Room Temperature Exsolution of Cds Nanodots on A‐site Deficient Cotton‐Ball Like Titanate Perovskite Nanoparticles for H2 Production Under Visible Light

Exsolution of nanoparticles followed by chemical treatment (“chemistry at a point”) is a very exciting approach to the smart design of functional materials such as visible light active photocatalysts. Unfortunately, the usually utilized thermal reduction approach is not feasible for low melting point metals and compounds such as Cd and CdO. Here a hydrothermal approach to prepare exsolved CdS nanodots on cotton ball‐like perovskite supports is described. The titanate‐based photocatalyst is synthesized using a hydrothermal process followed by room‐temperature sulfidation. The hydrothermal route directs A‐site doping of Cd2+ via hydroxyl group incorporation in the titanate lattice. Formation of CdS via exsolution provides a high H2 production mass activity of 3050 µmol g−1 h−1 under visible light with only 5 mol.% Cd doping of the titanate. Moreover, the strong CdS‐support interaction offers good cycling stability under UV–vis and visible light irradiation. This is the first report describing the exsolution of CdS nanodots at room temperature and shows its advantages for photocatalytic activity.


Introduction
With increasing demand for decarbonization, renewable solar energy for the production of green H 2 fuel via photocatalytic water splitting continues to attract significant interest, providing a potential means for green energy storage. [1,2]Among various visible light-absorbing semiconductors, CdS with its band gap of ≈2.24 eV is a potential candidate for this purpose. [3,4]However, photo-corrosion is a serious drawback of CdS, therefore DOI: 10.1002/aenm.202301381several strategies are adopted to improve its stability. [5,6]Moreover, there are reports on loading noble metal cocatalysts like Pt or Au on CdS to achieve high-performance photocatalysts, at the expense of increased cost of the photocatalysts. [7,8]11] In this aspect, SrTiO 3 would be suitable as it allows modification of its physicochemical properties by tuning the stoichiometry through A or B site cation substitution. [9,12]Additionally, aliovalent La 3+ doping results in A-site deficiency and hence lower the fermi level position, in turn increasing charge-carrier mobility and therefore influencing H 2 evolution activity. [13]Therefore, coupling these advantageous properties of (La x Sr y )TiO 3 with those of CdS, would provide a promising route to realizing an efficient catalytic system.However, the synthesis of Cd doped (La x Sr y )TiO 3 followed by exsolution of CdS is particularly challenging.Conventional high-temperature synthesis routes like the solid-state or pechini method are restricted by the low melting points of CdO and CdS.Because of this reason, a low-temperature sulfidation route is also necessary to exsolve out Cd 2+ from the titanate lattice as CdS.To date, the concept of exsolution has largely been reported as thermal reduction leading to metal nanoparticles. [14]Such nanoparticles have undergone chemical reaction whilst remaining constrained at the surface in the phenomenon of "chemistry at a point" and this approach has been successfully applied to WS 2 . [15,16]herefore, this current work would provide a new direction of research.
Here, we have developed a novel low-temperature hydrothermal synthesis route to achieve Cd 2+ doped La 0.4 Sr 0.4 TiO 3 followed by a room temperature sulfidation technique (Figure S2, Supporting Information) for in situ growth of CdS via exsolution.Furthermore, the visible light photoactivity of CdS exsolved photocatalysts was investigated and a superior mass activity for CdS content was observed.

Results and Discussion
La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- was prepared hydrothermally at 190 °C followed by 400 °C treatment in the air for 2 h as described in the Figure S1, (Supporting Information).X-ray diffraction (XRD) (Figure 1a) of as-synthesized samples confirmed phase pure perovskite oxide with cubic (Pm-3m) crystal symmetry.A comparative study of XRD with La 0.4 Sr 0.4 TiO 3 fabricated by solid-state synthesis, La 0.4 Sr 0.4 TiO 3 (SS), suggested cell expansion by showing increase in cell parameter value for hydrothermal (inset, Figure 1a) route in comparison to solid state method.It has been reported that such increment can be attained during hydrothermal reaction due to the incorporation of hydroxyl group in the lattice by replacing the oxygen vacancies. [17]Furthermore, broadening of diffraction peaks and shifting toward a higher 2 value (inset, Figure 1a) on Cd 2+ doping implies successful incorporation of Cd 2+ in the framework with contraction of lattice parameter and cell volume.This is only possible when Cd 2+ (0.95, 1.31 Å with coordination numbers VI and XII, respectively) with a lower ionic radius than Sr 2+ (1.44 Å) and La 3+ (1.36) occupies an A-site in the structure.Therefore, probably hydroxide group-assisted Asite doping of Cd 2+ occurred. [18]No visible CdO-related peak was observed for La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- further confirming the dissolution of Cd + in the lattice.Notably, the concentration of KOH solution was found to be crucial during synthesis to attain successful doping without any impurity.FTIR and TG analysis of the samples were further undertaken to ensure the fact of hydroxyl group incorporation.FTIR analysis (Figure 1b) showed presence of hydroxyl-related characteristics vibration bands at 3430 and 1646 cm −1 for La 0.4 Sr 0.4 TiO 3 and La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- , however, no such peaks were present for La 0.4 Sr 0.4 Ti 0.95 O 3 (SS). [19]The vibration peaks at 1370 cm −1 may arise from trace amounts of alcoholic -OH from ethylene glycol used during synthesis.The peaks in the range of 920-557 cm −1 are due to Ti-O vibrations. [20]s can be seen in Figure 1b, the peak at ≈810 cm −1 appears in the higher wavenumber region which may be due to the differences in crystallite size and the presence of -OH group.TG analysis (Figure 1c) also shows 1.96% weight loss in the temperature range 200-600 °C which corresponds to loss of ≈0.22 moles of [OH] from La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- .Therefore, these two results corroborated the XRD analysis suggesting incorporation of hydroxide group.Scanning electron microscopy (SEM) of La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- (Figure 2a,b) showed perovskite grains of ≈40-50 nm with unusual cotton ball-like morphologies.The magnified image in Figure 2b shows the presence of sheetlike structures aggregated to grow the hierarchical morphology.Scanning transmission electron microscopy (STEM) analysis (Figure 2c) further demonstrated the porous hierarchical structure of the perovskite grains.Figure 2d confirms the clear existence of nanosheets exposed on the surface of the cotton-ball-like nanoparticles.The atomic resolution image in Figure 2e shows the (pseudo)cubic arrangement of the metal ions where relatively larger bright spots and small lighter spots correspond to A and B site cations, respectively.Lattice fringes correspond to (110) plane with an interplanar spacing of ≈ 0.36 nm, higher than normal value can also be observed.Hence, we propose that during the hydrothermal synthesis, initially perovskite nanosheets are formed and those undergo self-aggregation to result in such a hierarchical structure, Figure 2f.EDX elemental mapping (Figure S3, Supporting Information) confirmed the homogenous distribution of the elements without any phase segregation and a composition corresponding to La 0.43 Sr 0.49 Cd 0.05 Ti 0.95 O x which is close to the expected composition.A small amount of potassium (≈2 wt.%) was found as an impurity from the KOH used for synthesis.Since surface area is an important parameter to assess the photocatalytic activity of a photocatalyst, N 2 sorption analysis was performed to measure the specific surface area and porosity of La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- .A typical Type IV adsorptiondesorption isotherm (Figure S4, Supporting Information) was obtained La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- implying the presence of mesoporosity.The corresponding pore size distribution plot (inset, Figure S4, Supporting Information) was derived from the adsorption branch of the isotherm.BET surface area value of 54.55 m 2 g −1 and average pore diameter ≈10 nm with pore volume 0.28 cm 3 g −1 were calculated.Additionally, the presence of smaller pores of diameter ≈2 and 5 nm can also be seen in the pore size distribution plot.
Therefore, such a large surface area and porous structure would offer easy access to the active site for photocatalysis.To avoid any thermal treatment, a room-temperature sulfidation route was designed to exsolve CdS from the perovskite oxide framework (Figure S2  formation of CdS on exsolution.It is noteworthy that diffusion of exsolving metal ions from deep inside the bulk to the surface is one of the crucial steps during exsolution. [21]This process requires a strong stimulus such as a concentration gradient or energy (i.e., temperature) to attain Gibb's free energy.In our case the affinity of Cd 2+ for S 2− facilitated the diffusion of Cd 2+ and simultaneously the unique nanoscale dimensions of the perovskite grains likely helped in lowering the free energy associated with such diffusion and exsolution.Moreover, the high porosity present on such hierarchical La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- structures also played an important role in exsolution by providing escape sites.Therefore, we could achieve room tem-perature exsolution of CdS on the surface.This result revealed the existence of CdS nanodots with a smaller band gap.Therefore, the color change of the powder from white to yellow along with the formation of CdS absorption in UV-vis absorption spectrum explicitly indicate the formation of CdS on exsolution.
Average particle size of exsolved CdS NPs was further calculated using Henglein's empirical formula (Equation.1) and found to be ≈3.5 nm.In the equation, R(CdS) and a are the radii and absorption maxima of exsolved CdS nanodots.Band gap of La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- before and after CdS exsolution was evaluated using the Tauc plot (Figure 3c) derived from by Kubelka-Munk transformation of UV-vis absorption spectra and extrapolating the linear part down to energy axis.The band gap values were found to be 3.  5).The fitted core level spectra for each of these samples are provided in Figure S6, Supporting Information.The S 2p core level spectra were obtained by performing rapid measurements at various points across an area rather than at a single point, due to the observed decrease in sulfur intensity after prolonged exposure of the samples to the X-ray beam.The binding energies and relative atomic concentrations of the component elements and their respective chemical species are listed in Table 1.The plotted core level spectra were normalized to the area of the Sr 3d 5/2 peak, owing to the fixed abundance of Sr and the consistent shape of the Sr 3d core level between the three samples, indicative of a constant chemical environment.The Sr 3d core levels (Figure 5a) of La 0.4 Sr 0.4 TiO 3 , La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3-, and La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- with exsolved CdS were fitted with a single doublet, comprising the Sr 3d 5/2 at a binding energy of ≈132.5 eV, separated from the higher binding energy Sr 3d 3/2 component (Figure S6, Supporting Information) by a spin-orbit splitting of 1.75 eV.This doublet peak is attributed to lattice Sr 2+ in all the samples. [22,23]A small difference of ±0.2 eV in the binding energy position of the Sr 3d doublet was observed between samples.This is within the expected error, and it is not considered to be indicative of any    126.17μmol g −1 h −1 under full range and visible light irradiation, respectively when 10% V TEOA was used as sacrificial agent (Figure S8, Supporting Information).These values were 1.2 and 4.5 times higher than bare CdS.Moreover, with 0.1 m Na 2 S /Na 2 SO 3 as a sacrificial agent, H 2 production rates of 4874 and 3058 μmol g −1 h −1 were obtained under full range and visible light, respectively, Figure 7b.Whereas, bare CdS exhibited 338.2 and 127.12 μmol g −1 h −1 , under full range and visible light, respectively.Therefore, almost 14 and 24 fold increases in mass activity for CdS amount were achieved.[36][37][38][39][40] Such improvements in photocatalytic property revealed the influence of strong interaction between CdS nanodots and perovskite support which has been established by XPS.This intrinsic interaction retarded recombination rate by easy separation of electron and holes, and therefore facilitated successful utilization of photogenerated electrons.Moreover, the small particle size of perovskite NPs also helps in the fast separation of charge carriers and effective utilization of active sites.Hence, an improved photocatalytic H 2 production activity was achieved.To prove the stability of ex- This result also supports the strong anchorage of CdS on perovskite support resulting from the exsolution that helps in protecting CdS, and therefore, suppresses photo-corrosion and leaching.Therefore, our approach not only offered enhanced photocatalytic activity but also assured photostability to our photocatalyst.

Conclusion
We demonstrated a smart approach for room temperature exsolution of CdS nanodots on perovskite NPs.Hydrothermal synthesis route followed by low-temperature calcination yielded cotton ball-like perovskite NPs of size ≈40-50 nm with porosity that also created a high specific surface area.Such small particle size of perovskite with porous framework facilitated faster diffusion of Cd 2+ ions and allowed to achieve room temperature sulfidation assisted CdS nanodot exsolution.Strong anchorage of CdS nanodots on perovskite support not only ensured photostability to CdS but also resulted in enhancement in photocatalytic H 2 production activity including under visible light which was confirmed by using different sacrificial agents.Therefore, this study has the potential for further development of the photocatalysts with improved activity.Moreover, the synthesis concept pre-sented in this study can be applied for low-temperature exsolution of other catalytically active NPs and therefore demonstrates wide applicability.
, Supporting Information).A visible color change of La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- from white to yellow powder (Figure S5, Supporting Information) suggested the
Figure 3a shows XRD pattern of the La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- obtained after sulfidation.No CdS-related diffraction peaks were observed indicating small particle size of exsolved CdS.UV-Vis absorption spectroscopic technique was used to identify CdS exsolution.UV-Vis absorption spectrum of La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- (Figure 3b) showed absorption only in UV range with an absorption maximum at ≈256 nm originating from the titanate component.Interestingly, there is a clear appearance of CdS absorbance band for La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- with exsolved CdS.The corresponding spectrum showed a strong visible light absorption with a maximum of around 450 nm (Figure 3b) with
31 and 3.19 eV for La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- and La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- with exsolved CdS, respectively.La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- with exsolved CdS showed an additional E g value at around 2.38 eV corresponding to the CdS nanodots.Therefore, exsolution of CdS led to the change in light absorption property of the material and was in good agreement with the color change observed after sulfidation.TEM analysis of CdS exsolved La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- was performed to check the microstructure and element distribution after sulfidation reaction, Figure 4. Corresponding EDX elemental mapping showed a homogeneous distribution of all elements including cadmium and sulfur, coming from CdS.The quantification of the elements is summarized in TableS1, Supporting Infor-mation.Under TEM it is very difficult to spot the nanodots as they are emerged on the thick perovskite oxide bulk surface.Moreover, the perovskite oxide nanoparticles are formed via accumulation of sheet-like structures which made it even more challenging to observe the CdS nanodot, and, for this reason, Cd and S dispersion in Figure4look agglomerated although in some parts small particles like dispersion can also be observed.XPS measurements of the Sr 3d, Ti 2p, La 3d, Cd 3d, O 1s, and S 2p core levels were performed for La 0.4 Sr 0.4 TiO 3 , La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- and La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- with exsolved CdS (Figure

solved
CdS on La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- , recovered photocatalysts were further characterized thoroughly after a water-splitting experiment under full range and visible light.UV-Vis (Figure S10, Supporting Information) showed no change in UV-Vis absorption edges after the photocatalysis reactions.TEM confirmed retention of microstructure and no elemental segregation or leaching was observed in corresponding EDX element mapping (Figure S11, Supporting Information), which proved the structural stability of CdS exsolved La 0.4 Sr 0.4 Cd 0.05 Ti 0.95 O 3- .Moreover, the distribution of Cd and S over CdS nanodots of size ≈3-4 nm was also evaluated during line scanning (FigureS12, Supporting Information) under TEM.TEM elemental mapping (FigureS13, Supporting Information) of recovered samples after testing under visible light also confirmed the stability under such reaction conditions.It is noteworthy that, the EDS analysis presented in FiguresS11,S13, (Supporting Information) show Cd: Sr ratios to be 0.110 and 0.10, respectively which is the same as the value obtained from Figure4before catalysis, Cd:Sr = 0.114.This justifies that there is almost no leaching of Cd in the solution.

Figure 7 .
Figure 7. a) Cycling stability test under visible light ( ≥ 400 nm) and b) comparative mass activity for CdS content with 0.1 m Na2S /Na2SO3 sacrificial agent.