Probing the Highly Efficient Electron Transfer Dynamics between Zinc

Aug 18, 2016 - Electron injection from the dye to the semiconductor is very fast (ket ≈ 1011 s–1), much faster than previously reported rates. The...
3 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/JPCA

Probing the Highly Efficient Electron Transfer Dynamics between Zinc Protoporphyrin IX and Sodium Titanate Nanosheets Sudipta Biswas,† Debdyuti Mukherjee,‡ Swati De,*,† and Arunkumar Kathiravan*,§ †

Department of Chemistry, University of Kalyani, Kalyani 741235, Nadia, West Bengal, India Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore 560012, India § National Centre for Ultrafast Processes, University of Madras, Taramani Campus, Taramani, Chennai 600113, Tamil Nadu, India ‡

S Supporting Information *

ABSTRACT: Sodium titanate nanosheets (NaTiO2 NS) have been prepared by a new method and completely characterized by TEM, SEM, XRD, EDX, and XPS techniques. The sensitization of nanosheets is carried out with Zn protoporphyrin IX (ZnPPIX). The emission intensity of ZnPPIX is quenched by NaTiO2 NS, and the dominant process for this quenching has been attributed to the process of photoinduced electron injection from excited ZnPPIX to the nanosheets. Time resolved fluorescence measurement was used to elucidate the process of electron injection from the singlet state of ZnPPIX to the conduction band of NaTiO2 NS. Electron injection from the dye to the semiconductor is very fast (ket ≈ 1011 s−1), much faster than previously reported rates. The large two-dimensional surface offered by the NaTiO2 NS for interaction with the dye and the favorable driving force for electron injection from ZnPPIX to NaTiO2 NS (ΔGinj = −0.66 V) are the two important factors responsible for such efficient electron injection. Thus, NaTiO2 NS can serve as an effective alternative to the use of TiO2 nanoparticles in dye sensitized solar cells (DSSCs).

1. INTRODUCTION

In this work, we report on the synthesis of sodium titanate nanosheets (NaTiO2 NS) via the hydrothermal method. However, NaTiO2 NS do not absorb visible light, and hence sensitization is necessary if they are to be considered for future use as potential solar cell materials. Dye sensitization is one of the most promising pathways for utilizing visible light for solar cells.26−28 Dye sensitization in DSSCs mostly arises due to fast electron transfer (ET) from visible-light-excited chromophores of the dye to the conduction band of TiO2, in which strong coupling between the dye and TiO2 is necessary. The most important factors that influence the efficiency of the solar cells are the dynamics of the electron injection (from the excited state of the dye to the conduction band of TiO2) and electron recombination (from injected electrons to the ground state of the dye) processes.29 Several studies in the past have addressed the issue of interfacial electron dynamics in dye/TiO2 systems.29−33 The former process initiates the electrical circuit in the solar cell, and its quantum yield determines the initial population of electrons, while the later process decreases the efficiency of electron collection. As the processes occur on a time scale from femtoseconds (electron injection) to nanoseconds (recombination),34 ultrafast time-resolved spectro-

Photochemistry of semiconductor electrodes is one of the most important topics in fields such as solar energy conversion.1−3 Titanium dioxide is strongly favored and has been investigated intensively since Fujishima and Honda discovered water photolysis on TiO2 semiconductor electrodes.4 Recently, nanosized TiO2 materials have attracted growing attention because of their potential applications in photocatalysis5−7 and in dye-sensitized solar cells (DSSCs).8,9 Since the first report10 on TiO2 nanotubes which were synthesized via the hydrothermal process, there has been intensive research on the growth mechanism and crystalline structure of one-dimensional nanosized TiO211,12 and various titanate nanomaterials.13−16 These have potential applications in many fields such as photocatalysis,13solar energy conversion17 and as potential anode materials for Na-ion batteries.18 Recently, there have been many reports on the synthesis and characterization of one-dimensional alkali metal titanate nanomaterials owing to their size and shape-dependent optical properties.19,20 Sodium titanates are interesting photoactive materials which often possess extended sheet structures made from TiO6 octahedra, sharing edges with each other and forming a zigzag ribbon structure in the layers with Na+ cations residing between the layers.21,22 Among various alkali metal titanate nanomaterials, sodium titanate is of particular interest for its wide applications.23−25 © 2016 American Chemical Society

Received: June 23, 2016 Revised: August 6, 2016 Published: August 18, 2016 7121

DOI: 10.1021/acs.jpca.6b06363 J. Phys. Chem. A 2016, 120, 7121−7129

Article

The Journal of Physical Chemistry A

The autoclave was heated for 10 h at 180 °C. The final concentration of TTEAIP was 5 × 10−3 M. 2.3. Characterization Techniques. The surface morphology of the materials was analyzed by scanning electron microscopy (SEM) (Carl Zeiss Ultra 55) equipped with energy dispersive X-ray spectroscopy (EDS). Transmission electron microscopy (TEM) images were obtained using a JEOL 2100F microscope operating at 200 kV. The TEM samples were prepared by drop coating the diluted dispersion on carboncoated copper grids 300 mesh and dried under vacuum for 12 h. X-ray diffraction (XRD) patterns were recorded using Philips (PAN analytical) instrument (Cu Kα radiation). UV−vis spectra were obtained using PerkinElmer (Lambda 750) double beam UV−vis−NIR spectrophotometer. X-ray photoelectron spectroscopic analysis was performed on a Kratos Axis Ultra DLD X-ray photoelectron spectrometer with monochromatic Al Kα (1486.708 eV) radiation. 2.4. Fluorescence and Transient Absorption Setups. The absorption spectra of the samples were recorded using an Agilent 8453 UV−visible diode array spectrophotometer. The fluorescence spectral measurements were carried out using Fluoromax-4 spectrophotometer (Horiba Jobin Yvon). For fluorescence studies, very dilute solutions were used to avoid spectral distortion due to the inner-filter effect and emission reabsorption. Time resolved picosecond fluorescence decays were obtained by the time-correlated single-photon counting (TCSPC) technique with microchannel plate photomultiplier tube (Hamamatsu, R3809U) as detector and a femtosecond laser as an excitation source. The second harmonic (400 nm) output from the mode-locked femtosecond laser (Tsunami, Spectra Physics) was used as the excitation source. The instrument response function for the TCSPC system is ∼50 ps. The data analysis was carried out using the software provided by IBH (DAS-6), which is based on the deconvolution technique using nonlinear least-squares method. Femtosecond fluorescence transients have been collected using the fluorescence up-conversion technique. In the femtosecond up-conversion setup used (FOG 100, CDP, Russia) the sample was excited using the second harmonic (400 nm) of a mode-locked Ti-sapphire laser (Tsunami, Spectra Physics). The fundamental beam (800 nm) was frequency doubled in a nonlinear crystal (1 mm BBO, θ = 25°, ϕ = 90°) and used for the excitation. The sample was placed inside a 1 mm thick rotating quartz cell. The fluorescence emitted from the sample was up-converted in a nonlinear crystal (0.5 mm BBO, θ = 38°, ϕ = 90°) using the fundamental beam as a gate

scopic techniques are generally used to evaluate the time constants of these processes. In the present work, time-resolved spectroscopy has been employed to study the interaction between a dye and NaTiO2 NS. Considering the efficiency shown by porphyrin derivatives as sensitizers for DSSCs,35−39 a Zn protoporphyrin IX (ZnPPIX) dye was chosen for this work (Scheme 1). The Scheme 1. Structure of Zn Protoporphyrin IX (ZnPPIX)

conjugated tetrapyrrole chromophore is responsible for the visible light absorption. This molecule has two carboxyl groups which may act as anchoring points for the Ti-based NS. Previously Akatsuka et. al37 had studied the photoelectrochemical properties of alternative multilayer films composed of Ti0.91O2 nanosheets and a Zn porphyrin derivative. In the present work NaTiO2 nanosheets have been used. Moreover, they prepared the Ti0.81O2 nanosheets by delamination of a layered protonic titanate,37 whereas in this work NaTiO2 was synthesized by a purely bottom-up approach starting from a titanium(IV) precursor.

2. MATERIALS AND METHODS 2.1. Materials. 30% hydrogen peroxide (H2O2) was obtained from Merck, India. Titanium(IV) triethanolaminatoisopropoxide (TTEAIP) and Zn protoporphyrin IX (ZnPPIX) were obtained from Sigma-Aldrich. Isopropyl alcohol was from Spectrochem, India, and sodium hydroxide was from Merck, India. 2.2. Synthetic Procedure. First a mixture of 10 mL of 30% H2O2 and 6.7 mL of NaOH (1 M) aqueous solution was prepared. Then to this mixture 25.3 μL of TTEAIP in 3.3 mL of isopropyl alcohol was added dropwise and the total mixture was then transferred to a Teflon-lined stainless steel autoclave.

Figure 1. (a) Absorption spectrum of the as-prepared TiO2 based nanomaterials and (b) the corresponding plot for band gap calculation. 7122

DOI: 10.1021/acs.jpca.6b06363 J. Phys. Chem. A 2016, 120, 7121−7129

Article

The Journal of Physical Chemistry A

From the deconvoluted spectra for Ti 2p, it is observed that Ti is in lower oxidation state, which is Ti3+ (as Ti3+ shows 2p3/2 peak at around 455 eV as is observed here in Figure 3 (a). Now, the approximate composition of the surface can be determined by dividing the individual peak areas, after appropriate background subtraction, by their respective atomic sensitivity factor (ASF).40 It is known that ASF values for Ti 2p, Na 1s, and O 1s are 1.8, 2.3, and 0.66, respectively. From the deconvoluted data shown in Figure 3a−c, and Table 1, we get the atomic ratio of

pulse. The upconverted light was dispersed in a monochromator and detected using photon counting electronics. The instrument response function of the apparatus is 300 fs. Transient absorption experiments were carried out using nanosecond laser flash photolysis (Applied Photophysics, U.K.). The third harmonic (355 nm) of a Q-switched Nd:YAG laser (Quanta-Ray, LAB 150, Spectra Physics, USA) with 8 ns pulse width and 150 mJ pulse energy was used to excite the samples. Samples were purged with argon gas for about 45 min prior to the laser irradiation. The transients were probed using a 150 W pulsed xenon lamp, a Czerny−Turner monochromator, and Hamamatsu R-928 photomultiplier tube as detector. The transient signals were captured with an Agilent infiniium digital storage oscilloscope, and the data were transferred to the computer for further analysis.

Na:Ti:O ≈ 1:1:2

Hence the probable formula of the nanomaterial is NaTiO2, where Ti is in +3 oxidation state.40 Summarizing, powder XRD pattern and XPS data provide concrete evidence that the nanomaterial synthesized is NaTiO2. To further determine the size/shape of the nanomaterials, TEM analysis was carried out. From the TEM images (Figure 4) it is clear that NaTiO2 nanosheets are formed. Now usually titanate materials exist as nanotubes. Sometime, during the detailed sequence of events leading to titanate nanotubes, titanate nanosheets are formed under alkaline conditions. Zhang et al.41 explained that single layered nanosheets experience an asymmetric chemical environment due to an imbalance of H+ and Na+ on two different surfaces of the sheet. This results in bending of the nanosheets to nanotubes to minimize the excess surface energy. However, for multilayered nanosheets, this problem can be overcome. The fact that we could stabilize the nanosheet morphology prompted us to explore further, i.e., whether single-layered or multilayered nanosheets are formed. Thus, to assess the layer morphology of the NaTiO2 nanosheets, field emission scanning electron microscopy (FESEM) was performed. Figure 5 shows the FESEM image of the as-prepared sodium titanate nanosheets. It shows the interconnected network formation which finally results in the formation of nanosheet. FESEM images confirm the existence of multilayered titania nanosheet structure, hence explaining their stability. Energy dispersive X-ray spectroscopy (EDX) was employed to ascertain the composition of the as-synthesized nanosheets. The EDX analysis [Supporting Information Figure Sa] indicates the presence of Na, Ti, and O, proving that the as-synthesized nanomaterials are composed of NaTiO2 only. The EDX analysis results correspond almost exactly to those obtained by XPS. 3.2. Sensitization of NaTiO2 Nanosheets with Zinc Protoporphyrin IX (ZnPPIX). 3.2.1. Absorption Spectra. The absorption spectrum of ZnPPIX exhibits an intense Soret band absorption at 421 nm (S0 → S2 transition), together with two weaker Q-bands at 545 and 582 nm (S0 → S1 transition) [Figure Sb in Supporting Information]. Typically, porphyrin molecules are very prone to form aggregates in solution.42 Hence, to understand the aggregation behavior of ZnPPIX in solution itself, we have measured the absorption spectrum of ZnPPIX at different concentrations in THF solvent by using absorption spectroscopy. As can be seen from Figure Sb in Supporting Information, while increasing the concentration of ZnPPIX, the intensity of absorbance increases but no new bands are observed. At the same time, Beer−Lambert law was also obeyed for ZnPPIX in the concentration range from 1 to 5 μM showing that the ZnPPIX is not significantly aggregated within this concentration range. In contrast, Akatsuka et. al37

3. RESULTS AND DISCUSSION 3.1. Characterization of the Nanomaterials. The absorption spectrum of the as-synthesized TiO2 based nanomaterials is shown in Figure 1 a. The absorption spectrum showed a strong and narrow absorption in the UV region. The corresponding Kubelka−Munk plot (Figure 1 b) yielded a band gap (Eg) value of 3.02 eV. Thus, the NaTiO2 NS produced in this work have a significantly lower Eg value than the Ti0.91O2 nanosheets produced by Akatsuka et. al37 (Eg ≈ 3.8 eV). This large Eg value had led to a significant decrease in the electron transfer efficiency from Ti0.91O2 nanosheets to organic dyes such as ZnTMPyP4+.37 The powder XRD pattern of the as-prepared TiO2 based nanomaterials is shown in Figure 2. It reveals peaks at 2θ =

Figure 2. Powder XRD pattern of the as-prepared TiO2 based nanomaterials.

16.40° and 33.16°. These correspond to the [003] and [006] plane of NaTiO2 (sodium titanate). There are several minor peaks, e.g., at 2θ = 36.05°, 40.56°, 57.15°, 60.5° and 71.52°, which correspond to the [012], [104], [018], [110], and [021] planes of NaTiO2, respectively. The data are consistent with the standard powder X-ray diffraction pattern (as given by the Joint Committee on Powder Diffraction Standards, JCPDS). This confirms that the synthesized TiO2 material is NaTiO2 not TiO2 as such. To further confirm this, the X-ray photoelectron spectra (XPS) of the samples were recorded. The XPS of the samples corresponding to Ti 2p, O 1s, and Na 1s are shown in Figure 3. Table 1 lists the peak positions, area, and full-widthat-half-maximum (fwhm) of the XPS. 7123

DOI: 10.1021/acs.jpca.6b06363 J. Phys. Chem. A 2016, 120, 7121−7129

Article

The Journal of Physical Chemistry A

Figure 3. XPS of the NaTiO2 nanomaterials corresponding to (a) Ti 2p, (b) O 1s, and (c) Na 1s.

band absorption is important.35 As can be seen from Figure 6, the absorption spectrum of ZnPPIX adsorbed onto NaTiO2 NS is broad and the peaks positions are not altered compared to the absorption maxima of ZnPPIX alone. This clearly indicates that there is no aggregation of ZnPPIX on the NaTiO2 NS surface. It is to be noted that peak broadening is always observed when a dye is attached onto a semiconductor nanoparticle surface.36 3.2.2. Steady State Fluorescence. Figure 7 shows the fluorescence spectrum of ZnPPIX in THF and with NaTiO2 NS (2 g/L). The fluorescence spectrum of ZnPPIX in THF shows two peaks at 590 and 640 nm (for λex = 400 nm). These bands can be attributed to the two S1 → S0 vibronic transition. However, in the presence of NaTiO2 NS, the fluorescence of ZnPPIX is completely quenched (Figure 8ii). This fluorescence quenching can occur due to many reasons: due to energy transfer from dye to NaTiO2 NS, due to dye to dye energy

Table 1. Peak Position, Area, and Full-Width at HalfMaximum (fwhm) of the XPS of the NaTiO2 Nanomaterials Na 1s Ti 2p O 1s

peak position (eV)

area

fwhm (eV)

1070.191 459.60 455.10 530.24

13406 4150 6282 8767

1.91 3.00 2.70 1.95

had reported significant J-aggregation of ZnTMPyP4+ dye on the surface of Ti0.91O2 nanosheets. The extinction coefficient of ZnPPIX is 2.5 × 105 M−1 cm−1 at 421 nm. Figure 6 shows the absorption spectrum of 2 × 10−4 M ZnPPIX in absence and presence of sodium titanate nanosheets (NaTiO2 NS). We have measured the absorption spectrum in the wavelength range 500−700 nm, i.e., Q-band region, because for use of protoporphyrins as photosensitizers in DSSCs, the Q-

Figure 4. (a) TEM image of as- prepared sodium titanate nanosheets. (b) Corresponding HRTEM image. 7124

DOI: 10.1021/acs.jpca.6b06363 J. Phys. Chem. A 2016, 120, 7121−7129

Article

The Journal of Physical Chemistry A

Figure 5. SEM images of as-prepared sodium titanate nanosheets: (a) lower resolution; (b) higher resolution.

Figure 6. Absorption spectra of 2 × 10−4 M ZnPPIX in the (i) absence and (ii) presence of NaTiO2 nanosheets.

Figure 8. Time resolved fluorescence decays of ZnPPIX in the (i) absence and (ii) presence of NaTiO2 nanosheets; (iii) is the instrument response function (IRF).

Figure 7. Fluorescence spectra of ZnPPIX in THF in the (i) absence and (ii) presence of NaTiO2 nanosheets; λex = 400 nm.

excluded. Also the absorption and fluorescence spectra [Figure Sb in Supporting Information and Figure 7] of the dye conform to that reported in the literature. Hence there is no probability of self-quenching by the dye. We therefore conclude that the fluorescence quenching is solely due to electron transfer from ZnPPIX to NaTiO2 NS. Moreover, the excited state oxidation potential of ZnPPIX (−1.93 V) is more negative than the conduction band (CB) potential of NaTiO2 NS (−1.27 V),37 ensuring a positive driving force for electron injection from the excited ZnPPIX to NaTiO2 NS (Scheme 2). In order to substantiate this conclusion, time-correlated single-photon counting (TCSPC) measurements were performed. 3.2.3. Time Resolved Fluorescence Studies. TCSPC is a convenient method to elucidate the electron injection rates in dye/semiconductor systems,43 since the only species that emits in the dye/semiconductor systems is the dye singlet state, which is formed after absorption of a photon. As we know, dye singlet states can inject electrons into the conduction band of semiconductor and these electrons are responsible for the photocurrent.37 Therefore, it is important to measure the rate of electron injection. Previous studies involving transient absorption spectroscopy showed that the rate of electron injection is characterized by nonexponential kinetics.43 Alternatively, the time-resolved fluorescence technique can also be used to study the electron injection dynamics and quite often the results obtained from time-resolved fluorescence measurements are in good agreement with the transient absorption results.44 Thus, it is established that one can

transfer, self-quenching due to aggregation, or due to electron transfer from the excited singlet state of ZnPPIX to the NaTiO2 NS. There is no spectral overlap between the emission spectrum of ZnPPIX and the absorption spectrum of NaTiO2 NS, so energy transfer between ZnPPIX and NaTiO2 NS cannot be responsible for the fluorescence quenching of ZnPPIX by NaTiO2 NS. A competition between dye to dye energy transfer and electron transfer may be responsible. However, since the absorption maximum of the dye does not vary too much when ZnPPIX is adsorbed on NaTiO2 NS, in all probability dye aggregation does not occur; hence dye to dye energy transfer or self-quenching due to dye aggregation can be 7125

DOI: 10.1021/acs.jpca.6b06363 J. Phys. Chem. A 2016, 120, 7121−7129

Article

The Journal of Physical Chemistry A Scheme 2. Schematic Energy Level Diagram for Electron Injection Process

measure the electron injection rate precisely using timeresolved fluorescence. In this work, we have employed the same technique to characterize the ZnPPIX/NaTiO2 NS system. The measured time-resolved fluorescence decays of ZnPPIX in THF and on NaTiO2 NS are shown in Figure 8. The fluorescence decays have been deconvoluted from the instrument response function. The excitation wavelength (λex) is 400 nm and probing wavelength (λem) is 590 nm for both samples, since the fluorescence emission maximum (λem) of ZnPPIX/NaTiO2 NS is similar to that for free ZnPPIX and the absorption maximum does not vary too much when the ZnPPIX is adsorbed onto NaTiO2 NS. The fluorescence decay of ZnPPIX in THF can be fit to a single exponential function [F(t) = A exp(−t/τ)], and τ value for ZnPPIX was found to be 2 ns. This is assigned to the singlet (Sl) lifetime of the excited state of ZnPPIX. For efficient performance of a dye in DSSCs, the essential requirement is that electron injection from the dye to the semiconductor should be very fast (0.1−10 ps) compared to the excited state lifetime of the dye (1−20 ns). In this case the excited state lifetime for ZnPPIX being 2 ns, the situation may be favorable for efficient electron injection to the NaTiO 2 NS conduction band. This has earlier been theoretically predicted to be true by the energy level diagram shown in Scheme 2. For the ZnPPIX/NaTiO2 NS system, the observed fluorescence decay is completely quenched when compared to ZnPPIX in neat solvent under the same experimental conditions (Figure 8). This indicates that very fast electron injection takes place from the S1 state of ZnPPIX to NaTiO2 NS conduction band. Earlier reports have indicated that such quenching of dye fluorescence is due to fast electron transfer from dye to semiconductors.45−47 The fluorescence decay of ZnPPIX in the presence of NaTiO2 NS is very fast compared to the employed instrument response, i.e., 50 ps, and thus cannot be resolved by the TCSPC method. Hence, femtosecond fluorescence upconversion has to be employed to measure the electron injection kinetics for ZnPPIX/NaTiO2 NS. Figure 9 depicts the fluorescence decays of ZnPPIX in THF and on NaTiO2 NS suspension (measured using a fs up-conversion setup in a 60 ps time window upon excitation at 400 nm). The emission was monitored at λem = 590 nm. Figure 9a shows the femtosecond fluorescence decay of ZnPPIX in THF solution. The decay is not complete within the measured window (250 ps), and this is expected because we know from TCSPC that the fluorescence decay time of ZnPPIX is 2 ns. Then the fluorescence decay dynamics of ZnPPIX adsorbed onto NaTiO2 NS was measured

Figure 9. Time-resolved fluorescence decay of ZnPPIX in THF measured in a femtosecond fluorescence up-conversion setup (a) in the absence of NaTiO2 NS and (b) in the presence of NaTiO2 NS.

to gain insight on the rate of interfacial electron transfer between ZnPPIX and NaTiO2 NS. Figure 9b shows the femtosecond fluorescence decay of ZnPPIX adsorbed onto NaTiO2 NS. By comparing this to the decay of the dye in absence of NaTiO2 NS (Figure 9a), we note that the decay of ZnPPIX in the presence of NaTiO2 NS is very fast and is completed within the 60 ps time scale. This fast decay is assigned to electron injection from the S1 state of ZnPPIX to NaTiO2 NS. Previous workers have also assigned such fast decays (