 950, PerkinElmer, USA). The dye-loading density q_t at time t was obtained from the next equation:where C₀ is the initial concentration; A₀ and A_t is the absorbance of Q bands at t = 0 and t = t, respectively; V is the volume of dye solution; S is the total size of the films.

Data analysis

Intraparticle diffusion model, pseudo-first order model, and pseudo-second order model were used to analysis the adsorption data. The former helps to identify the steps involved during adsorption, while the latter two concern the kinetics of intrinsic adsorption reaction. The intraparticle diffusion model is expressed in following form:^15,16

q_t = k_pt^1/2 + C (2)

where q_t is the adsorption density (mol cm⁻²) at time t (min), k_p is the diffusion rate constant (mol cm⁻² min^−1/2), and C is a constant. The validity of intraparticle diffusion model can be checked by the linearity of the plot of q_t vs. t^1/2.

In the pseudo-first order model, the dye-loading rate is proportional to the dye-loading density:¹⁷

where k₁ is the pseudo-first order rate constant (min⁻¹), and q_e is the saturated dye-loading density. Integrating eqn (3) for the boundary conditions t = 0 to t = t and q_t = 0 to q_t = q_t gives

Rate parameters, k₁ and q_e can be directly obtained from the slope and intercept of the log(q_e − q_t) vs. t plot.

In the pseudo-second order model, the dye-loading rate at time t is in the second order of dye-loading density:¹⁸

where k₂ (cm² mol⁻¹ min⁻¹) is the rate constant of pseudo-second order. Integrating eqn (5) and applying the boundary conditions t = 0 to t = t and q_t = 0 to q_t = q_t gives

Eqn (6) can be rearranged to obtain a linear form

According to eqn (7), k₂ and q_e can be obtained from the intercept and slope of the plot t/q_t against t. The fitting validity of these models can be checked by each linear plot of log(q_e − q_t) vs. t, and t/q_t vs. t, respectively.

FTIR measurement

Infrared spectra of pure ZZX-N7, ZZX-N7-coated electrodes were collected on an attenuated total reflectance Fourier transform infrared spectrometry (VERTEX 70, Bruker). The ZZX-N7 powder was held in KBr pellet. The fresh prepared dye-coated electrodes were rinsed with dry MeCN, and dried with air flow subsequently. Then electrodes were immediately subjected to measurement to avoid the adsorption of H₂O. Every spectrum for samples was acquired in transmission mode with a resolution of 4 cm⁻¹ and spectral range of 4000–400 cm⁻¹. The obtained spectra were then transferred to absorption mode for analysis.

Fabrication of solar cells

Working electrodes (8 μm-thick transparent layer + 3 μm-thick scattering layer) were dipped in a TiCl₄ aqueous solution for 30 minutes at 70 °C. The electrode was then flushed with de-ionized (DI) water and ethanol and dried with an air flow. Electrodes were then sintered at 500 °C for 30 minutes. After being cooled to 80 °C, the electrodes were immersed in the dye solutions (0.2 mM) for 12 hours. The films were flushed with acetonitrile thoroughly and dried in air. The counter electrodes were prepared by casting a Pt solution on clean FTO glass and sintered at 450 °C for 20 minutes. Two electrodes were sandwiched using a 45 μm thick hot-melt ring (Surlyn, DuPont). The internal space was filled with liquid electrolytes (1.0 M 1,3-dimethylimidazolium iodide, 0.03 M iodine, 0.5 M tert-butylpyridine, 0.05 M LiI, 0.1 M guanidiniumthiocyanate, in an 85 : 15 acetonitrile/valeronitrile) using a vacuum back-filling system through two pre-drilled holes on the counter electrode. The holes were sealed with a Surlyn sheet and a thin glass cover.

Device characterizations

Current–voltage cures were obtained by using an AM 1.5G solar simulator equipped with a 450 W xenon light (Oriel, model 9119) and an AM 1.5G filter (Oriel, model 91192). During the I–V measurement, a 0.09 cm² mask was used to get a uniform working area for all the cells. In the incident photon to electron conversion efficiency (IPCE) measurement, light from a 300 W xenon lamp (ILC Technology, U.S.A.) was focused through a Gemini-180 double monochromator (Jobin Yvon Ltd., U.K.) onto the cell under test. The monochromator was incremented through the visible spectrum to generate IPCE spectra. A white light bias (1% sunlight intensity) was applied onto the sample during the testing with an AC model (10 Hz).

Photovoltage decay and charge extraction measurements

The determination of the electron life time and charge amounts was carried out by performing transient photovoltage decay measurement and charge extraction experiment, respectively. The measurement detail has been reported in previous work.¹⁹

Results and discussion

Adsorption kinetics and mechanisms

Adsorption kinetics was measured under three different solvents (i.e., THF, EtOH, and n-butanol). Fig. 1a shows adsorption profiles of ZZX-N7. Both adsorptions in THF and EtOH were completed within 12 hours, while saturated dye-loading density (q_e) of adsorption in n-butanol was obtained by prolonging the loading time to 24 hours. The q_e was 0.70 × 10⁻⁷ mol cm⁻², 2.23 × 10⁻⁷ mol cm⁻², and 1.78 × 10⁻⁷ mol cm⁻² for adsorptions in THF, EtOH, and n-butanol, respectively.

Fig. 1 (a) Adsorption profiles of ZZX-N7 in different solvents; (b) test of intraparticle diffusion model. Solid symbol: experimental data; solid line: fitting result.

We found that solvent has a significant influence on the adsorption rate. In THF solution, the amount of adsorbed ZZX-N7 reached ∼86% q_e within the first one hour. The q_t increased smoothly from 1–6 hours. After 6 hours, the adsorption was saturated. Unlike the adsorption in THF, adsorptions in EtOH and n-butanol exhibited slow adsorption rate in an initial phase. Especially for the adsorption in n-butanol, the adsorbed amount was nearly zero at first 1 hour. It took about 4 hours and 20 hours to reach 80% q_e for adsorptions in EtOH and n-butanol, respectively.

To get insights into the difference in adsorption rate under different loading solvents, adsorption data were fitted to intraparticle diffusion model to identify the steps involved during the adsorption process. Theoretically, as shown in Scheme 2, adsorption of dyes onto TiO₂ nanoparticles involves four steps:^20–22 (i) diffusion of dye molecules from bulking solution to the boundary layer film (a film of solvent) surrounding the TiO₂ particle, bulk diffusion; (ii) transport of the dye from the boundary layer onto the external TiO₂ surface, boundary layer diffusion; (iii) adsorption at active sites; (iv) diffusion of the unbound dye into the pores, intraparticle diffusion. It is generally accepted that process (iii) is much faster than the other three steps, and it unlikely to be the rate-determining step. Bulk diffusion is also often ignored in adsorption systems with sufficient stirring, but this is not the situation in present work. According to the previous studies, the plot of q_t vs. t^1/2 will exhibit multiple-linearity for adsorptions involving two or more steps.^15,23

Scheme 2 Adsorption steps for dyes to mesoporous TiO₂ films: (i) bulk diffusion; (ii) boundary layer diffusion; (iii) adsorption at active sites; (iv) intraparticle diffusion.

Fig. 1b exhibits q_t vs. t^1/2 plots under different loading solvents. It was observed that all three adsorptions exhibited multiple-linearity. The plot of adsorption in THF showed three linear sections with different slopes. The first line, which is the steepest, represents the diffusion process of dye molecules through the boundary layer to the TiO₂ surface (i.e., boundary layer diffusion).²² The second line represents the intraparticle diffusion of dye molecules into pores. The smaller slope of this line is in consistent with the gradual adsorption in this stage as shown in Fig. 1a. The third section is a flat line, where adsorption is saturated. It should be noted that the multiple-linearity doesn't mean boundary layer diffusion and intraparticle diffusion happened separately. Instead, boundary layer diffusion and intraparticle diffusion are more likely to be inseparable and slopes of lines help determine the predominant process in a certain period. From Fig. 1b it was found that plots of adsorptions in EtOH and n-butanol cannot be fitted to a straight line at initial adsorption stage. Considering the relative low adsorbed amount at this period, this stage should be the combination of bulk diffusion and boundary layer diffusion. The similar phenomenon wasn't observed in THF solution. It should be because of THF's relatively low viscosity (THF (0.53 cP) < EtOH (1.17 cP) < n-butanol (2.95 cP)), which is favorable for fast mass transfer. To demonstrate this assumption, we reduced the initial concentration of THF solution from 0.2 mM to 0.1 mM and 0.037 mM as low concentration also causes slow mass transfer. It was observed that the non-linearity at initial stage became more significant (Fig. S1†). Thus, the initial stage is mainly controlled by bulk diffusion. Following the initial non-linearity, three lines were also observed for adsorption in EtOH solution, which successively represents boundary layer diffusion, intraparticle diffusion, and equilibrium stage. The q_t vs. t^1/2 plot of adsorption in n-butanol solution resembled the one in EtOH solution except the equilibrium stage wasn't observed in the current time range. Therefore, it can be concluded that the overall adsorption rate is limited by the combination of mass transport processes. Increasing the initial concentration and/or lowering the viscosity of loading solvent are favorable for fast adsorption.

The adsorption data were also analyzed using pseudo-first order model and pseudo-second order model to study the kinetics of intrinsic adsorption reaction. Fig. 2a and b exhibit the fitting results of pseudo-first order model and pseudo-second order model, respectively. The detailed parameters are shown in Table 1. It was observed from Fig. 2b that the adsorption in THF solution can be well described by pseudo-second order model for the entire period of adsorption. Both the high correlation coefficients (R², 0.993) and similarity between calculated saturated dye-loading density (q_ecal, 0.74 × 10⁻⁷ mol cm⁻²) and experimental one (q_e, 0.70 × 10⁻⁷ mol cm⁻²) indicate the excellent correlation between the experimental data and the model. This result suggests that the adsorption of ZZX-N7 onto TiO₂ surface in THF is a second order reaction.

Fig. 2 (a) Test of pseudo-first order model; (b) test of pseudo-second order model.

Table 1 Kinetic parameters for pseudo-first and pseudo-second order models

Solvent	qe (10^−7 mol cm^−2)	Pseudo-first order model			Pseudo-second order model
		qecal (10^−7 mol cm^−2)	k1 (10^−3 min^−1)	R^2	qecal (10^−7 mol cm^−2)	k2 (10^−3 cm^2 mol^−1 min^−1)	R^2
THF	0.70	0.68	56.2	0.980	0.74	53.77	0.993
EtOH	2.23	2.57	9.57	0.889	3.65	0.87	0.869
n-Butanol	1.78	1.69	3.50	0.962	2.63	0.13	0.957

---

For adsorptions in EtOH and n-butanol, we found that they cannot be described by either pseudo-first order model or pseudo-second order model. Instead, as shown in Fig. 3, the kinetics can be divided into three portions: (i) in initial phase, the kinetics cannot be described by either pseudo-first or pseudo-second order model; (ii) at the second portion, the kinetics can be best described by the pseudo-first order model; (iii) at the third portion, the kinetics can be best described by pseudo-second order model. The detailed piecewise fitting results are shown in Table 2. It was observed that the q_ecal obtained from the pseudo-second order portion (i.e., the third portion) is more similar with the experimental q_e. This result suggests that the second order mechanism is predominant for adsorption reaction in EtOH and n-butanol solutions. The deviation from pseudo-second order model in the initial phase has been studied by Azizian.²⁴ The author ascribed the deviation to the heterogeneity of adsorbent. In this regard, adsorption kinetics in THF solution should also exhibit similar feature, which wasn't observed in present work. Thus, other factors may be responsible for the deviation. The reaction kinetics can also be affected by the relative concentration of reactants.²⁵ We then plotted t/q_t against t for adsorptions in 0.1 mM and 0.037 mM THF solutions, and non-linearity at an initial stage was also observed (Fig. S2†). Therefore, we may conclude that the deviation from pseudo-second order model should be because of slow bulk diffusion in these solutions, causing porphyrin's relative low concentration at TiO₂ surface.

Fig. 3 Piecewise fitting of adsorption kinetics in EtOH and n-butanol. Solid symbol: experimental data; solid line: fitting result of pseudo-first order model; dash line: fitting result of pseudo-second order model.

Table 2 Kinetic parameters for piecewise fitting

Solvent	qe (10^−7 mol cm^−2)	Pseudo-first order portion			Pseudo-second order portion
		qecal (10^−7 mol cm^−2)	k1 (10^−3 min^−1)	R^2	qecal (10^−7 mol cm^−2)	k2 (10^3 cm^2 mol^−1 min^−1)	R^2
EtOH	2.23	6.82	12.32	0.993	2.22	510.28	0.999
n-Butanol	1.78	2.26	2.17	0.998	1.80	1.42	0.999

---

Effect of solvents on ZZX-N7 monolayer on TiO₂

To study the influence of loading solvents on the structure of the porphyrin monolayers on TiO₂ surface, absorption spectra ZZX-N7 on 2.5 μm-thick TiO₂ electrodes were measured under different solvents. As shown in Fig. 4, the electrode obtained from ZZX-N7/THF solution showed similar absorption spectra with the counterpart in solution, but the blue-shoulder at the Soret band and broadened absorption bands indicate the existence of dye aggregates.^26,27 In spite different loading solvents were used, the blue shoulder suggested that ZZX-N7 formed H-aggregates on electrodes.^28,29 Upon changing the loading solvent from THF to n-butanol and EtOH, absorption bands were much broader and features of Soret band became blurred. Thus, more serious aggregation was expected on electrodes obtained from n-butanol and EtOH. The increased aggregates should be because of the higher dye-loading.^30,31 High dye-loading is favorable for better light-harvesting, but this beneficial effect may be compensated by the increased energy loss in dye aggregates, which would reflect on the photovoltaic properties.

Fig. 4 Absorption spectra of ZZX-N7 in EtOH solution and counterparts on TiO₂. The immersing time was 12 hours.

The attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy was used to study the adsorption mode of dye molecules to TiO₂ surface. Fig. 5 gives ATR-FTIR spectra of ZZX-N7 powder and ZZX-N7-coated electrodes. The spectrum of ZZX-N7 powder in a KBr pellet exhibits a strong signal at 1689 cm⁻¹, which is attributable to the ν(C=O) stretching mode of the carboxylic acid group.^32,33 This signal disappears after the dye was coated on the TiO₂, while two new bands located at 1595–1597 cm⁻¹ and 1385–1390 cm⁻¹ were observed. These two bands should be the antisymmetric and symmetric ν(–CO₂⁻) modes of carboxylate, respectively.³² The disappearance of ν(C=O) stretching signal indicates that all dye molecules were bound to TiO₂ and excludes the existence of dye multilayer. On the other hand, the appearance of antisymmetric and symmetric ν(–CO₂⁻) signals reveals that ZZX-N7 molecules are bound to TiO₂ surface via a bridging bidentate mode even different loading solvents were used and.^34–36 We also measured the ATR-FTIR spectra of ZZX-N7-coated electrodes with a short loading time (Fig. S3†), which resembled the ones with long loading time. Thus, dye adsorption may exhibit different kinetics in different time stage, but dyes are bound to TiO₂ in the same mode (i.e., bridging bidentate mode).

Fig. 5 ATR-FTIR spectra of ZZX-N7-coated electrodes obtained from (a) THF, (b) EtOH, and (c) n-butanol for 12 hours, and (d) ZZX-N7 powder.

Photovoltaic properties under different loading solvent

The solvents effects on cell performances were investigated. The best performance under different solvents are increased in the order THF-cell (V_oc = 592 mV, J_sc = 7.81 mA cm⁻², FF = 70.97, η = 3.46%) < EtOH-cell (V_oc = 732 mV, J_sc = 15.39 mA cm⁻², FF = 63.33, η = 7.51%) < n-butanol-cell (V_oc = 728 mV, J_sc = 15.79 mA cm⁻², FF = 67.62, η = 8.18%). This trend is in consistent with the average parameters detailed in Table 3. The considerably inferior performance of THF-cell is due to much lower V_oc and J_sc (Fig. 6a). EtOH-cell and n-butanol-cell exhibited similar V_oc and J_sc, but the η of n-butanol-cell is slightly higher because of increased fill factor (FF). The lower FF of EtOH-cell should be due to more serious dye aggregates in EtOH-cell.³²

Table 3 Average photovoltaic performances (with standard deviation) of ZZX-N7 using different loading solvents. Three independent devices were measured to obtain the average values

Solvent	^a (mA cm^−2)	Jsc (mA cm^−2)	Voc (mV)	FF (%)	η (%)
a is derived from the integration of IPCE values.
THF	7.64 ± 0.14	7.72 ± 0.10	588 ± 5	70.64 ± 0.67	3.38 ± 0.07
EtOH	15.10 ± 0.20	15.31 ± 0.25	730 ± 2	63.40 ± 0.74	7.45 ± 0.05
n-Butanol	15.12 ± 0.23	15.33 ± 0.35	725 ± 4	66.76 ± 0.32	7.88 ± 0.25

---

Fig. 6 I–V curves (a) and IPCE spectra (b) of ZZX-N7-sensitized solar cells under different loading solvents.

As shown in Fig. 6b, EtOH-cell and n-butanol-cell exhibited similar IPCE spectra, while the IPCE of THF-cell was much lower than the former two, which is in consistent with the J_sc. According to adsorption kinetics study, the dye-loading density was 0.70 × 10⁻⁷ mol cm⁻², 2.23 × 10⁻⁷ mol cm⁻², and 1.78 × 10⁻⁷ mol cm⁻² for adsorptions in THF, EtOH, and n-butanol, respectively. The much lower dye-loading in THF solution implies inferior light-harvesting ability of THF-cell, which was demonstrated by the light-harvesting efficiency (LHE) spectra (Fig. S4†). Thus, the low J_sc of THF-cell should be ascribed to lower dye-loading. The LHE of EtOH-electrode was slightly higher than n-butanol-electrode. The IPCE spectra, however, indicated that the advantage in light harvesting wasn't successfully transferred to higher IPCE. It has been reported that higher dye-loading may induce more serious dye aggregation on TiO₂ electrode, and reduced electron injection efficiency is expected.³⁰

Transient photovoltage decay measurements were performed to get insights into the difference in V_oc. It was widely recognized that a change in V_oc may originate from the shift of TiO₂'s Fermi level or different charge recombination kinetics.³⁷ As shown in Fig. 7a, THF-cell and n-butanol-cell exhibited same Q–V_oc plot, suggesting a fixed difference between the TiO₂ Fermi level and the electrolyte Fermi-level. Comparing the former two plots, the one of EtOH-cell feature a lower V_oc at the same extracted charge (Q), indicating a downshift of TiO₂ Fermi level in EtOH-cell. As reported in previous work,^38–40 the downshift of TiO₂ Fermi level should be ascribed to the adsorption of protons coming from EtOH. Fig. 7b shows the life time of injected electrons. It can be observed from the figure that EtOH-cell exhibited longest electron life time, indicating the most retarded charge recombination at TiO₂/electrolyte interface. n-Butanol-cell gave slightly smaller life time at given V_oc as compared to EtOH-cell. But this difference was compensated by higher TiO₂ Fermi level in n-butanol cell. Thus n-butanol cell and EtOH-cell exhibited nearly the same V_oc.

Fig. 7 Comparison of charges extracted from ZZX-N7-sensitized TiO₂ films at a certain V_oc (a) and electron lifetime against V_oc (b).

Conclusion

We have systematically studied the effect of loading solvents (THF, EtOH, and n-butanol) on adsorption structure, adsorption kinetics, and photovoltaic properties. ATR-FTIR measurements suggested that dye molecules were adsorbed onto TiO₂ surface via bidentate mode under different solvents. It was found that solvents played a significant role on adsorption kinetics. The results indicated that the saturated dye-loading was increased from THF to n-butanol and EtOH. The higher dye-loading in the latter two solvents induced more dye aggregates on TiO₂ surface as presented by absorption spectra of dye-coated electrodes. Except for the dye-loading, differences were also found in adsorption rate. By fitting the adsorption kinetic data to intraparticle diffusion model, it was found that all three adsorptions were controlled by diffusion process, and the relative lower adsorption rates in EtOH and n-butanol were because of their high viscosity. Kinetic data were also analyzed by pseudo-first and pseudo-second order model to study the intrinsic adsorption reaction kinetic. The adsorption reaction in THF solution can be best described by pseudo-second order model in the entire period, while deviations from second order at initial stage were found for adsorptions in EtOH and n-butanol. The power conversion efficiency was increased from THF (3.46%) < EtOH (7.51%) < n-butanol (8.18%). The higher efficiency of n-butanol-cell was ascribed to appropriate dye-loading.

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflict of interest

The authors declare no competing financial interest.

Acknowledgements

This work was partially supported by the National Basic Research Program of China (973 program), Grant No. 2011CBA00703, the Fundamental Research Funds for the Central Universities, Grant No. HUST: 2014TS016 (Z. Z.), Eastern Illinois University the President's Fund for Research and Creative Activity (H. H.), Science and Technology Project Supported by Education Department of Jiangxi Province, Grant No. GJJ150644 (W. L.).

References

1. B. O'Regan and M. Grätzel, A Low-Cost, High-Efficiency Solar Cell Based on Dye-Sensitized Colloidal TiO₂ Films, Nature, 1991, 353, 737–740.
2. M. K. Nazeeruddin, A. Kay, R. R. Humpbry-Baker, E. Miiller, P. Liska, N. Vlachopoulos and M. Grätzel, Conversion of Light to Electricity by cis-X₂Bis(2,2′-bipyridyl-4,4′-dicarboxylate)ruthenium(II) Charge-Transfer Sensitizers (X = C1⁻, Br⁻, I⁻, CN⁻, and SCN⁻) on Nanocrystalline TiO₂ Electrodes, J. Am. Chem. Soc., 1993, 115, 6382–6390.
3. P. Wang, S. M. Zakeeruddin, J. E. Moser, R. R. Humpbry-Baker, P. Comte, V. Aranyos, A. Hagfeldt, M. K. Nazeeruddin and M. Grätzel, Stable New Sensitizer with Improved Light Harvesting for Nanocrystalline Dye-Sensitized Solar Cells, Adv. Mater., 2004, 16, 1806–1811.
4. M. K. Nazeeruddin, F. D. Angelis, S. Fantacci, A. Selloni, G. Viscardi, P. Liska, S. Ito, B. Takeru and M. Grätzel, Combined Experimental and DFT-TDDFT Computational Study of Photoelectrochemical Cell Ruthenium Sensitizers, J. Am. Chem. Soc., 2005, 127, 16835–16847.
5. F. Gao, Y. Wang, D. Shi, J. Zhang, M. Wang, X. Jing, R. Humphry-Baker, P. Wang and S. M. Zakeeruddin, et al., Enhance the Optical Absorptivity of Nanocrystalline TiO₂ Film with High Molar Extinction Coefficient Ruthenium Sensitizers for High Performance Dye-Sensitized Solar Cells, J. Am. Chem. Soc., 2008, 130, 10720–10728.
6. T. Higashino and H. Imahori, Porphyrins as Excellent Dyes for Dye-Sensitized Solar Cells: Recent Developments and Insights, Dalton Trans., 2015, 44, 448–463.
7. T. Bessho, S. M. Zakeeruddin, C.-Y. Yeh, E. W.-G. Diau and M. Grätzel, Highly Efficient Mesoscopic Dye-Sensitized Solar Cells Based on Donor–Acceptor-Substituted Porphyrins, Angew. Chem., Int. Ed., 2010, 49, 6646–6649.
8. A. Yella, H.-W. Lee, H. N. Tsao, C. Yi, A. K. Chandiran, M. K. Nazeeruddin, E. W.-G. Diau, C.-Y. Yeh, S. M. Zakeeruddin and M. Grätzel, Porphyrin-Sensitized Solar Cells with Cobalt(II/III)-based Redox Electrolyte Exceed 12 Percent Efficiency, Science, 2011, 334, 629–634.
9. S. Mathew, A. Yella, P. Gao, R. Humphry-Baker, B. F. E. Curchod, N. Ashari-Astani, I. Tavernelli, U. Rothlisberger, M. K. Nazeeruddin and M. Grätzel, Dye-Sensitized Solar Cells with 13% Efficiency Achieved through the Molecular Engineering of Porphyrin Sensitizers, Nat. Chem., 2014, 6, 242–247.
10. C.-L. Wang, C.-M. Lan, S.-H. Hong, Y.-F. Wang, T.-Y. Pan, C.-W. Chang, H.-H. Kuo, M.-Y. Kuo, E. W.-G. Diau and C.-Y. Lin, Enveloping Porphyrins for Efficient Dye-Sensitized Solar Cells, Energy Environ. Sci., 2012, 5, 6933–6940.
11. W. Li, Z. Liu, H. Wu, Y.-B. Cheng, Z. Zhao and H. He, Thiophene-Functionalized Porphyrins: Synthesis, Photophysical Properties, and Photovoltaic Performance in Dye-Sensitized Solar Cells, J. Phys. Chem. C, 2015, 119, 5265–5273.
12. H. Imahori, S. Kang, H. Hayashi, M. Haruta, H. Kurata, S. Isoda, S. E. Canton, Y. Infahsaeng and A. Kathiravan, et al., Photoinduced Charge Carrier Dynamics of Zn–Porphyrin–TiO₂ Electrodes: The Key Role of Charge Recombination for Solar Cell Performance, J. Phys. Chem. A, 2011, 115, 3679–3690.
13. J. Lu, X. Xu, K. Cao, J. Cui, Y. Zhang, Y. Shen, X. Shi, L. Liao, Y. Cheng and M. Wang, D–π–A Structured Porphyrins for Efficient Dye-Sensitized Solar Cells, J. Mater. Chem. A, 2013, 1, 10008–10015.
14. L. Cabau, C. V. Kumar, A. Moncho, J. N. Clifford, N. López and E. Palomares, A Single Atom Change “Switches-On” the Solar-to-Energy Conversion Efficiency of Zn–Porphyrin Based Dye Sensitized Solar Cells to 10.5%, Energy Environ. Sci., 2015, 8, 1368–1375.
15. G. McKay, M. S. Otterburn and A. G. Sweeney, The Removal of Color from Effluent Using Various Adsorbents III Silica: Rate Processes, Water Res., 1980, 14, 15–20.
16. A. E. Daifullah, S. El-Reefy and H. Gad, Adsorption of p-Nitrophenol on Inshas Incinerator Ash and on Pyrolysis Residue of Animal Bones, Adsorpt. Sci. Technol., 1997, 15, 485–496.
17. C. Y. Chang, W. T. Tsai, C. H. Ing and C. F. Chang, Adsorption of Polyethylene Glycol (PEG) from Aqueous Solution onto Hydrophobic Zeolite, J. Colloid Interface Sci., 2003, 260, 273–279.
18. F.-C. Wu, R.-L. Tseng and R.-S. Juang, Kinetic Modeling of Liquid-Phase Adsorption of Reactive Dyes and Metal Ions on Chitosan, Water Res., 2001, 35, 613–618.
19. C.-Y. Chen, M. Wang, J.-Y. Li, N. Pootrakulchote, L. Alibabaei, C. Ngocle, J.-D. Decoppet, J.-H. Tsai, M. Grätzel and C.-G. Wu, et al., Highly Efficient Light-Harvesting Ruthenium Sensitizer for Thin Film Dye-Sensitized Solar Cells, ACS Nano, 2009, 3, 3103–3109.
20. E. Guibal, C. Milot and J. M. Tobin, Metal-Anion Sorption by Chitosan Beads: Equilibrium and Kinetic Studies, Ind. Eng. Chem. Res., 1998, 37, 1454–1463.
21. N. Kannan and M. M. Sundaram, Kinetics and Mechanism of Removal of Methylene Blue by Adsorption on Various Carbons—a Comparative Study, Dyes Pigm., 2001, 51, 25–40.
22. W. H. Cheunga, Y. S. Szeto and G. McKay, Intraparticle Diffusion Processes during Acid Dye Adsorption onto Chitosan, Bioresour. Technol., 2007, 98, 2897–2904.
23. G. McKay, Adsorption of Dyestuffs from Aqueous Solutions Using Activated Carbon, J. Chem. Technol. Biotechnol., 1983, 33, 196–204.
24. S. A. Azizian, Novel and Simple Method for Finding the Heterogeneity of Adsorbents on the Basis of Adsorption Kinetic Data, J. Colloid Interface Sci., 2006, 302, 76–81.
25. C.-R. Lee, H.-S. Kim, I.-H. Jang, J.-H. Im and N.-G. Park, Pseudo First-Order Adsorption Kinetics of N719 Dye on TiO₂ Surface, ACS Appl. Mater. Interfaces, 2011, 3, 1953–1957.
26. M. Kasha, Energy Transfer Mechanisms and the Molecular Exciton Model for Molecular Aggregates, Radiat. Res., 1963, 20, 55–71.
27. C.-F. Lo, L. Luo, E. W.-G. Diau, I.-J. Chang and C.-Y. Lin, Evidence for the Assembly of Carboxyphenylethynyl Zinc Porphyrins on Nanocrystalline TiO₂ Surfaces, Chem. Commun., 2006, 1430–1432.
28. N. C. Maiti, S. Mazumdar and N. Periasamy, J- and H-Aggregates of Porphyrin–Surfactant Complexes: Time-Resolved Fluorescence and Other Spectroscopic Studies, J. Phys. Chem. B, 1998, 102, 1528–1538.
29. U. Sigge, U. Bindig, C. Endisch, T. Komatsu, E. Tsuchida, J. Voigt and J.-H. Fuhrhop, Photophysical and Photochemical Properties of Porphyrin Aggregates, Ber. Bunsenges. Phys. Chem., 1996, 100, 2070–2075.
30. Q. Wang, W. M. Campbell, E. E. Bonfantani, K. W. Jolley, D. L. Officer, P. J. Walsh, K. Gordon, R. H. Baker, M. K. Nazeeruddin and M. Grätzel, Efficient Light Harvesting by Using Green Zn–Porphyrin-Sensitized Nanocrystalline TiO₂ Films, J. Phys. Chem. B, 2005, 109, 15397–15409.
31. T. Hasobe, H. Imahori, S. Fukuzumi and P. V. Kamat, Nanostructured Assembly of Porphyrin Clusters for Light Energy Conversion, J. Mater. Chem., 2003, 13, 2515–2520.
32. K. S. Finnie, J. R. Bartlett and J. L. Woolfrey, Vibrational Spectroscopic Study of the Coordination of (2,2′-Bipyridyl-4,4′-dicarboxylic acid)ruthenium(II) Comple Surface of Nanocrystalline Titania, Langmuir, 1998, 14, 2744–2749.
33. M. K. Nazeeruddin, R. Humphry-Baker, P. Liska and M. Grätzel, Investigation of Sensitizer Adsorption and the Influence of Protons on Current and Voltage of a Dye-Sensitized Nanocrystalline TiO₂ Solar Cell, J. Phys. Chem. B, 2003, 107, 8981–8987.
34. A. Vittadini, A. Selloni, F. P. Rotzinger and M. Grätzel, Formic Acid Adsorption on Dry and Hydrated TiO₂ Anatase (101) Surfaces by DFT Calculations, J. Phys. Chem. B, 2000, 104, 1300–1306.
35. H. Imahori, S. Hayashi, H. Hayashi, A. Oguro, S. Eu, T. Umeyama and Y. Matano, Effects of Porphyrin Substituents and Adsorption Conditions on Photovoltaic Properties of Porphyrin-Sensitized TiO₂ Cells, J. Phys. Chem. C, 2009, 113, 18406–18413.
36. M. K. Nazeeruddin, R. Humphry-Baker, D. L. Officer, W. M. Campbell, A. K. Burrell and M. Grätzel, Application of Metalloporphyrins in Nanocrystalline Dye-Sensitized Solar Cells for Conversion of Sunlight into Electricity, Langmuir, 2004, 20, 6514–6517.
37. Z.-S. Wang, N. Koumura, Y. Cui, M. Takahashi, H. Sekiguchi, A. Mori, T. Kubo, A. Furube and K. Hara, Hexylthiophene-Functionalized Carbazole Dyes for Efficient Molecular Photovoltaics: Tuning of Solar-Cell Performance by Structural Modification, Chem. Mater., 2008, 20, 3993–4003.
38. (a) A. J. Mozer, P. Wagner, D. L. Officer, G. G. Wallace, W. M. Campbell, M. Miyashita, K. Sunahara and S. Mori, The Origin of Open Circuit Voltage of Porphyrin-Sensitised TiO₂ Solar Cells, Chem. Commun., 2008, 4741–4743; (b) X. Chen, C. Jia, Z. Wan, J. Feng and X. Yao, Effects of Different Solvent Baths on the Performances of Dye-Sensitized Solar Cells: Experimental and Theoretical Investigation, Org. Electron., 2014, 15, 2240–2249.
39. C. Redmond and D. Fitzmaurice, Spectroscopic Determination of Flatband Potentials for Polycrystalline TiO₂ Electrodes in Nonaqueous Solvents, J. Phys. Chem., 1993, 97, 1426–1430.
40. B. Enright, C. Redmond and D. Fitzmaurice, Spectroscopic Determination of Flat Band Potentials for Polycrystalline TiO₂ Electrodes in Mixed Solvent Systems, J. Phys. Chem., 1994, 98, 6195–6200.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra23650b

‡ These authors contributed equally to this work.

---