EXAFS, ab Initio Molecular Dynamics, and NICIS Spectroscopy

Article

EXAFS, ab initio Molecular Dynamics and NICIS Spectroscopy Studies on an Organic Dye Model at the Dye-sensitized Solar Cell Photoelectrode Interface Peng Liu, Viktor Johansson, Herri Trilaksana, Jan Rosdahl, Gunther G Andersson, and Lars Kloo ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 23 May 2017 Downloaded from http://pubs.acs.org on May 29, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

EXAFS, ab initio Molecular Dynamics and NICIS Spectroscopy Studies on an Organic Dye Model at the Dye-Sensitized Solar Cell Photoelectrode Interface Peng Liu,1 Viktor Johansson,1 Herri Trilaksana,2 Jan Rosdahl,1 Gunther G. Andersson,2 and Lars Kloo1∗ 1 Applied

Physical Chemistry, Center of Molecular Devices, Department of Chemistry, School of Chemical Science and Engineering, KTH-Royal Institute of Technology, SE10044 Stockholm, Sweden

2 Flinders

Centre for Nano Scale Science and Technology, Flinders University, PO Box 2100, Adelaide SA 5001, Australia

Keywords: EXAFS, ab initio molecular dynamics, NICIS, dye sensitized solar cell

ABSTRACT The organization of dye molecules in the dye layer adsorbed on the semiconductor substrate in dye-sensitized solar cell has been studied using a combination of theoretical methods and experimental techniques. The model system is based on the simple D-π-A dye L0, which has been chemically modified by substituting the acceptor group –CN by –Br (L0Br) in order to offer better X-ray contrast. Experimental EXAFS data based on the Br K-edge backscattering show no obvious difference between dye-sensitized titania powder and titania film samples, thus allowing model systems to be based on powder slurries. Ab initio molecular dynamic (aiMD) calculations have been performed to extract less biased information from the experimental EXASF data. Using the aiMD calculation as input, the EXAFS structural models can be generated a priori that match the experimental data. Our study shows that the L0Br dye adsorbs in the trans-L0Br configuration and that adsorption involves both a proximity to other L0Br dye molecules and the titanium atoms in the TiO2 substrate. These results indicate direct coordination of the dye molecules to the TiO2 surface in contrast to previous results on metal-organic dyes. The molecular coverage of L0Br on mesoporous TiO2 was also estimated using NICIS spectroscopy. The NICISS results emphasized that the L0Br dye on nanoporous titania mainly forms monolayers with a small contribution of multilayer coverage.

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

INTRODUCTION Energy and environment issues have evolved to become the main challenges for a sustainable social development in the coming decades. Due to the finite fossil energy resources and anthropogenic perturbation of the atmospheric carbon dioxide levels, green and sustainable energy sources are urgently required. The (essentially) infinite and clean solar energy available makes the conversion of solar light into energy-rich substances (solar fuels) or electricity (solar cells) the most promising solution for the future global energy supply. The dye-sensitized solar cells (DSSCs) emerged as a realistic future solar cell alternative through the seminal report by O’Regan and Grätzel in 19911. Since then, a new research field has bloomed and diverged.2-4 Until now, the record efficiency of dye-sensitized solar cells is >14%, based on a co-photosentization strategy.5 In DSSCs, the light harvesting material is a dye which is assumed to be chemically attached to a semiconductor substrate. The charge separation is achieved through injection of the light-excited electrons into the conduction band of the semiconductor substrate, and a transport of the remaining holes in the dye molecules via liquid electrolyte redox couples or hole-transport materials. Therefore, the dye molecules have a very central role in DSSCs. Initially, commonly used dyes were based on Ru coordination centers 6-8. Later, organic dyes were developed offering higher molar extinction coefficients and a potential lower cost with respect to large-scale production. Among those, the D-π-A type of dye is so far the most promising for highly efficient DSSCs.9-10 This type of dye consists of three main components: a donor part (D), a linker part (π, typically a conjugated system) and an acceptor part (A) that acts not only as electron withdrawing unit but also as the expected anchoring group to the semiconductor substrate. Also, other typical dyes, such as perylenes or porphyrines, have been adopted into the D-π-A type for applications in DSSCs.11-12 The morphology of sensitizers adsorbed on the titania mesoporous films has also significant effects on the performance of the DSSC devices.13-15 However, the exact nature of sensitizer adsorption has so far not been fully clarified. In the early days of DSSCs, the assumption that sensitizers self-assemble into a monolayer on the metal-oxide (typically titania) surface via chemisorption was widely accepted.16-18 The formation of monolayers, involving a mixture of coordination models by the dye acceptor group, was thought to cause full coverage of dye molecules. It is notable that these assumptions are all based on modelling of indirect experimental studies and must be regarded as wishful thinking. Recently, previous studies in our collaborating groups have shown that the adsorption mechanism of the extensively used N719 dye in fact is best described as a multilayer coverage, using depth profiling of a direct experimental technique


Page 2 of 19

Page 3 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


based on neutral impact collision ion spectroscopy (NICISS).19 In that study, it was also shown that the dye molecules do not cover the entire titania surface even at the maximum coverage level (according to the Langmuir isotherm model), where the dye molecules form multilayer islands surrounded by non-covered surface, instead of perfect monolayers.20 In addition, a recent AFM study shows around 50 nm size of molecular islands in Z907 sensitized semiconductors.21 These results are in good agreement with our combined photovoltaic and NICISS studies on the dye Z90720. The electrostatic and non-covalent interaction of dye molecules with respect to both adjacent molecules and titania can also lead to the formation of dye aggregates.22-24 In this work, we used the model dye L0Br as sensitizer, which is a modified analogue to the simplest D-π-A dye, L0 (Figure 1). The acceptor group –CN has been substituted by –Br in order to enhance the contrast in both EXAFS and NICISS experiments. Thus, the introduced bromine atom (in the L0Br dye) will work as a “spy” atom in the experiments.25 It should be highlighted that any structural model suitable for modelling the adsorption of the L0Br dye molecules to the TiO2 surface will be quite complex. The normal methods of structural analysis based on least-squares fitting of models to the experimental data will thus require many parameters and as a consequence, the reliability of the models proposed will be weak. In this work, we instead use ab initio Molecular Dynamics (aiMD) to generate a priori EXAFS traces for realistic structural models for direct comparison with experimental data (without any model fitting). The sensitivity of this approach offers a higher degree of credibility to the conclusions drawn and limits the problem of overinterpretation of EXAFS data. As seen below, full understanding of the systems studied requires a combination of theoretical models and experimental techniques. The introduction of the –Br unit also a photoisomerization phenomenon in the D-π-A dye molecules based on a cyanoacrylic acceptor group could be revealed and was carefully investigated using a series of spectroscopy experiments25. These results also had consequences for how the current experiments were performed to avoid cis/trans complications from photoisomerization.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1. The molecular structures of L0 (left) and L0Br (right).

EXPERIMENTAL DETAILS

SECTION

AND

COMPUTATIONAL

Chemicals The L0Br dye was purchased from Dyenamo. The DSL 18NR-T paste was purchased from Dyesol. Other chemicals were purchased from Sigma-Aldrich. All chemicals were used without further purification. L0Br was throughout the experiments kept in the dark in order to avoid photoisomerization.

X-ray Absorption Measurements and Data Analysis Extended X-ray Absorption Fine Structure (EXAFS) measurements were performed at the Stanford Synchrotron Radiation Laboratory (SSRL) operating at 3.0 GeV and 50-100 mA current, and the Br K-edge X-ray absorption EXAFS data were collected at room temperature in transmittance and fluorescence mode on beamline 4.1. A Si(220) double crystal monochromator was used. For slurry samples, the L0Br sensitized TiO2 (Degussa P25) nanoparticles were investigated as a thick slurry in acetonitrile. For the DSSC electrode model samples, a 1 µm (+/- approximately 0.5 µm) thick dye-sensitised TiO2 film on an FTO glass substrate was obtained by doctor blading analogously to how photoelectrodes for normal DSSCs are made (DSL 18NR-T paste from Dyesol). The L0Br dye was used as sensitizer in both types of samples.


Page 4 of 19

Page 5 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The acquired data was processed and analyzed using the Athena and Artemis programs,26 in which the FEFF6 and IFEFFIT codes are implemented.27 The raw EXAFS data were analyzed using the SixPACK program package. The model EXAFS traces were generated using the FEFF program. The standard model EXAFS equation is shown below: ܵ଴ଶ ܰ௝ ݂௝ ሺ݇ሻ݁ ିଶ௞ ‫ݔ‬ሺ݇ሻ = ෍ ܴ݇௝ଶ ௝

మ ఙమ ೕ

sinൣ2ܴ݇௝ + ߜ௝ ሺ݇ሻ൧

where Nj is the number of neighboring atoms, Rj is the distance to the neighboring atoms, fj(k) is the effective scattering amplitude, σj2 is the Debye-Waller factor (DW) and S02 is the electronic reduction factor.

Ab initio Molecular Dynamic Calculations All ab initio molecular dynamics (aiMD) simulations were performed using the canonical ensemble thermostat (NVT) and 6-311G basis sets as implemented in the TeraChem code.28 The TiO2 cluster (Ti56O112) used as surface model consisted of 5 layers, of which the uppermost two layers were non-tethered. Effects of solvent was included for one system using the PCM formalism modelling acetonitrile. The total simulation time was 10 ps, using a step size of 2 fs. The first ps was used for thermal equilibration at 298K, while the remaining 9 ps were used for calculating the a priori theoretical EXAFS traces of the model systems. For the qualitative analysis, sets of 200 randomly chosen configurations for each system were used to calculate an average EXAFS trace.

NICISS NICISS is a technique for determining the concentration depth profiles of the elements at soft matter surfaces up to a depth of about 20 nm and with a depth resolution of a few Ångström in the near-surface area. NICISS has been employed extensively for determining the elemental concentration depth profiles at soft matter surfaces.29 This ion-spectroscopic technique utilizes ionized helium atoms as projectiles and detects the backscattered neutral helium atoms.30 The energy of the projectiles backscattered from the atoms in the target is determined by their time of flight (ToF) from the target to the detector. The projectiles loose energy during the backscattering process, and the energy transfer depends on the mass of the target atom. This first type of energy loss is used to identify the element from



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

which a projectile is backscattered. Additionally, the projectiles continuously loose energy on their trajectory through the adsorbed bulk film due to small-angle scattering from and electronic excitations of the molecules constituting the target. This energy loss, known as stopping power, is used to determine the depth of the atom from which the projectile is backscattered. In combination, these two types of energy losses are used to determine the concentration depth profiles of the elements. A typical NICIS spectrum consists of energy-loss spectra for each element constituting the sample. The shape of the spectrum obtained for a single element is determined by its concentration depth profile in the sample. The depth resolution of NICIS spectra is mainly influenced by small-angle scattering of the projectile. Small-angle scattering of the projectile is least in the top several Å of the surface, leading to a depth resolution of a few Å. The spectra are also influenced by the distribution of inelastic energy losses during the backscattering process and the straggling of the energy loss caused by lowangle scattering and electronic excitations.31 The inelastic energy loss during the backscattering process can be determined from the NICIS spectra of the respective elements in the gas phase. The energy-loss straggling of low energy projectiles has been determined experimentally.32 The inelastic energy loss during backscattering and the energy-loss straggling of projectiles in the bulk is known experimentally, and thus it is possible to deconvolute the measured profiles to obtain more details of the concentration depth profiles. The deconvolution is performed with a program based on a genetic algorithm.30 In the genetic algorithm the result of the deconvolution is independent of the starting conditions of the program and the statistical error of all suitable solutions can be used to determine the error bars of the deconvoluted profile. In order to determine the energy of a projectile backscattered from an element in the outermost layer of the sample, NICIS spectra of the elements in a low density gas jet are employed.30

Materials and Sample Preparation for NICISS The experiments were based on nanoporous titania analogous to those used in the DSSCs. The mesoporous layer of TiO2 was prepared using the DSL 18NR-T paste from Dyesol (same as for the film samples in the EXAFS experiments). The titania nanoparticles are predominantly of anatase structure33 and films were prepared on ITO glass substrates by doctor blading. The samples were heated in an oven for 30 minutes at (450 ± 20)◦C. The dye molecules (sensitizers) were adsorbed onto the surface of the metal oxide, by immersing the samples into a dye solution for 24 hours at the respective dye concentration directly after the sintering process. The


Page 6 of 19

Page 7 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


range of dye solution concentration was from 0.1 mM down to 0.0025 mM, and each dye solution was freshly prepared. The sample was rinsed by methanol after it was removed from the solution. The rinsing is needed to remove L0Br excess from the mesoporous surface.

Spherical Shape Nature of the Titania Particles The shape of the surface, i.e., the spherical shape of the titania particles, needs to be taken into account in the data analysis. Obviously, such a surface is not microscopically flat. Because of the particle curvature the helium ions will travel a longer path through the dye layer than representing the actual thickness of this layer when impacting the surface at any point that is not at the center of a particle. Consequently, the spectra will show bromine seemingly located at a depth greater than the thickness of the dye layer. The concentration depth profile of the layer formed by the dye molecules on the titania nanoparticles can be derived by relating the length of the projectile trajectory to the depth in the dye layer, as shown in the referrences.20 The presence of bromine in the L0Br molecule enables NICISS to communicate information about the concentration depth profile of L0Br more accurately.25



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RESULTS AND DISCUSSION Experimental EXAFS Data

Figure 2. Experimental EXAFS data based on two different substrates.

Figure 2 shows the k3-weighted Br-K edge EXAFS data for the two different types of samples. The data for sensitized TiO2 slurries show a distinct oscillation pattern in the whole k-range studied. However, the TiO2 film samples display a very noisy response in the high k-range. In general, the two types of samples offer the same scattering information except of a slight shift of the peaks to lower k-values for the slurry samples. Those results indicate that the local environment in both types of samples is essentially analogous. The main structural analysis will therefore be made on the results from the slurry samples.


Page 8 of 19

Page 9 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3. Three different model systems for the aiMD simulations: (a) DyeTiO2 (Model A), (b) Dimers (Model B), and (c) Stacking (Model C). Here, the L0Br dye molecules have the trans configuration.

Figure 3 shows three configurations used as model systems for the aiMD simulations. Model A represents one L0Br dye molecule adsorbed to a (100) surface of a TiO2 slab (Figure 3a). Model A was also investigated including acetonitrile solvent effects (Figure S4), but the effects of including the solvent were negligible. The L0Br-dimer configuration, Model B, is based on two dimers formed by four L0Br molecules (Figure 3b), while the L0Br-packed configuration, Model C, is based on four parallel L0Br molecules (Figure 3c). All L0Br molecules were modelled in the trans configuration. In this paper, we used the configurations obtained from the aiMD simulations as the structural model input for calculations of a priori model EXAFS traces.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Ab initio Molecular Dynamics Models of EXAFS Results

Figure 4. Direct comparison between experimental and modelled EXAFS traces based on the different models (a) Model A, (b) Model B, (c) Model C, (d) Model A+B and (e)

Model A+C

Figure 4 shows a direct comparison between experimental and modelled EXAFS traces. In particular, Figure 4 a, b, and c show the outcome of the three individual structural models, in comparison with the experimental data. It is clear from the comparisons in Figure 4 that a single structural model will not generate an EXAFS trace that fully matches the experimental one. Thus, combination of models need to be considered. All model traces have been achieved by linear combinations. All models considered based on different degrees of freedom have been summarized in the Supporting Information. According to these results, an increase in the number of degrees of freedom cannot significantly improve the resemblance to the experimental traces. Therefore, we choose to use a combination of two models as our preferred model. The Figures 4d and 4e show the best models when only two degrees of freedom are considered. The good match between the theoretical and experimental data indicates that those two models are equally good for explaining the experimental observations. The main results from the comparison of theoretical and experimental results show that both Br-Ti and Br-Br correlations dominate the EXAFS traces, and that offers three more general conclusions. Firstly, the majority of L0Br molecules must have other L0Br molecules in the proximity, and thus they must be aggregated either in the form of a self-assembled


Page 10 of 19

Page 11 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


monolayer on the TiO2 surface (SAM; Model C) or as multilayer islands (Model B). Secondly, the fact that a simple linear combination of Model A and Model B, or Model A and Model C, describes the experimental data quite well gives an indication that a significant part of the L0Br molecules also are close to the TiO2 surface. Thirdly, combining the two conclusions above it seems likely that the L0Br molecules form a SAM. However, this indication has been further investigated using NICISS (vide infra). In this context it should be mentioned that a linear combination of the Model B (trans) and Model B (cis) also would describe the experimental EXAFS trace in an adequate way. However, such a model can be excluded on the grounds of both the NICISS results described below and the careful exclusion of light preventing trans-to-cis photoisomerization during the experiments.

Figure 5. Direct comparison between experimental and modelled EXAFS traces based on analogous structural models as in Figure 3 but now with the L0Br dye in cis configuration (a) Model A, (b) Model B, (c) Model C, (d) Model A+B and (e) Model A+C.

A previous study shows that photoisomerization may have significant consequences for the dye-sensitized solar cells, and therefore we need to also investigate how structural data would be affected if the L0Br dye molecules photoisomerizes into a cis configuration25. It should be noted that all experiments were performed without exposing the L0Br samples to light in order to minimize spurious effects from photoisomerization. Figure 5 shows the direct comparison between experimental and modelled EXAFS traces based on dye molecules in the cis-L0Br configuration. It is clear that if the L0Br molecules were



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

photoisomerized to the cis configuration to a larger extent, it would have become clearly visible in the EXAFS traces of the samples. The main conclusion is that L0Br has remained in the trans configuration (in which form it is synthesized) throughout the experiments. The results in Figure 5 also have a more indirect implication: the combined experimental and theoretical approach, generating a priori EXAFS traces with a marginal fitting procedure, to model EXAFS data is highly sensitive and would highlight unwanted effects that an analysis based on least-squares fitting of a structural model possibly would not.

NICIS Spectroscopy NICIS spectra of the L0Br dye adsorbed onto the mesoporous titania sample is shown in the supplementary section in Figure S1. Features due to the presence of the main elements constituting the samples, such as carbon, oxygen, titanium, and bromine, can clearly be observed in the spectra. In general, the coverage of the titania with dye can be determined from the concentration depth profiles of bromine in L0Br, as well as the thickness and structure of the dye layer. The procedure taking the spherical shape of the substrate into account has been applied to the spectra because of the porous nanostructured nature of the substrate. The procedure has been described in detail in previous papers.19,20

Figure 6. Concentration depth profiles of bromine of L0Br deconvoluted and corrected for the spherical shape nature of the substrate. The profiles of L0Br layers adsorbed from 0.0025 mM, 0.03 mM, 0.08 mM and 0.1 mM solutions are shown.


Page 12 of 19

Page 13 of 19

In Figure 6, the concentration depth profiles of bromine in the L0Br dye layers on titania are shown. The dye layers are adsorbed from dye solutions with four different concentrations. The concentration depth profiles were deconvoluted for the energy resolution of NICISS and corrected for the spherical shape nature of the substrate. The concentration depth profiles show a narrow distribution of bromine of about 2 Å. However, the uncertainty of the deconvolution procedure has to be taken into account and the full-width-half maximum (FWHM) could be as broad as 5 Å, which corresponds to an average thickness of the dye layer of not more than a single monolayer. In this context, it should be noted that bromine, in the dye anchoring group, is expected to reside within a few Ångströms to the TiO2 surface if directly adsorbed. Low dye concentrations at a depth >7 Å can be interpreted in terms of a small contribution from multilayers as well. Based on the concentration depth profiles shown above, the fraction of the surface covered with multilayers is less than 20%. The threshold of 7 Å for separating monolayer and multilayer is based on approximating a single L0Br molecule as ellipsoid with the long axis being ~ 13 Å and the short axis ~ 7 Å. Approximating the L0Br molecule is chosen this way this way in order not to underestimate the multilayer contribution. Approximating the L0Br molecule as a sphere would result in a diameter of ~ 10 Å. The multilayer contribution then would be less than 10%. The surface coverage of the L0Br dye on the nanoporous TiO2 has been determined by integrating the bromine concentration depth profile. The molecular coverage is plotted as a function of the L0Br dye concentration in the dye baths used is shown in Figure 7.

8

molecular coverage x 10-10(mol cm-2)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7 6 5 4 3 2 1 0

0

0.02

0.04

0.06

0.08

concentration (mM)

0.1

Figure 7. Adsorption isotherm of the L0Br molecule coverage from 0.1 down to 0.0025 mM concentrations in the dye bath used on nanoporous titania. The fitted curve is only intended to guide the eyes.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

In general, the plot of dye coverage increases with increasing dye bath concentration, and it reaches a maximum coverage at around 4x10-10 mol cm-2. This corresponds to a dye solution concentration range of 0.07 to 0.1 mM. At the lowest concentration, about 0.0025 mM, the coverage is around 1.0 x 10-10 mol cm-2 The plot of coverage as a function of logarithmic concentration in Figure 7 shows that a finite detectable coverage exists even at the lowest concentration investigated. In addition, at higher concentrations (>0.07 mM) two data points with higher coverage level (red dot point) were recorded. The concentration depth profiles of these measurements was distinctly different to the other depth profiles showing higher concentrations at a depth >15 Å. These concentration depth profiles show agglomeration of the dye molecules. None of the preparation conditions could be identified as the cause for this observed agglomeration of the L0Br dye. The concentration depth profiles determined via NICISS not only are used to determine the amount of dye adsorbed to the titania of the shape of the concentration depth profiles informs whether the dye adsorbs in mono- or multilayers.19

CONCLUSIONS The properties of the dye-semiconductor interface have been investigated using EXAFS combined with aiMD calculations and NICISS. aiMD theoretical and the experimental data adequately match when direct TiO2 adsorption and aggregation models of trans-L0Br are combined, and the main conclusion is that L0Br adsorbs into a SAM in contrast to results from photoelectrodes dyes with metal-organic dyes. Determination of the molecular coverage on L0Br on the mesoporous TiO2 substrate using NICISS was carried out, and after data analysis the concentration depth profile shows that the L0Br dye adsorbed on nanoporous titania predominantly exist in the form of a monolayer with a small contribution of multilayers. The range of dye coverage is between 1x10-10 mol cm-2 and 4x10-10 mol cm-2 corresponding to between 1.6 and 0.4 nm2 per L0Br molecule. In comparison, a perfect monolayer dye coverage is estimated to 0.5 nm2 per molecule (aiMD results). The NICISS results strengthen the combined EXAFS/aiMD results in terms of the formation of a SAM of L0Br on the titania surface, and NICISS further communicates that the coverage of the titania surface is close 100% in the isotherm plateau region indicating that organic dyes, such as L0Br, should affect both the titania energy levels and recombination loss reaction rates in dye-sensitized solar cells.


Page 14 of 19

Page 15 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ASSOCIATED CONTENT Supporting Information Detailed analysis of the NICISS data. This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author [email protected] Notes The authors declare no competing financial interest.

ACKNOWLEDGEMENTS The Swedish Research Council, the Swedish Energy Agency, the Chinese Scholarship Council (CSC) and Knut and Alice Wallenberg Foundation are gratefully acknowledged for their financial support. Use of the Stanford Synchrotron Radiation Lightsource, SLAC National Accelerator Laboratory, is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Contract No. DE-AC02-76SF00515. One of the authors – H. Trilaksana - acknowledges the support through a scholarship from the Indonesian Government.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

REFERENCES (1) O'Regan, B.; Grätzel, M. A Low-Cost, High-Efficiency Solar Cell Based on DyeSensitized Colloidal TiO2 Films, Nature 1991, 353, 737-740. (2) Grätzel, M. Dye-Sensitized Solar Cells, J. Photochem. Photobiol., C 2003, 4, 145-153. (3) Hagfeldt, A.; Boschloo, G.; Sun, L.; Kloo, L.; Pettersson, H. Dye-Sensitized Solar Cells, Chem. Rev. 2010, 110, 6595-6663. (4) Hardin, B. E.; Snaith, H. J.; McGehee, M. D. The Renaissance of Dye-Sensitized Solar Cells, Nat Photon 2012, 6, 162-169. (5) Kakiage, K.; Aoyama, Y.; Yano, T.; Oya, K.; Fujisawa, J.-i.; Hanaya, M. HighlyEfficient Dye-Sensitized Solar Cells with Collaborative Sensitization by Silyl-Anchor and Carboxy-Anchor Dyes, Chem. Commun. 2015, 15894-15897. (6) Nazeeruddin, M. K.; Zakeeruddin, S. M.; Humphry-Baker, R.; Jirousek, M.; Liska, P.; Vlachopoulos, N.; Shklover, V.; Fischer, C.-H.; Grätzel, M. Acid−Base Equilibria of (2,2`-Bipyridyl-4,4`-Dicarboxylic Acid)Ruthenium(Ii) Complexes and the Effect of Protonation on Charge-Transfer Sensitization of Nanocrystalline Titania, Inorg. Chem. 1999, 38, 6298-6305. (7) Nazeeruddin, M. K.; Pechy, P.; Renouard, T.; Zakeeruddin, S. M.; HumphryBaker, R.; Comte, P.; Liska, P.; Cevey, L.; Costa, E.; Shklover, V.; Spiccia, L.; Deacon, G. B.; Bignozzi, C. A.; Gratzel, M. Engineering of Efficient Panchromatic Sensitizers for Nanocrystalline TiO2-Based Solar Cells, J. Am. Chem. Soc. 2001, 123, 1613-1624. (8) Nazeeruddin, M. K.; De Angelis, F.; Fantacci, S.; Selloni, A.; Viscardi, G.; Liska, P.; Ito, S.; Bessho, T.; Gratzel, M. Combined Experimental and DFT-TDDFT Computational Study of Photoelectrochemical Cell Ruthenium Sensitizers, J. Am. Chem. Soc. 2005, 127, 16835-16847. (9) Hagberg, D. P.; Edvinsson, T.; Marinado, T.; Boschloo, G.; Hagfeldt, A.; Sun, L. A Novel Organic Chromophore for Dye-Sensitized Nanostructured Solar Cells, Chem. Commun. 2006, 2245-2247. (10) Kloo, L. On the Early Development of Organic Dyes for Dye-Sensitized Solar Cells, Chem. Commun. 2013, 49, 6580-6583. (11) Wu, S.-L.; Lu, H.-P.; Yu, H.-T.; Chuang, S.-H.; Chiu, C.-L.; Lee, C.-W.; Diau, E. W.G.; Yeh, C.-Y. Design and Characterization of Porphyrin Sensitizers with a Push-Pull Framework for Highly Efficient Dye-Sensitized Solar Cells, Energy Environ. Sci. 2010, 3, 949-955. (12) Imahori, H.; Umeyama, T.; Ito, S. Large Π-Aromatic Molecules as Potential Sensitizers for Highly Efficient Dye-Sensitized Solar Cells, Acc. Chem. Res. 2009, 42, 1809-1818. (13) Lu, H.-P.; Tsai, C.-Y.; Yen, W.-N.; Hsieh, C.-P.; Lee, C.-W.; Yeh, C.-Y.; Diau, E. W.G. Control of Dye Aggregation and Electron Injection for Highly Efficient Porphyrin Sensitizers Adsorbed on Semiconductor Films with Varying Ratios of Coadsorbate, J. Phys. Chem. C 2009, 113, 20990-20997.


Page 16 of 19

Page 17 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(14) Labat, F.; Le Bahers, T.; Ciofini, I.; Adamo, C. First-Principles Modeling of DyeSensitized Solar Cells: Challenges and Perspectives, Acc. Chem. Res. 2012, 45, 12681277. (15) Lee, K. E.; Gomez, M. A.; Elouatik, S.; Demopoulos, G. P. Further Understanding of the Adsorption Mechanism of N719 Sensitizer on Anatase TiO2 Films for DSSC Applications Using Vibrational Spectroscopy and Confocal Raman Imaging, Langmuir 2010, 26, 9575-9583. (16) Katono, M.; Bessho, T.; Meng, S.; Humphry-Baker, R.; Rothenberger, G.; Zakeeruddin, S. M.; Kaxiras, E.; Grätzel, M. D-Π-a Dye System Containing CyanoBenzoic Acid as Anchoring Group for Dye-Sensitized Solar Cells, Langmuir 2011, 27, 14248-14252. (17) Krüger, J.; Bach, U.; Grätzel, M. Modification of TiO2 Heterojunctions with Benzoic Acid Derivatives in Hybrid Molecular Solid-State Devices, Adv. Mater. 2000, 12, 447-451. (18) Hwang, K.-J.; Shim, W.-G.; Jung, S.-H.; Yoo, S.-J.; Lee, J.-W. Analysis of Adsorption Properties of N719 Dye Molecules on Nanoporous Tio2 Surface for DyeSensitized Solar Cell, Appl. Surf. Sci. 2010, 256, 5428-5433. (19) Ellis-Gibbings, L.; Johansson, V.; Walsh, R. B.; Kloo, L.; Quinton, J. S.; Andersson, G. G. Formation of N719 Dye Multilayers on Dye Sensitized Solar Cell Photoelectrode Surfaces Investigated by Direct Determination of Element Concentration Depth Profiles, Langmuir 2012, 28, 9431-9439. (20) Johansson, V.; Ellis-Gibbings, L.; Clarke, T.; Gorlov, M.; Andersson, G. G.; Kloo, L. On the Correlation between Dye Coverage and Photoelectrochemical Performance in Dye-Sensitized Solar Cells, Phys. Chem. Chem. Phys. 2014, 16, 711-718. (21) Ni, J.-S.; Hung, C.-Y.; Liu, K.-Y.; Chang, Y.-H.; Ho, K.-C.; Lin, K.-F. Effects of Tethering Alkyl Chains for Amphiphilic Ruthenium Complex Dyes on Their Adsorption to Titanium Oxide and Photovoltaic Properties, J. Colloid Interface Sci. 2012, 386, 359-365. (22) Pastore, M.; De Angelis, F. Aggregation of Organic Dyes on TiO2 in DyeSensitized Solar Cells Models: an ab initio Investigation, ACS Nano 2010, 4, 556-562. (23) Sirohi, R.; Kim, D. H.; Yu, S.-C.; Lee, S. H. Novel Di-Anchoring Dye for DSSC by Bridging of Two Mono Anchoring Dye Molecules: A Conformational Approach to Reduce Aggregation, Dyes and Pigments 2012, 92, 1132-1137. (24) Otsuka, A.; Funabiki, K.; Sugiyama, N.; Yoshida, T.; Minoura, H.; Matsui, M. Dye Sensitization of ZnO by Unsymmetrical Squaraine Dyes Suppressing Aggregation, Chem. Lett. 2006, 35, 666-667. (25) Zietz, B.; Gabrielsson, E.; Johansson, V.; El-Zohry, A. M.; Sun, L.; Kloo, L. Photoisomerization of the Cyanoacrylic Acid Acceptor Group - a Potential Problem for Organic Dyes in Solar Cells, Phys. Chem. Chem. Phys. 2014, 16, 2251-2255. (26) Ravel, B.; Newville, M. Athena, Artemis, Hephaestus: Data Analysis for X-Ray Absorption Spectroscopy Using Ifeffit, J. Synchrotron Radiat. 2005, 12, 537-541. (27) Newville, M. EXAFS Analysis Using Feff and Feffit, J. Synchrotron Radiat. 2001, 8, 96-100. (28) Ufimtsev, I. S.; Martinez, T. J. Quantum Chemistry on Graphical Processing Units. 3. Analytical Energy Gradients, Geometry Optimization, and First Principles Molecular Dynamics, J. Chem. Theory Comput. 2009, 5, 2619-2628.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(29) Andersson, G.; Ridings, C. Ion Scattering Studies of Molecular Structure at Liquid Surfaces with Applications in Industrial and Biological Systems, Chem. Rev. 2014, 114, 8361-8387. (30) Andersson, G.; Morgner, H. Impact Collision Ion Scattering Spectroscopy (ICISS) and Neutral Impact Collision Ion Scattering Spectroscopy (NICISS) at Surfaces of Organic Liquids, Surf. Sci. 1998, 405, 138-151. (31) Andersson, G.; Morgner, H. Determining the Stopping Power of Low Energy Helium in Alkanethiolates with Neutral Impact Collision Ion Scattering Spectroscopy (NICISS), Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 1999, 155, 357-368. (32) Andersson, G. Energy-Loss Straggling of Helium Projectiles at Low Kinetic Energies, Phys. Rev. A 2007, 75, 032901. (33) Gottesman, R.; Haltzi, E.; Gouda, L.; Tirosh, S.; Bouhadana, Y.; Zaban, A.; Mosconi, E.; De Angelis, F. Extremely Slow Photoconductivity Response of CH3NH3PbI3 Perovskites Suggesting Structural Changes under Working Conditions, J. Phys. Chem. Lett. 2014, 5, 2662-2669.


Page 18 of 19

Page 19 of 19

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


T OC


EXAFS, ab Initio Molecular Dynamics, and NICIS Spectroscopy

Recommend Documents