TiO₂-based photoanodes modified with GO and MoS₂ layered materials

Abstract

MoS₂/TiO₂ and GO/TiO₂ nanocomposites were synthesized via environmentally friendly electrodeposition (GO, MoS₂) and – for the purpose of comparison – hydrothermally (MoS₂). The modified Hummers' method was used to form GO from expanded graphite. A hydrothermal process of MoS₂ preparation from Na₂MoO₄ and CH₄N₂S with the use of a surfactant or NH₂OH·HCl and NH_3(aq) was applied. The prepared powders were characterized by means of X-ray diffraction, Raman spectroscopy, and scanning electron microscopy. Both GO and MoS₂ were found to form 2D layered materials. Electrochemical deposition of two-dimensional compounds on the surface of TiO₂ was conducted with the use of a suspension of 2D MoS₂ or GO in KNO₃ (pH close to neutral) at 1.2 V for 40 s. Hydrothermal conditions were applied for MoS₂ deposition as well. The morphology and photoelectrochemical properties of GO- and MoS₂-modified TiO₂ photoanodes were studied, and measurements using electrochemical impedance spectroscopy were performed. Lower charge transfer resistance as well as a significant enhancement of photoelectrochemical response were confirmed for MoS₂/TiO₂ nanocomposites in comparison to TiO₂. The highest photocurrent was achieved for 2D/TiO₂ prepared with the use of MoS₂, characterized by a well-defined microstructure and higher crystallinity. Hydrothermally modified photoanodes were found to be stable under photoelectrochemical measurement conditions.

---

1. Introduction

Graphene-like layered structures are currently one of the most sought-after class of materials. This group includes the following compounds: graphene,^1,2 graphene-like metal dichalcogenides (LMD: MoS₂, WS₂, MoSe₂, WSe₂),³ black phosphorus,⁴ and hexagonal boron nitride (BN).⁵ They can play a key role in leading scientific fields, such as chemistry, materials science, solid state and condensed matter physics, molecular electronics, spintronics, and tribology. This is because of their unique chemical, electrical and mechanical properties, which can be easily tuned by varying the synthesis conditions. In addition, their layered structure provides an abundance of reactive sites, which are particularly desirable in photo-induced reactions such as photocatalysis or hydrogen generation due to photoelectrochemical water splitting.^2,6

In the field of photochemical water decomposition, there are at least three factors that affect the performance of a semiconducting photoelectrode in a photoelectrochemical cell (PEC). These factors are: a suitable band-gap energy, which assures a wide range of absorption of light, a limited recombination process of photo-generated carriers, and, finally, immunity to corrosion and photocorrosion in water-based electrolytes. Titanium dioxide is a well-known non-corrosive, wide-band-gap semiconductor that is stable in aqueous electrolytes, and is considered to be the best candidate for a photoelectrode material. The disadvantage of absorption limited to the UV range can be overcome in many ways, e.g. by combining TiO₂ with layered materials to form photo-active composites/heterojunctions with the intended properties. Such materials are highly desirable due to the exceptionality of their chemical, electrical and mechanical properties, which can be widely tuned. Their leading representative is graphene oxide (GO), which can be easily formed into controlled two-dimensional layered structures. What is more important, a material from this group can act as a support for the semiconductor. Sun et al. showed that the electron transfer rate between TiO₂ and the graphene sheet depends on the dimensions of the semiconductor.⁷ 0D nanoparticles, 1D nanotubes, and 2D nanosheets of TiO₂ were supported on graphene. The authors suggested that 2D–2D nanocomposites are characterized by faster charge separation.

Graphene oxide (GO) is one of the more common derivatives of materials from the afore-mentioned class, and it can be most effectively synthesized using the modified Hummers' method.⁸ Another interesting layered material is molybdenum disulfide (MoS₂). Various approaches can be applied for MoS₂ preparation:^6,9,10 mechanical, chemical or electrochemical exfoliation via lithium intercalation, thermal decomposition of molybdenum salts, CVD, sulfurization of Mo or its thin film oxide, and chemical reactions under hydrothermal conditions. Some of them do not allow pure and uncontaminated material to be synthesized. Contamination with non-stoichiometric molybdenum oxide and molybdenum trisulfide is very common. In some cases, heavy chemical treatment with compounds such as hydrogen sulfide (H₂S) is required. However, the hydrothermal method seems to yield pure, single-phase MoS₂ without using any harsh chemicals in the synthesis.¹¹

The choice of how GO and MoS₂ as well as GO/TiO₂ and MoS₂/TiO₂ heterojunctions are to be obtained is a separate issue. In the case of graphene oxide, different methods can be applied: liquid phase deposition (LPD),¹² colloidal blending,¹³ physical mixing,¹⁴ combustion,¹⁵ microwave-assisted combustion,¹⁵ self-assembly method,¹⁶ and sol–gel.¹⁷ For MoS₂ preparation, one of the following approaches can be used: hydrothermal method,¹⁸ self-assembly method,¹⁸ physical mixing,¹⁹ mechanochemical method,²⁰ and photodeposition.²¹ Designing the procedure of deposition of a GO or MoS₂ layer on the surface of TiO₂ can also pose a problem. Among the many ways in which GO/TiO₂ can be obtained, the following ones may be mentioned: thermal hydrolysis of a Ti source in the presence of a solution containing GO,²² spin-coating,¹⁶ photodeposition,²³ or impregnation.²⁴ In the case of MoS₂/TiO₂, the following methods may be used: the doctor blade technique,²⁵ atomic layer deposition (ALD),²⁶ photodeposition,²⁷ and chemical bath deposition.²⁸

Electrodeposition seems to be an interesting routine for both GO and MoS₂ deposition. This approach is well-known and has been used successfully with graphene oxide,^29–31 but not as effectively in the case of MoS₂. Attempts made so far with the use of both aqueous and non-aqueous precursors of molybdenum have produced unsatisfactory results. Ghosh et al. performed DC electrolysis of a mixture of sodium molybdate and sulfide in a two-electrode system (Pt) on different metallic substrates.³² The deposited films were found to be amorphous and contaminated with MoO₂. Even though annealing at elevated temperatures increased crystallization significantly, molybdenum dioxide remained. Ponomarev et al. electrodeposited thin films from ammonium tetrathiomolybdate and Na₂S solutions on a Mo foil and conductive glass in a three-electrode cell (Pt, SCE).³³ The prepared layers were found to be amorphous and required annealing, but the presence of MoO_xS₂ (x = 0.2–0.6) and MoO_3−y (where y < 1) was subsequently confirmed. Maijenburg et al. used a formamide solution containing ammonium tetrathiomolybdate, potassium chloride, and ammonium chloride to electrodeposit MoS₂ on nanocubes of Au in a three-electrode system (Pt, Ag/AgCl).³⁴ Due to the fact that the prepared layers were annealed in an Ar atmosphere at 400 °C, it may be presumed that the samples had an amorphous structure. Murugesan et al. made some efforts to perform electrodeposition on glassy carbon from ionic liquids (PP₁₃-TFSI) in combination with molybdenum glycolate in a three-electrode cell (Pt, QRE).³⁵ As a result, a pure and stoichiometric MoS₂ layer was obtained. However, the process was carried out at an elevated temperature ranging from 50 to 100 °C. Shariza et al. also used electrodeposition at a slightly elevated temperature (40 °C) from a mixture of molybdic acid and sodium thiosulfate on conductive glass and stainless steel in a three-electrode system (graphite, SCE).³⁶ Nevertheless, MoS₂ was very poorly crystallized. There were also some attempts to use electropolymerization as a modification of this method, but – yet again – amorphous materials were obtained, and they were additionally contaminated with MoS₃.³⁷ It can therefore be concluded that electrodeposition with the use of Mo and S precursors does not provide a simple way to obtain a pure and well-crystallized MoS₂ layer. That is why it is necessary to choose an approach in which the electrochemical process is carried out with the use of a previously prepared powder. There are only several examples of such attempts. Güler et al. deposited MoS₂ powder on stainless steel with the use of a suspension in a Watts bath.³⁸ Likewise, Kanagalasara et al. tried to electrodeposit MoS₂ powder in a bath based on sulfate and chloride.³⁹ In both cases, a commercial powder was used and homogeneous layers were obtained with clearly visible particles of molybdenum disulfide. Nonetheless, a highly acidic medium considered harmful to the environment was applied in both cases.

The aim of the presented work was to prepare graphene oxide and molybdenum disulfide and deposit them on the surface of titanium dioxide with the use of a rarely employed electrochemical deposition procedure. The proposed method of MoS₂ electrodeposition is novel, environmentally friendly, and does not require toxic reagents and solvents. For comparison, the hydrothermal approach was also applied in the case of MoS₂. The influence of the deposition technique on the electro- and photoelectro-chemical properties of GO/TiO₂ and MoS₂/TiO₂ was studied.

2. Experimental

2.1. Materials

Na₂MoO₄·2H₂O (ACS reagent), CH₄N₂S (pure), PEG-1000, NH₂OH·HCl (analytically pure, ACS reagent), NH_3(aq) (25%, analytically pure), C₂H₅OH (99.8%, analytically pure), H₂SO₄ (95%, analytically pure), KMnO₄ (95%, analytically pure), HCl (35–38%, pure), H₂O₂ (30%, analytically pure), C₃H₆O (analytically pure), C₃H₈O (analytically pure), C₂H₆O₂ (99.0% analytically pure), Na₂SO₄ (analytically pure), and KNO₃ were purchased from Avantor Performance Materials (Gliwice, Poland). NH₄F (≥98.0%, ACS reagent) and Ti foil (0.127 mm, 99.7%) were bought from Sigma-Aldrich. Argon (pure) was purchased from Air Liquide.

2.2. Synthesis of MoS₂ powders

MoS₂ powders were synthesized under different hydrothermal conditions with the use of different reactants. In the first approach, MoS₂ powder (marked “1M”) was obtained from the following reactants: Na₂MoO₄ (0.03 mol) as a molybdenum precursor, CH₄N₂S as a sulfur precursor, and PEG-1000 as a surfactant. The chosen molar ratio was 1 : 4.56 : 0.03, respectively. Appropriate amounts of reagents were dissolved in 200 ml of distilled water and magnetically stirred for 15 min with a speed of 500 rpm. Afterwards, the mixture was transferred into a Teflon-lined hydrothermal reactor and heated at 200 °C for 6, 12, 18, or 24 h. In the second method, the final product was marked “2M” and the following reactants were used: Na₂MoO₄ (0.06 mol) and CH₄N₂S as precursors, NH₂OH·HCl as a reductant, and NH_3(aq) as a pH adjuster. In a typical procedure, Na₂MoO₄ (1) and NH₂OH·HCl (2) in a molar ratio of 1 : 2 were dissolved separately in 100 ml of distilled water and magnetically stirred for 15 min. Mixture 2 was then added to mixture 1, transferred into a 250 ml round-bottom flask and heated at 90 °C under constant stirring. After that, the mixture was cooled to room temperature and the pH was adjusted to 6 with an appropriate amount of NH_3(aq). Subsequently, the mixture was heated again to 90 °C and 0.20 mol of CH₄N₂S was added. After 10 min, it was transferred into the Teflon-lined hydrothermal reactor. The process was carried out at 180 °C for 6, 12, 18, and 24 h. In both methods, the pressure was autogenous. Finally, the obtained materials were washed with a mixture of distilled water and absolute ethanol in proportions of 50 : 50 vol%, collected via filtration and dried at 60 °C for 12 h. All experimental conditions in which sample preparation was performed are collected in Table 1.

Table 1 Experimental conditions for the preparation of materials

GO powder
Precursors/sources	Method	Temperature (°C)	Time (h)
EG1000, KMnO4, H2SO4	Modified Hummers'	35	20

---

MoS2 powder
Name	Precursors/reagents	Temperature (°C)	Time (h)
1M6	Na2MoO4, CH4N2S, PEG-1000	200	6
1M12			12
1M18			18
1M24			24
2M6	Na2MoO4, CH4N2S, NH2OH·HCl, NH3(aq)	180	6
2M12			12
2M18			18
2M24			24

---

TiO2 substrate			Sensitization of TiO2
Type	Synthesis	Conditions	Type	Conditions/precursors	Name
Nanotubes (TiO2-NT)	Electrochemical oxidation of Ti foils	30 V, 60 min in two-electrode system (Ti, Pt) immersed in electrolyte (ethylene glycol + 1 wt% NH4F + 2 wt% H2O)	Electrochemical	1.2 V, 40 s 0.5 mg ml^−1 sensitizer in 50 ml of water + 200 ml of 1 M KNO3	GO/TiO2-NT
Layers (TiO2-L)	Thermal oxidation of Ti foils	600 °C, 8 h, air flow of 80 sccm	Electrochemical	1.2 V, 40 s 0.5 mg ml^−1 sensitizer in 50 ml of water + 200 ml of 1 M KNO3	1Me/TiO2-L
					2Me/TiO2-L
			Hydrothermal	200 °C, 12 h, the same precursors as for 1M	1Mh/TiO2-L
				180 °C, 12 h, the same precursors as for 2M	2Mh/TiO2-L

---

2.3. Synthesis of GO in the form of flakes

GO was synthesized with the use of a modified Hummers' method from thermally expanded graphite (at 1000 °C) – EG1000 – as a carbon source. First, 0.40 g of EG1000 and 0.01 mol of KMnO₄ were mixed together. Then, 30 ml of H₂SO₄ was gradually poured into the mixture, which was stirred for 30 min in an ice bath in order to avoid boiling. The mixture was kept at 35 °C in a water bath under stirring for another 20 h. Another dose of KMnO₄ (0.01 mol) was added under the same conditions and the mixture was again transferred into the water bath for 1.5 h. Afterwards, 20 ml of distilled water and 10 ml of H₂O₂ were added. The obtained material was sonicated for 30 min, washed successively in HCl, H₂O₂, and distilled water, and finally centrifuged (4000 rpm, 15 min). Finally, the dense gel-like dispersion was sonicated for 30 min and dried under vacuum at 50 °C.

2.4. Synthesis of TiO₂-L and TiO₂-NT

Two types of TiO₂ substrates were prepared: TiO₂ nanotubes (TiO₂-NT)⁴⁰ and TiO₂ layers (TiO₂-L).⁴¹ In both cases, titanium foils were oxidized after successive ultrasonic degreasing in the following media: tap water with detergent, tap water, distilled water, acetone, and isopropanol. TiO₂-L substrates were obtained by thermally oxidizing the Ti foils for 8 h in a tubular furnace at 600 °C and in an air flow of 80 sccm. TiO₂-NT substrates were obtained by means of anodization (electrochemical oxidation) at 30 V for 60 min, which was carried out in a two-electrode system (Ti foil – anode, Pt – cathode) immersed in an electrolyte composed of ethylene glycol, 1 wt% of NH₄F, and 2 wt% of H₂O. The obtained amorphous nanotubes were ultrasonically cleaned and annealed for 3 h at 450 °C and in an Ar flow of 100 sccm to enhance crystallinity.

2.5. Preparation of M/TiO₂-L and GO/TiO₂-NT

The deposition of MoS₂ on the surface of TiO₂-L (M/TiO₂-L) was performed either hydrothermally or electrochemically. In the hydrothermal approach, TiO₂-L was placed in a Teflon-lined reactor and MoS₂ was deposited on TiO₂-L during synthesis under the same conditions as those described above for the 1M and 2M powders. After that, the samples were rinsed with distilled water and dried in air. The resulting samples were marked “1Mh/TiO₂-L” or “2Mh/TiO₂-L” depending on the selected hydrothermal conditions.

Electrochemical deposition was conducted in a three-electrode system in which TiO₂-NT or TiO₂-L served as anodes, Pt as a counter electrode, and Ag/AgCl as a reference electrode. The electrolyte prepared for the deposition of MoS₂ was composed of a 50 ml dispersion of 0.5 mg ml⁻¹ MoS₂ (1M or 2M) in distilled water, which had been sonicated for 15 min, and 200 ml of 1 M KNO₃. MoS₂ was deposited from this mixture and onto a TiO₂-L surface by applying a constant DC voltage of 1.2 V for 40 s, under stirring. Finally, the electrodes were washed with distilled water and dried at room temperature. The resulting samples were marked “1Me/TiO₂-L” and “2Me/TiO₂-L” for 1M and 2M powders, respectively. GO/TiO₂-NT was obtained in much the same way, but by electrochemically depositing GO on the surface of TiO₂-NT.

2.6. Characterization techniques

The crystal structure of the obtained materials was investigated with the use of an X'Pert Pro diffractometer (PANalytical) over the angle range of 4.5–90°. Scanning electron microscopic observations were performed by means of a Nova NanoSEM 200 (FEI) with an EDX detector (EDAX). Raman spectra were registered with a Jobin-Yvon LabRam HR800 spectrometer with a green laser (λ = 532 nm). The electrical properties of photoanodes were studied by means of electrochemical impedance spectroscopy (EIS). Data were obtained using a Solartron 1296 dielectric interface and a Solartron 1260 frequency response analyzer. A frequency range of 1–10⁶ Hz was covered, with a 10 mV amplitude. The impedance spectra were analyzed using ZView software. An equivalent circuit containing a resistor (R), an inductor (L) and a constant phase element (CPE) was applied for the purpose of fitting. Experiments were performed in a custom-made two-electrode photoelectrochemical cell (Pt electrode and TiO₂-based electrode immersed in 0.8 M Na₂SO₄). The photoelectrochemical properties of materials were evaluated in a three-electrode PEC. PECs were composed of a photoactive anode, Pt wire, SCE electrode, and a 0.8 M Na₂SO₄ electrolyte. Current–voltage (I_ph–V_b) characteristics were measured in the dark and under illumination with a white light simulator. Flat-band potential (V_fb) was defined as the potential value at the onset of anodic photocurrent^42,43 and the error of design was of the order of 20 mV.

3. Results and discussion

SEM observations of graphite morphology (Fig. 1a) show that it is formed of flake-like elements that disappear after transformation into GO. The crystal structure of graphite and graphene oxide was analyzed based on XRD data presented in Fig. 1b. The diffraction pattern of graphite (EG1000) is characterized by the presence of three peaks at 26.5, 42.4, and 44.5°. These peaks are typical of hexagonal graphite (JCPDS-ICDD #03-065-6212). The chemical oxidation of EG1000 into graphene oxide is accompanied by changes in the diffraction pattern; the peaks originating from graphite disappear and in their place a peak at ∼11° appears, which is typical for graphene oxide. It is also evident that the full width at half maximum (FWHM) of the GO peak is much higher than in the case of EG1000, which indicates that the crystallites of the former are much smaller.

Fig. 1 SEM images (a), diffraction patterns (b), and Raman shift (c) of graphite and graphene oxide.

Raman spectroscopy was used to gain additional information about the structure; the obtained spectra of carbon-based materials are presented in Fig. 1c. According to the literature^2,44 there are only several Raman-active bands that can be identified for graphite-like materials. At about 1350 cm⁻¹, there is a D band due to first-order resonance and a primary in-plane vibrational mode G band is located at 1580 cm⁻¹, and a second-order overtone of a different in-plane vibration 2D band is located near 2700 cm⁻¹. The D band originates from structural imperfections.

A comparison of graphite and graphene oxide spectra shows a number of differences. First of all, all bands in the GO spectra are much less sharp, which indicates that it is crystallized with a lower degree of order than EG1000.⁴⁵ Secondly, the D bands have a much greater intensity in the GO spectra. The relative intensity of D and G bands (I_D/I_G) for EG1000 is equal to 0.07, while for GO it equals 0.96. This effect can be attributed both to the lower symmetry of GO, which is due to the presence of defects in its structure,⁴⁴ and much smaller crystallites than those in the case of EG1000.⁴⁶ In our previous paper,² we calculated the average crystalline size based on the intensity of G and D bands; it was found to be 281 nm for EG1000 and 20 nm for GO. Another interesting point is that for higher degrees of oxidation, the intensity of the G band not only increases but also gradually moves toward a higher wavenumber, whilst the D band moves towards a lower wavenumber.^47–49 The G band positions for graphene oxide and expanded graphite are 1600 cm⁻¹ and 1580 cm⁻¹, respectively, whilst the corresponding values for the D band are 1355 cm⁻¹ and 1350 cm⁻¹. Finally, in the GO spectra the D band is accompanied by two other disorder peaks – D + G at 2940 cm⁻¹ and 2D′ at 3195 cm⁻¹. In the second-order region (2300–3300 cm⁻¹), a so-called 2D band is also observed at ∼2700 cm⁻¹.⁵⁰

There are numerous parameters that can be taken into consideration during the analysis of Raman spectra. For example, the wavenumber corresponding to a band,^10,51–53 its integral^54,55 and relative intensities,⁵⁶ and changes in FWHM.⁵⁷ In the case of MoS₂ Raman spectra analysis, the first of these is commonly used due to the fact that the other parameters change arbitrarily,⁵⁸ and do not yield any conclusive information. Fig. 2 presents the Raman spectra of MoS₂ synthesized under hydrothermal conditions according to the procedures used to prepare samples 1M (a) and 2M (b). Typical bands originating from the layered structure are observed for bulk material at 383 and 408 cm⁻¹. The former one is marked E¹_2g and is attributed to in-plane optical vibration of the Mo and S atoms in the basal plane, while the latter – A_1g – is related to out-of-plane optical vibration of S atoms along the vertical axis.⁵² According to the literature, the difference between the E¹_2g and A_1g band positions decreases together with the decrease in the number of layers.^10,52–54 In the case of the studied materials, both bands are red-shifted in comparison with the literature data, i.e.: E¹_2g = 374–377 cm⁻¹ and A_1g = 403–405 cm⁻¹. However, similar observations were made by Yang et al. and Thripuranthaka et al. for MoS₂ measured under strain and at higher temperatures, respectively.^59,60

Fig. 2 Raman spectra of MoS₂ synthesized by means of methods 1M (a) and 2M (b).

The difference between E¹_2g and A_1g for 1M changes from 31 to 26 cm⁻¹ (±1 cm⁻¹) with increasing time synthesis, while for the 2M procedure it drops from 29 to 27 cm⁻¹ (±1 cm⁻¹). This suggests that for an extended time of synthesis, the number of layers decreases.

In addition to bands originating from MoS₂ vibrations, there are other bands present in the spectra. Bands at the following wavenumbers are visible: 148 cm⁻¹, which is attributed to – LA (M) (LA(M) – longitudinal acoustic mode, which is an oscillation for which wave polarization is parallel to wave propagation); 198, 237, and 285 cm⁻¹, attributed to the oxygen bond; 337 cm⁻¹, attributed to the bending of O–Mo–O oscillation; the origin of the band near 454 cm⁻¹ is still an open question; ∼659 cm⁻¹ indicates the A_1g(M) + LA(M) mode, and finally, 816 together with 994 cm⁻¹ represent stretching vibrations of the Mo=O bonds.⁶¹ According to Sen et al., such bonds between molybdenum and oxygen are possible when oxygen is adsorbed on the surface of partially non-stoichiometric MoS₂.⁶² It is also clear that when the time of hydrothermal synthesis increases, additional bands become less evident or disappear.

SEM images of MoS₂ powder are presented in Fig. 3a. As can be seen, particles of molybdenum disulfide grow in the shape of cauliflower-like nanostructures with flagella-like particles composed of cauliflower floret-like units. For both 1M and 2M powders it can be noticed that longer synthesis times lead to an increased size of particles. The change is, however, much more pronounced in the case of the 1M powder, i.e. from 14 to 33 μm, than for the 2M powder – from 12 to 17 μm. What can also be noticed is that almost every cauliflower floret-like element is built from nanoflagella. In the case of both methods, 6 h of synthesis is not sufficiently long for the flagella to develop fully. After 12 h, they are well-formed and prolonging the synthesis time further increases the diameter of the flagella from 56 to 84 nm for 1M. In the case of 2M diameter of flagella increases from 30 (6 h) to 97 nm (18 h) and decreases afterwards down to 50 nm (24 h).

Fig. 3 SEM images (a), diffraction patterns observed for MoS₂ powders prepared using the hydrothermal process: 1M (b) and 2M (c).

Diffraction patterns registered for the 1M and 2M powders are presented in Fig. 3b and c, respectively. All observed XRD patterns are typical for nanoscale MoS₂ (JCPDS-ICDD #01-089-2905), in which very broad peaks are observed. In the case of the 1M powder synthesized for 6 h (1M6), there are some additional peaks (marked with an asterisk) that remain unidentified. Increasing the hydrothermal reaction time leads to the gradual disappearance of nearly all of those peaks. For 2M powders, the only peaks that are observed originate from molybdenum disulfide. Aside from the presence of the unidentified peaks, all of these diffraction patterns are very similar. What is also important is that there is a significant shift of the (002) reflection from ∼14.4° towards a lower angle, down to the range of 9.5–10.7°. This effect is a direct consequence of an increased interplanar distance d₀₀₂.⁶³ For the 1M powder, the distance between layers varies in the range of 7.25–9.30 Å. In the case of the 2M powder, d₀₀₂ varies between 7.30 and 9.03 Å. These observations indicate the presence of few-layered MoS₂, which is consistent with the results obtained by means of Raman spectroscopy.

According to the literature,^63,64 an increased interplanar distance d₀₀₂ could be caused by the presents of surfactant ions and other ions or particles between the layers. On the other hand, in a study concerning WS₂, Liu et al. suggested that ammonia ions can enter the space between layers of WS₂ in the form of NH₄⁺ and cause increased d-spacing.⁶⁵

X-ray diffraction patterns of TiO₂ layers are presented in Fig. 4a. TiO₂ prepared via the process of electrochemical oxidation (TiO₂-NT) after annealing at 450 °C crystallizes in the tetragonal structure of anatase (JCPDS-ICDD #01-084-1286). TiO₂ prepared via thermal oxidation is composed of tetragonal rutile (JCPDS-ICDD #01-076-0321) with anatase as a minor phase and traces of titanium lower oxides – Ti₆O (JCPDS-ICDD #01-072-1807) and Ti₃O (JCPDS-ICDD #01-073-1583). Additionally, reflections originating from the Ti substrate are also present (JCPDS-ICDD #03-065-9622) in both samples.

Fig. 4 XRD patterns observed for TiO₂ substrates (a), and SEM images of unmodified and modified TiO₂-NT (b) and TiO₂-L (c).

Fig. 4b shows the SEM images of TiO₂ nanotubes before and after the deposition of graphene oxide. As can be seen, uniform nanotubes form in the process of electrochemical anodization. The average diameter of the nanotubes is ∼55 nm and tube length measured form cross-section images is ca. 740 nm. After the electrochemical deposition of GO, it can be noticed that the surface of TiO₂-NT is only partially covered with GO. Fig. 4c presents SEM images of the TiO₂ layer before and after the deposition of MoS₂. The surface of TiO₂-L is uniform and small grains are clearly visible. The thickness of the layer is ∼160 nm. It is apparent that the deposition of MoS₂ with the use of two different methods results in a drastically different morphology of the surface. The electrochemical deposition of molybdenum disulfide leads to slight changes in the SEM images. Semi-transparent flake-like objects appear; the surface of TiO₂-L can clearly be seen through them. The hydrothermal deposition of MoS₂ causes the growth of a dense layer, which is cracked, probably due to excessive thickness. A closer look reveals small irregular particles of MoS₂ for 1Mh/TiO₂-L and well-formed flower-like MoS₂ nanostructures with distinct flakes. It can be stated that the electrochemical deposition of MoS₂ allows a uniformly distributed layer to form.

The difference in the amount of MoS₂ deposited with the use of these two distinct methods is also evident from the results of EDX analysis. In the case of electrochemical deposition, only molybdenum is observed (sulfur is too light), the content of which varies in the range of 0.1–0.6 at%, while after hydrothermal deposition both elements are easily identified, with a Mo content of ∼18 at% and an S content of ∼38–46 at%.

The results of impedance spectroscopic measurements are presented in the form of Bode (Fig. 5) and Nyquist (Fig. 6) plots.

Fig. 5 Impedance spectra for anodes based on TiO₂ – Bode plots.

Fig. 6 Impedance spectra for anodes based on TiO₂ – Nyquist plot.

The complex admittance plots (Fig. 6) suggested that the experimental results might be approximated by circles with centers located below the Y′′ axis. The simplest equivalent circuit which gives such circles is composed of a resistor and a CPE (constant phase element) in series.² Additionally, an inductance element (L) connected in series is present due to the metallic conductivity of the wires between the electrodes and the measuring instrument. L assumes a constant value of 3.40–3.90 μH. The nature of a CPE associated with the space charge region in a semiconductor is defined by the equation:

Y_CPE = C(jω)ⁿ (1)

where C and n denote constants, j represents an imaginary unit, and ω stands for angular frequency.

A constant phase element is used instead of a simple capacitor, C (n = 1), to compensate for the non-ideal capacitive response of the interface. The parameter n of a CPE element (eqn (1)), takes values close to 1 for TiO₂ films. Its value of 0.79 for electrodes based on nanotubes indicates that the system is not homogeneous or that there is distribution of values related to physical properties. The possible explanations of the CPE behavior are: electrode roughness,⁶⁶ varying thickness or composition of the coating,⁶⁷ and non-homogeneous reaction rates on the surface.⁶⁸ The fitted values of electrical components are listed in Table 2.

Table 2 Parameters of an equivalent circuit (Fig. 6) fitted to the experimental data of the impedance spectrum^a

Sample	R [Ω]	R2 [Ω]	CPE1		CPE2
			n1	C1 (S s^−n1)	n2	C2 (S s^−n2)
a R – the resistance of the electrolyte and the TiO2 electrode. R2 – the resistance of the second component. CPE1 – a constant phase element related to the space charge region in a semiconductor.
TiO2-NT	31.74	—	0.79	6.21 × 10^−5	—	—
TiO2-NT/GO	31.45	38.98	0.84	1.20 × 10^−5	0.46	3.33 × 10^−4
TiO2-L	22.43	—	0.95	7.92 × 10^−7	—	—
1Mh/TiO2-L	18.05	368.9	0.81	2.27 × 10^−5	—	—
2Mh/TiO2-L	17.60	1565	0.85	9.68 × 10^−6	—	—
1Me/TiO2-L	18.59	—	0.90	1.19 × 10^−6	—	—
	R1 = 22.4	109.2
2Me/TiO2-L	16.90	—	0.91	1.10 × 10^−6	—	—
	R1 = 22.4	68.85

---

The impedance spectra of anodes are affected by the deposition of the second component of a composite. In the case of TiO₂ electrodes with the electrochemical deposition of MoS₂ and GO, the diameter of the semicircles increases. For an EIS function in the Y” = f(Y′) form, this indicates that the response is decreasing. On the other hand, the shape of the impedance spectrum does not change for MoS₂/TiO₂-L. This suggests that the equivalent circuit is more complex in this case. The spectra of GO/TiO₂-NT are described by a circuit containing an additional resistor (R₂) and CPE₂ connected in series; these elements are related to the second component of the composite. On the other hand, the impedance spectrum of the TiO₂ sample modified by means of the hydrothermal method is characterized by two time constants (see Fig. 5b plots for 1Mh/TiO₂-L and 2Mh/TiO₂-L). Additionally, one semicircle is observed in the first quadrant of Z′′ < 0, as shown in the inset of Fig. 6, which represents a simple slow electron transfer related to capacitance behavior. A different model was therefore adopted. The equivalent circuit employed for the analysis of the spectra of the samples featuring MoS₂/TiO₂-L composites fabricated via hydrothermal deposition is composed of a resistor (R₁) and inductor (L₁) connected in series with a parallel R₂-CPE₁ element. This can be explained by a number of phenomena, depending on the nature of the system being investigated. The electrical response of these electrodes involves the after-effect process. On the other hand, in the case of electrodes modified by means of electrochemical deposition, the responses are connected with instantaneous processes. The determined parameters are given in Table 2.

In order for a semiconductor to be applied as a photoanode in a PEC, it must meet certain requirements, which are schematically presented in Fig. 7a. One of these criteria is that the values of anodic current measured repetitively in the dark (the so-called dark current) must reach values equal or close to zero. In addition, the values of anodic current measured under illumination with white or monochromatic light should be significantly higher than those observed for dark current. In order to verify whether TiO₂ and TiO₂-based materials modified with MoS₂ and GO fulfill such criteria, current–voltage measurements in the dark were performed for all samples. Fig. 7b shows a comparison of current–voltage characteristics measured without illumination. Four curves overlap, while the remaining two are very distinct. Both TiO₂-NT and TiO₂-L as well as electrodes modified electrochemically with MoS₂ meet the requirements set for a good photoactive electrode in PEC. The current densities of the 1Mh/TiO₂-L and 2Mh/TiO₂-L anodes assume high values. This indicates that MoS₂-covered TiO₂ is electrochemically unstable. The hydrothermal modification does not provide suitable photoelectrochemical properties.

Fig. 7 Current–voltage characteristics of a TiO₂-L photoanode, as measured in a PEC (a), comparison of dark currents of the obtained samples (b), and flat-band potentials together with I_modified to I_TiO2 ratios (c).

As far as morphology is concerned, there are two key aspects in the process of preparing a heterojunction composed of two different semiconductors: sufficient contact must be ensured between the two, and the amount of each of them must be adequate. When these conditions are fulfilled, a synergistic effect is possible and, moreover, an effective separation of charge carriers between the two composite components takes place. Finally, this leads to an extended lifetime of carriers.^69–71

During the synthesis of MoS₂/TiO₂ composites with the use of the hydrothermal procedure, it was found that the deposited layer of MoS₂ was dense and thick. The instability of the hydrothermally sensitized layers caused a chemical reaction as a result of which MoS₂ was transferred into the electrolyte during photoelectrochemical tests; the layers subsequently degraded. For this reason, the hydrothermally modified samples were excluded from further analysis.

The results of photoelectrochemical measurements performed based on the photocurrent–voltage of TiO₂ and TiO₂-based photoanodes are presented in Fig. 7c in the form of a comparison of flat-band potential and the ratio of the MoS₂-modified TiO₂ photocurrent to the unmodified TiO₂ photocurrent (I_modified-to-I_TiO2 ratio) at 0 and 1.3 V. For unmodified samples, it was assumed that the photocurrent ratio is equal to 1. Flat-band potential (V_fb) is a very important factor from the point of view of processes occurring in a PEC; one of the effects of the shift of V_fb towards a more negative potential is better photoelectrochemical performance.^43,72 Its value was assigned as shown in Fig. 7a.

As can be seen, both TiO₂ anodes were characterized by similarly negative flat-band potentials, with values of −462 mV for TiO₂-NT, and −442 mV for TiO₂-L (Fig. 7c). The flat-band potential of GO/TiO₂-NT did not significantly differ from that of TiO₂-NT. On the other hand, MoS₂ deposition on the surface of TiO₂-L had a positive impact on the flat-band potential position. For both 1Me/TiO₂-L and 2Me/TiO₂-L, V_fb shifted towards more negative potentials, down to −480 mV and −518 mV, respectively. In the case of sensitization with the use of either GO and MoS₂, the photocurrent of the modified electrodes was greater than that observed for the unmodified ones. The comparison of the corresponding ratios shows that the best performance was achieved for 2Me/TiO₂-L, for which the value of the photocurrent ratio was equal to 1.85 at 0 V and 1.80 at 1.3 V. For 1Me/TiO₂-L and GO/TiO₂-NT, the ratios were comparable. The ratios at 0 V were equal to 1.34 for 1Me/TiO₂-L and 1.39 for GO/TiO₂-NT, while at 1.3 V they were 1.28 and 1.32, respectively. The better performance of 2Me/TiO₂-L in comparison to 1Me/TiO₂-L can be explained by the microstructure and degree of crystallization. The analysis of the SEM images (Fig. 3a) shows that the nanoflagella-like elements in the 2M12 powder are formed better than in the case of 1M12. This can be a consequence of a higher degree of crystallization, which can be partially observed in Fig. 3b and c. This influence of crystallization on photoanode behavior was also observed by Gorzkowska-Sobas et al. in the case of TiO₂ thin films.⁷³ Furthermore, the electrical resistance of R and R₂ estimated from the electrical equivalent circuit is clearly lower for 2Me/TiO₂ than it is for 1Me/TiO₂ (Table 2). Thus, the transport of electron carriers is also faster. Combined with the lowest value of the flat-band potential, this results in the highest photocurrent response.

On the basis of the electrochemical impedance spectroscopic and photoelectrochemical measurements, the architecture of the modified electrodes was proposed. Fig. 8 shows a schematic representation of the cross-section of photoelectrodes based on TiO₂. The photoelectrodes consist of TiO₂ (nanotubes or compact oxide) and a layer formed with two-dimensional materials, i.e. either MoS₂ or GO. The interfacial properties of modified TiO₂ entail a different microstructure of electrodes, which determines their photoelectrochemical properties. MoS₂ and GO layers deposited onto TiO₂ via electrochemical methods cover the electrode surface only partially. TiO₂ with a second phase in the form of islands facilitates the penetration of the electrolyte into the electrode and enhances photoactivity. Densely packed MoS₂ nanoparticles produced by means of the hydrothermal technique allow the formation of a barrier at the TiO₂/electrolyte interface and inhibit photoelectrochemical processes.

Fig. 8 Architecture of the investigated electrodes based on TiO₂ (cross-section) and corresponding electrical equivalent circuits.

4. Conclusions

MoS₂/TiO₂ and GO/TiO₂ composites were successfully fabricated via a simple and environmentally friendly electrochemical deposition and – for comparison – MoS₂ was also deposited using a hydrothermal process. For this purpose, GO was prepared from expanded graphite with the use of the modified Hummers' method, and MoS₂ was synthesized under hydrothermal conditions. It was found that – compared to pure TiO₂ – MoS₂/TiO₂ and GO/TiO₂ exhibited a modified morphology and electrochemical response, lower charge transfer resistance, and significantly enhanced photoelectrochemical behavior. MoS₂ deposited on the surface of TiO₂ via electrodeposition demonstrated a beneficial effect with regard to photoanode performance in PECs. The same was not true in the case of the hydrothermal approach. This work provides a new strategy for MoS₂/TiO₂ formation; this approach can improve the photocurrent response of TiO₂ and extend their application in the photocatalytic decomposition of water.

Acknowledgements

The project was financed by the National Science Centre (NCN) based on decision number DEC-2011/01/D/ST5/05859. The authors would like to kindly acknowledge Katarzyna Kolacz for her input into GO synthesis.

References

1. G. Xie, K. Zhang, B. Guo, Q. Liu, L. Fang and J. R. Gong, Adv. Mater., 2013, 25, 3820–3839.
2. K. Kołacz, M. Gajewska, S. Komornicki and M. Radecka, Int. J. Hydrogen Energy, 2016, 41, 7538–7547.
3. D. Chen, W. Chen, L. Ma, G. Ji, K. Chang and J. Y. Lee, Mater. Today, 2014, 17, 184–193.
4. A. Castellanos-Gomez, L. Vicarelli, E. Prada, J. O. Island, K. L. Narasimha-Acharya, S. I. Blanter, D. J. Groenendijk, M. Buscema, G. A. Steele, J. V. Alvarez, H. W. Zandbergen, J. J. Palacios and H. S. J. van der Zant, 2D Mater., 2014, 1, 025001.
5. S. M. Kim, A. Hsu, M. H. Park, S. H. Chae, S. J. Yun, J. S. Lee, D.-H. Cho, W. Fang, C. Lee, T. Palacios, M. Dresselhaus, K. K. Kim, Y. H. Lee and J. Kong, Nat. Commun., 2015, 6, 8662.
6. M. Chhowalla, H. S. Shin, G. Eda, L.-J. Li, K. P. Loh and H. Zhang, Nat. Chem., 2013, 5, 263–275.
7. J. Sun, H. Zhang, L. H. Guo and L. Zhao, ACS Appl. Mater. Interfaces, 2013, 5, 13035–13041.
8. W. S. Hummers and R. E. Offeman, J. Am. Chem. Soc., 1958, 80, 1339.
9. M. Ye, D. Winslow, D. Zhang, R. Pandey and Y. Yap, Photonics, 2015, 2, 288–307.
10. S. A. Han, R. Bhatia and S.-W. Kim, Nano Convergence, 2015, 2, 17.
11. S. Wang, G. Li, G. Du, X. Jiang, C. Feng, Z. Guo and S. J. Kim, Chin. J. Chem. Eng., 2010, 18, 910–913.
12. G. Jiang, Z. Lin, C. Chen, L. Zhu, Q. Chang, N. Wang, W. Wei and H. Tang, Carbon, 2011, 49, 2693–2701.
13. T. D. Nguyen-Phan, V. H. Pham, E. W. Shin, H. D. Pham, S. Kim, J. S. Chung, E. J. Kim and S. H. Hur, Chem. Eng. J., 2011, 170, 226–232.
14. S. Morales-Torres, L. M. Pastrana-Martinez, J. L. Figueiredo, J. L. Faria and A. M. T. Silva, Appl. Surf. Sci., 2013, 275, 361–368.
15. X. Pu, D. Zhang, Y. Gao, X. Shao, G. Ding, S. Li and S. Zhao, J. Alloys Compd., 2013, 551, 382–388.
16. Y. P. Zhang, J. J. Xu, Z. H. Sun, C. Z. Li and C. X. Pan, Prog. Nat. Sci.: Mater. Int., 2011, 21, 467–471.
17. H. Il Kim, S. Kim, J. K. Kang and W. Choi, J. Catal., 2014, 309, 49–57.
18. D. Robert, Catal. Today, 2007, 122, 20–26.
19. M. Shanmugam, T. Bansal, C. A. Durcan and Y. Bin, Nanotechnol. (IEEE-NANO), 2012 12th IEEE Conf., 2012, pp. 1–4.
20. Y. Zhu, Q. Ling, Y. Liu, H. Wang and Y. Zhu, Phys. Chem. Chem. Phys., 2015, 17, 933–940.
21. G. Malekshoar and A. K. Ray, Chem. Eng. Sci., 2016, 152, 35–44.
22. V. Stengl, S. Bakardjieva, T. M. Grygar, J. Bludská and M. Kormunda, Chem. Cent. J., 2013, 7, 41.
23. Y. Park, S.-H. Kang and W. Choi, Phys. Chem. Chem. Phys., 2011, 13, 9425–9431.
24. P. Song, X. Zhang, M. Sun, X. Cui and Y. Lin, Nanoscale, 2012, 4, 1800.
25. Z. He, W. Que, Y. Xing and X. Liu, J. Alloys Compd., 2016, 672, 481–488.
26. W.-H. Ryu, J.-W. Jung, K. Park, S.-J. Kim and I.-D. Kim, Nanoscale, 2014, 6, 10975–10981.
27. S. Kanda, T. Akita, M. Fujishima and H. Tada, J. Colloid Interface Sci., 2011, 354, 607–610.
28. T. Du, N. Wang, H. Chen, H. He, H. Lin and K. Liu, J. Power Sources, 2015, 275, 943–949.
29. Y. Chen, Y. Tang, S. Luo, C. Liu and Y. Li, J. Alloys Compd., 2013, 578, 242–248.
30. J.-H. Yun, R. J. Wong, Y. H. Ng, A. Du and R. Amal, RSC Adv., 2012, 2, 8164.
31. X. Luan, L. Chen, J. Zhang, G. Qu, J. C. Flake and Y. Wang, Electrochim. Acta, 2013, 111, 216–222.
32. S. K. Ghosh, C. Srivastava, S. Nath and J. P. Celis, Int. J. Electrochem., 2013, 2013, 1–7.
33. E. A. Ponomarev, M. Neumann-Spallart, G. Hodes and C. Lévy-Clément, Thin Solid Films, 1996, 280, 86–89.
34. A. W. Maijenburg, M. Regis, A. N. Hattori, H. Tanaka, K. S. Choi and J. E. Ten Elshof, ACS Appl. Mater. Interfaces, 2014, 6, 2003–2010.
35. S. Murugesan, A. Akkineni, B. P. Chou, M. S. Glaz, D. A. Vanden Bout and K. J. Stevenson, ACS Nano, 2013, 7, 8199–8205.
36. S. Shariza and T. Anand, Chalcogenide Lett., 2011, 8, 529–539.
37. D. Merki, S. Fierro, H. Vrubel and X. Hu, Chem. Sci., 2011, 2, 1262.
38. E. S. Güler, E. Konca and İ. Karakaya, Int. J. Electrochem. Sci., 2013, 8, 5496–5505.
39. V. Kanagalasara and T. V. Venkatesha, J. Solid State Electrochem., 2012, 16, 993–1001.
40. A. Kusior, A. Wnuk, A. Trenczek-Zajac, K. Zakrzewska and M. Radecka, Int. J. Hydrogen Energy, 2015, 40, 4936–4944.
41. A. Trenczek-Zajac and J. Banas, Funct. Mater. Lett., 2016, 9, 1641004.
42. R. C. Pawar and C. S. Lee, Heterogeneous Nanocomposite-Photocatalysis for Water Purification, 2015.
43. M. Radecka, M. Rekas, A. Trenczek-Zajac and K. Zakrzewska, J. Power Sources, 2008, 181, 46–55.
44. I. Childres, L. Jauregui, W. Park, H. Cao and Y. Chen, New Dev. Photon Mater. Res., 2013, 1–20.
45. R. Beams, L. G. Canç Ado and L. Novotny, J. Phys.: Condens. Matter, 2015, 27, 83002–83026.
46. I.-Y. Jeon, Y.-R. Shin, G.-J. Sohn, H.-J. Choi, S.-Y. Bae, J. Mahmood, S.-M. Jung, J.-M. Seo, M.-J. Kim, D. Wook Chang, L. Dai and J.-B. Baek, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 5588–5593.
47. K. Krishnamoorthy, M. Veerapandian, K. Yun and S. J. Kim, Carbon, 2013, 53, 38–49.
48. S. Claramunt, A. Varea, D. López-Díaz, M. M. Velázquez, A. Cornet and A. Cirera, J. Phys. Chem. C, 2015, 119, 10123–10129.
49. S. M. Hafiz, R. Ritikos, T. J. Whitcher, N. M. Razib, D. Chia, S. Bien, N. Chanlek, H. Nakajima, T. Saisopa, P. Songsiriritthigul, N. M. Huang and S. A. Rahman, Sens. Actuators, B, 2014, 193, 692–700.
50. J. M. Munuera, J. I. Paredes, S. Villar-Rodil, M. Ayán-Varela, A. Martínez-Alonso and J. M. D. Tascon, Nanoscale, 2016, 2982–2998.
51. W. Qiao, S. Yan, X. Song, X. Zhang, X. He, W. Zhong and Y. Du, Appl. Surf. Sci., 2015, 359, 130–136.
52. K. Gołasa, M. Grzeszczyk, K. P. Korona, R. Bożek, J. Binder, J. Szczytko, A. Wysmołek and A. Babiński, Acta Phys. Pol., A, 2013, 124, 849–851.
53. C. Rice, R. J. Young, R. Zan, U. Bangert, D. Wolverson, T. Georgiou, R. Jalil and K. S. Novoselov, Phys. Rev. B: Condens. Matter Mater. Phys., 2013, 87, 1–5.
54. L. Liang and V. Meunier, Nanoscale, 2014, 6, 5394.
55. M.-R. Gao, M. K. Y. Chan and Y. Sun, Nat. Commun., 2015, 6, 7493.
56. H. Wang, Z. Lu, S. Xu, D. Kong, J. J. Cha, G. Zheng, P. C. Hsu, K. Yan, D. Bradshaw, F. B. Prinz and Y. Cui, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 19701–19706.
57. W. Zhao, Z. Ghorannevis, K. Amara and J. Pang, Nanoscale, 2013, 5, 9677–9683.
58. E. S. W. Kong, Nanomaterials, Polymers and Devices: Materials Functionalization and Device Fabrication, John Wiley & Sons, Inc, 2015.
59. L. Yang, X. Cui, J. Zhang, K. Wang, M. Shen, S. Zeng, S. A. Dayeh, L. Feng and B. Xiang, Sci. Rep., 2014, 4, 5649.
60. M. Thripuranthaka, R. V. Kashid, C. Sekhar Rout and D. J. Late, Appl. Phys. Lett., 2014, 104, 081911.
61. P. Qin, G. Fang, W. Ke, F. Cheng, Q. Zheng, J. Wan, H. Lei and X. Zhao, J. Mater. Chem. A, 2014, 2, 2742.
62. H. S. Sen, H. Sahin, F. M. Peeters and E. Durgun, J. Appl. Phys., 2014, 116, 083508.
63. M. Acerce, D. Voiry and M. Chhowalla, Nat. Nanotechnol., 2015, 10, 313–318.
64. G. Huang, T. Chen, W. Chen, Z. Wang, K. Chang, L. Ma, F. Huang, D. Chen and J. Y. Lee, Small, 2013, 9, 3693–3703.
65. Q. Liu, X. Li, Z. Xiao, Y. Zhou, H. Chen, A. Khalil, T. Xiang, J. Xu, W. Chu, X. Wu, J. Yang, C. Wang, Y. Xiong, C. Jin, P. M. Ajayan and L. Song, Adv. Mater., 2015, 27, 4837–4844.
66. W. H. Mulder, J. H. Sluyters, T. Pajkossy and L. Nyikos, J. Electroanal. Chem., 1990, 285, 103–115.
67. W. Strunz, C. A. Schiller and J. Vogelsang, Electrochim. Acta, 2006, 51, 1437–1442.
68. C. H. Kim, S. I. Pyun and J. H. Kim, Electrochim. Acta, 2003, 48, 3455–3463.
69. B. Pourabbas and B. Jamshidi, Chem. Eng. J., 2008, 138, 55–62.
70. A. Trenczek-Zajac, A. Kusior and M. Radecka, Int. J. Hydrogen Energy, 2016, 41, 7548–7562.
71. A. Trenczek-Zajac, A. Kusior, A. Lacz, M. Radecka and K. Zakrzewska, Mater. Res. Bull., 2014, 60, 28–37.
72. M. Finlayson and B. Wheeler, J. Phys. Chem., 1985, 89, 5676–5681.
73. A. Gorzkowska-Sobas, E. Kusior, M. Radecka and K. Zakrzewska, Surf. Sci., 2006, 600, 3964–3970.

---