First principles study of the adsorption and dissociation mechanisms of H₂S on a TiO₂ anatase (001) surface

Abstract

The adsorption and dissociation mechanisms of H₂S on a TiO₂ (001) surface were elucidated using first principles calculation based on the density functional theory. The interaction of the intermediates involving S, HS and OH species has been discussed in detail. For these species, the most favorable adsorption sites were determined and led to further computations involving dissociation of H₂S into HS and H and bonding of H-atom with the OH to form water on the surface. The creation of vacancies along with the presence of S on the surface enhanced the H₂S adsorption energy slightly. However, addition of OH in the system caused H₂S to bind on the surface sufficiently strongly. Additionally, H₂S decomposition was found to be a spontaneous process in the presence of OH radicals.

---

1 Introduction

Hydrogen sulphide (H₂S) is an essential factor of tail gas poising in many industrial processes and is one of the major products extracted during fuel gas processing.^1,2 The removal of H₂S from the tail gas and many industrial emissions using coal and sulphur products has gained a lot of importance recently because of its toxicity and adverse effects. Due to growing awareness about the depleting environmental resources, the environmental protection agencies are imposing more and more stringent air pollution regulations which include release of gases containing sulphur as well.³ A lot of research is being carried out throughout the world making use of various semiconductors for degrading pollutants. Among them, TiO₂ in the form of anatase has attracted wide interest due to its strong oxidizing power under ultra violet (UV) irradiation, its chemical stability and absence of toxicity.^4–9 Additionally, due to its high activity in degrading organic and inorganic air pollutants, titanium dioxide is used as a photocatalyst. It is biologically and chemically inert and cheaper as well.^10–12

Canela and co-authors have carried out a study making use of the TiO₂ nanoparticles for studying the gas phase destruction of H₂S gas under UV and visible light conditions and found it to be quite efficient.¹³ In another study conducted by our group to investigate the catalytic potential of TiO₂ nanoparticles, we also observed a significant H₂S destruction efficiency at high temperatures.^14,15 We also carried out studies to evaluate the photocatalytic potential of TiO₂ nanofibers and found them to be considerably efficient for the destruction of H₂S gas.¹⁶

Adsorption and dissociation of H₂S has been extensively studied on various metals including Cu (111) and (100), where dissociation of H₂S resulted when Cu attracted one of its H atom.^17,18 H₂S decomposition on ZnO surface concluded in H₂O formation.¹⁹ Masoud and co-authors found up to seven molecules of H₂S binding to Pt atom.²⁰ Additionally, various transition metals were taken into consideration for H₂S decomposition reported in ref. 21 and the process was found facile over them. Huang and co authors studied the adsorption and reaction of H₂S on TiO₂ rutile (110) and anatase (101) surfaces.²² The authors suggested two different pathways for H₂S dissociation on TiO₂ surface resulting in formation of H₂O and H₂, respectively, which was comparatively harder for rutile (110) surface than anatase (101) due to lower energy barrier in the latter case.

Our experimental studies and findings are based on TiO₂ nanoparticles and nanofibers in anatase phase. We have carried out these studies because metastable anatase form is found to be stable at nanoscale,^23,24 because of relatively low surface energies of its free (101) and (100) facets.^22,25–27 Rutile phase is most stable for larger diameters.²⁴ Although, (101) faces in anatase are dominantly present, (001) surface exposes potentially more reactive facet having surface energy of 0.98 J m⁻² because of stress on surface cleavage. This energy is almost double the value of ∼0.5 J m⁻², found for stable (101) facet.²⁸ In addition, many theoretical studies found (001) surface to be more reactive.^29,30 Moreover, for (001) surface, the two fold coordinated oxygen atoms (indicated as O_2c in Fig. 1) with 156° Ti–O–Ti angles are prone to reaction with adsorbates.²⁸

Fig. 1 TiO₂ slab and top view of (001) surface of anatase showing different positions of Ti and O atoms.

After exploring the H₂S destruction potential of pure and doped TiO₂ nanocatalysts using various catalytic and photocatalytic experiments as narrated above, we used density functional theory (DFT) studies to further evaluate the H₂S destruction and validate our experimental findings on the TiO₂ anatase (001). The aim of the present work is to explore the adsorption and dissociation mechanism of H₂S. The effect of co-adsorption of various species and defect creation has been evaluated on adsorption energy and destruction of H₂S.

2 Computational details

Our calculations were carried out using the Vienna ab-initio simulation package (VASP)³¹ which uses plane wave as basis set to perform an iterative solution of the Kohn–Sham equations. Plane waves with a kinetic energy ≤400 eV were included in the calculations. The generalized gradient approximation (GGA) of Perdew and Wang (PW91) formalism was used to calculate the exchange–correlation energy.³² Blöchl's projector-augmented wave (PAW) method was used to describe the electron–ion interactions for Ti, O, S and H atoms.³³ This is essentially a scheme combining the accuracy of all electron methods and the computational simplicity of the pseudo potential approach.³⁴

P(2 × 2) unit cell of anatase (001) which contains four Ti_5c atoms on the surface, has been used in these computations. Each of these Ti_5c atom is bonded to two raised 2c and two 3c lowered oxygen atoms in the (100) and (010) directions, respectively. Unit cell consists of 16 TiO₂ units and periodically repeated slabs which are separated by ∼12 Å width. The reciprocal space for the p(2 × 2) unit cell was sampled with (4 × 4 × 1) k-point meshes generated automatically using the Monkhorst–Pack method.³⁵ The upper half of the Ti and O atoms in the unit cell were relaxed. The relative positions of the Ti ions were fixed initially as those in the bulk, with optimized lattice parameters of 3.80 Å and 9.49 Å (the experimental values are 3.78 Å and 9.51 Å).³⁶ The optimized lattice parameter was calculated using the tetragonal anatase unit cell and its reciprocal space was sampled with a (8 × 8 × 3) k-point grid. A first-order Methfessel–Paxton smearing-function with a width ≤0.1 eV was used to account for fractional occupancies.³⁷

Partial geometry optimizations were performed using the RMM-DIIS algorithm³⁸ and were stopped when all the forces were <0.05 eV Å⁻¹. Vibrational frequencies for transition states (TS) were calculated within the harmonic approximation. Hessian matrix of the adsorbate was calculated neglecting the adsorbate-surface coupling.³⁹ The climbing image nudged elastic band (cNEB) method was used in this study to determine minimum energy paths.⁴⁰ Closed shell H₂S and H₂O molecules were optimized at the gamma point by non-spin polarized calculations using a 10 × 10 × 10 ų cubic unit cell. Spin-polarized calculations in a 10 × 12 × 14 ų orthorhombic unit cell were used for open shell species, H, O, S, OH and HS. All the calculations were performed at T = 0 K.

The slab of TiO₂ (001) is shown in Fig. 1. The face (001) consists of periodically repeated slabs consisting of 16 units. The bond angles between Ti_5c–O_2c–Ti_5c, Ti_5c–O_3c–Ti_5c are 150 Å and 156.9 Å, respectively, while the distances between Ti_5c–O_2c, Ti_5c–O_3c are 1.97 Å and 1.94 Å, respectively.

The computations to determine the most stable adsorption configuration for H₂S, HS, S, H and OH were carried out by placing each of them on four different possible locations, namely, hollow, five-fold bridging titanium (Ti_5c), two-fold bridging oxygen (O_2c) and three-fold bridging oxygen (O_3c). The adsorption energies (E_ads) were calculated by using the following equation:²⁸

E_ads = [E_a/s − E_s − nE_a(gas)]/n (1)

where, E_ads = adsorbed energy of the fragment adsorbed, E_a/s = total energy of the system when ‘a’ fragment on TiO₂ slab, E_s = energy of the TiO₂ slab, E_a(gas) = energy of ‘a’ fragment in gas phase, n = no. of ‘a’ co-adsorbates on the surface.

3 Results and discussions

3.1 Adsorption of intermediates and H₂S on the perfect surface

H-atom adsorption. The most stable conformation on the chosen surface (see Fig. 2(a)) was observed when H-atom stays in perpendicular fashion on O_2c site with H–O_2c bond length of 0.97 Å, which is in agreement with previous investigations.⁴¹ The calculated E_ads (adsorption energies) was measured to be −2.75, −2.41, −2.25 and −0.27 eV on O_2c, O_3c, hollow and Ti_5c (tilted), respectively. For the most stable observation, our computed value of E_ads is 0.62 eV higher than the previously reported value for H on TiO₂ anatase (001).⁴² However, these values are in close agreement with those reported by Hussain and co-authors.⁴³

Fig. 2 Adsorption properties of (a) H-atom, (b) OH, (c) OH (tilted) and (d) S-atom. The cases having maximum adsorption energies are presented in the figure only.

OH adsorption. As the most stable geometry for single H-atom adsorption corresponds with the two way coordinated oxygen (O_2c) of TiO₂. Formation of hydroxyl seems to be more favorable than the formation of hydride on Ti site because H-atom adsorption on Ti_5c was negligibly small. This observation is in agreement with previous investigations.⁴⁴ The OH adsorption calculations yield E_ads of 0.39, 0.07, −1.01, and −1.45 eV for O_2c, hollow, O_3c and Ti_5c, sites, respectively. The E_ads on O_2c and hollow sites is positive indicating that the reaction is endothermic giving a repulsive interaction, while on O_3c and Ti_5c it is negative indicating that the reactions are exothermic. OH having tilted coordination at Ti_5c position is the most stable configuration (−1.85 eV) making an angle (Ti–O–H) of 108° compared to 180° (see Fig. 2(b) and (c)).²⁸ This substantial variation in the adsorption energies on different locations causes hindrance regarding easy diffusion of OH on the surface in diverse directions. Fig. 2 summarizes the most stable adsorption geometries for single H-atom, OH and S-atom on the surface.

S-atom adsorption. It was observed in the previous experimental investigations that by doping TiO₂ with S, the H₂S destruction efficiency of the TiO₂ catalysts⁴⁵ was improved. To validate and confirm our experimental measurements, DFT calculations have been performed. The calculated adsorption energies of single S atom for the four different positions namely Ti_5c, hollow, O_3c and O_2c were found to be −0.91 −1.04, −1.67 and −2.46 eV (Fig. 2(d)), respectively. We show that S interacts significantly with all the tested locations, exhibiting the strongest bonding on O_2c position. In its most stable configuration, O–S distance is 1.72 Å, as shown in Fig. 2(d).

HS adsorption. From our calculations, we found three structures that are energetically more favorable, HS on Ti_5c, O_3c and O_2c sites. Their calculated absorption energies of HS on these sites in respective order are 0.15, −0.29 and −1.26, eV. This shows that adsorption energy varies significantly with the adsorption position. The most stable configuration is HS–O_2c. Thus practically, HS diffusion becomes difficult on the surface. In this case, the HS radical moved from the hollow position towards the O_2c position during the optimization process and got tilted adsorbed on the O_2c site forming an angle O_2c–S–H of 98.3° with a bond length of 1.67 Å while the H–S bond length is 1.35 Å, as shown in Fig. 3(a). It shows that HS on the hollow position was unstable which led to the movement of HS towards O_2c. However, the E_ads calculated for HS was 0.77 eV higher than the value found in a previous study where HS was placed to form a bridge between Ti_5c and O_3c.²²

Fig. 3 Most stable geometry configurations of (a) HS (tilted), (b) H₂S and (c) H₂S–H₂S on TiO₂ surface along with their adsorption energies.

H₂S adsorption. In the case of H₂S interaction with the surface, the most favorable configuration corresponds to the case when H₂S was placed on O_3c. The adsorption energies calculated for the different sites, hollow, O_2c, Ti_5c and O_3c are 2.66, 0.04, −0.13 and −0.17 eV, respectively. The interaction at Ti_5c slightly decreases relative to the most stable configuration while remaining locations interact repulsively. Our calculated value for the adsorption energy is lower than that of −0.32 eV reported for H₂S on Ti_5c site of TiO₂ anatase (101).²² Nevertheless, an agreement with the previous investigation²² regarding weak interaction is reached. The H–S–H bond angle was found to be 93.6 Å with a H–S bond length of 1.35 Å and 2.91 Å for S–Ti_5c as shown in Fig. 3(b). Therefore, the H₂S on O_3c structure can be characterized as physisorbed species with small adsorption energy. Then, it can be concluded that H₂S hardly perturbs this surface.

Additionally, the coverage effect on the adsorption energy of H₂S has also been computed by placing two H₂S molecules in one unit cell. H₂S–H₂S co-adsorption was investigated keeping both H₂S on adjacent Ti_5c positions as well as one H₂S on Ti_5c and the other on O_2c. The adsorption energy per H₂S molecule increased significantly, as compared to the energies obtained by placing single H₂S molecule in the unit cell. E_ads increased from −0.13 to −0.25 eV per molecule when both H₂S were placed on adjacent Ti_5c locations, while from +0.04 to −0.45 eV for second configuration showing significant enhancement due to coverage increase. The dipole–dipole interaction between the polar H₂S molecules might be responsible for this increased stability. The hydrogen bonding might also be involved in stabilizing the system on doubling coverage. As shown in Fig. 3(c), the H₂S molecule on O_2c site is tilted bound making an angle H–S–O of 73.2°. The Ti_5c–S distance is 2.72 Å and the H–S distance is 1.35 Å having an angle H–S–H at 97°. The S–O_2c distance is 3.08 Å. The adsorption energies, bond lengths and angles measurements have been tabulated in Table 1. It can be inferred that H₂S molecules have attractive interaction and the system stabilizes substantially for the second configuration.

Table 1 Different species with their respective adsorption sites and E_ads

Surface	Species	Site	Eads
	H	O2c	−2.75
	OH	Ti5c (tilted)	−1.85
	S	O2c	−2.46
	HS	Hollow	−1.26
	H2S	O3c	−0.17
	H2S–H2S (per molecule)	Ti5c–O2c	−0.47
	S–H2S	O2c–O3c	−0.91
	S–HS	O2c–O3c	−0.9
	H2S + OH	No S, H2S–O2c, OH–Ti5c	−0.33
	H2S + OH-90	No S, H2S–Ti5c, with H2S 90° rotated, OH–O2c	−0.09
	S + H2S + OH	S–O2c, H2S–O2c, OH–Ti5c	−1.54
	S + H2S + OH-90	S–O2c, H2S–O2c, with H2S 90° rotated, OH–O2c	−1.07
Defected surface	H2S	No S, H2S–Ti5c	−0.20
	H2S	No S, H2S–O2c(vac)	−0.27
	H2S	No S, H2S–O3c	−0.31
	S + H2S	S–O2c, H2S–O2c(vac)	−0.21
	S + H2S	S–O2c, H2S–Ti5c	−0.24
	S + H2S	S–O2c, H2S–O3c	−0.96
	S + H2S + OH	S–O2c, H2S–Ti5c, OH–O2c	−4.64
	S + H2S (90°) + OH	S–O2c, H2S–O2c(vac), with H2S 90° rotated, OH–O2c	−2.67
	H2S (90°) + OH	No S, H2S–O2c(vac), with H2S 90° rotated, OH–O2c	−2.25
Perfect surface TiO2 (p2 × 2)	S + HS + H(l)	S–O2c, one H of H2S–Ti5c on left O2c	−1.49
	S + HS + H(r)	S–O2c, one H of H2S–Ti5c on right O2c	−1.49
	S + S + H + H	S–O2c, one H each of H2S–O3c on left O2c and on right O2c
Defected surface	S + HS + H(lt)	S–O2c, one H of H2S–Ti5c on left O2c(vac)	−1.6
	S + HS + H(rt)	S–O2c, one H of H2S–Ti5c on right O2c	−1.6
	HS (90°) + H2O	No S, H2S–O2c(vac), with H2S 90° rotated and one H of H2S added to OH–O2c	−3.62
	S + HS (90°) + H2O	S–O2c, H2S– O2c(vac), with H2S 90° rotated one H of H2S added to OH–O2c	−3.79

---

Effect of S addition on the adsorption energy of H₂S and HS. To study the effect of S-atom co-adsorbed with H₂S and HS, we placed S on its most stable O_2c site while H₂S and HS were placed on O_3c site, respectively and two separate calculations were performed. However, during the process of optimization, both H₂S and HS moved slightly and settled as shown in Fig. 4. The most stable co-adsorption configurations were achieved with E_ads of H₂S and HS −0.91 eV and −0.90 eV, respectively. It is worth mentioning here that addition of S has significantly stabilized H₂S and HS as their respective E_ads increased from −0.17 to −0.91 eV and from −0.29 to −0.90 eV, respectively. Other combinations of different available sites for co-adsorption of S with H₂S and HS were also considered but were found substantially less important. Fig. 4 summarizes the most relevant adsorption geometries. It should also be noted that system with HS on the Ti_5c yielded to an endothermic adsorption of 1.60 eV but gave E_ads of −0.33 eV with co-adsorbed S. These findings are in good agreement with our previous experimental results where S-doping increases the H₂S destruction efficiency.⁴⁵ These results allow us to conclude that presence of S on the TiO₂ surface enhances the H₂S and HS interactions significantly.

Fig. 4 Adsorption properties of H₂S and HS systems in the co-adsorbed state with S-atom are given in part (a) and (b), respectively.

Effect of OH addition on the adsorption energy of H₂S. The co-adsorption of OH and H₂S was investigated by placing OH on its most stable Ti_5c location and H₂S also on the similar adjacent site with two different orientations (see Fig. 5(a) and (b)). It has been observed by some authors that presence of moisture in the system has a positive effect on the H₂S adsorption.^46,47 As shown in Fig. 5(a), it is observed that adding OH has a favorable effect on E_ads of H₂S which increases from −0.13 to −0.33 eV. This was not the case for the other configuration (see Fig. 5(b)) where the effect was not pronounced. By observing the position of H-atom as shown in Fig. 5(a), the increase in adsorption energy of H₂S could be attributed to presence of attractive forces between H and OH radical present nearby. The increase in d_S–H (1.42 Å) bond indicate the attractive interaction between O and H (of H₂S) atoms leading to stabilize the system. This behavior becomes evident from Fig. 5(b) where H₂S co-adsorbs with OH with the respective H-atoms of H₂S tilted towards OH. The geometrical information regarding these co-adsorption configurations can be seen in Fig. 5(a) and (b).

Fig. 5 Adsorption properties of H₂S in presence of OH. In (a) H of H₂S points towards O-atom of OH. In (b) H₂S is rotated at 90° and OH stands rotated at 180° relative to its position indicated in part (a).

Effect of OH addition on the adsorption energy of H₂S in presence of S. In previous sub-sections we have shown that the interaction between H₂S with surface increased due to S and OH addition. Here, we computed the consequences of S insertion in the systems studied in previous sub-section i.e., co-adsorption of S and OH simultaneously with H₂S. Adding S caused significant rise in H₂S adsorption energy. As two cases shown in Fig. 6(a) and (b), the H₂S adsorption energy increased from −0.34 to −1.54 eV and from −0.09 to −1.07 eV, respectively. In the first case (see Fig. 6(a)), the H₂S position is almost similar as explained in the previous sub section, showing the OH affinity for the H atom. The H₂S molecule is tilted towards the OH with one of its H-atoms pointing in the direction of OH. The S–Ti_5c distance is 2.78 Å and H₂S is comparatively less tilted in this case forming an angle S–Ti_5c–O_2c of 79.1°. The H–S–H angle is 93.4° and H₂S forms the tilted geometry with the surface, making angles H–S–Ti_5c (H towards OH) and H–S–Ti_5c as 92° and 100.7°, respectively. The H–S distance also increases from 1.35 to 1.41 Å due to attraction between H and OH. The distances O–Ti_5c and O–H are 1.92 and 0.97 Å, respectively. The O–H–Ti_5c angle is 113°. S-atom is also tilted towards Ti_5c making an angle of O_2c–S–Ti_5c of 65.6° and forms a bridge between O_2c and Ti_5c, while the S–O_2c bond length is 1.6 Å.

Fig. 6 Adsorption geometries of H₂S in presence of OH and S. In (b) part H₂S is rotated at 90° relative to its orientation in part (a) in addition to change in position of S and OH direction where O–H stands rotated at 180°.

In the second case where H₂S was rotated at 90° as shown in Fig. 6(b), we observe a considerable difference in geometries if we compare to the same system in the absence of S as discussed previously. Both of H atoms in H₂S and OH are tilted towards each other. The S–Ti₅ distance is 2.91 Å and in this case the H atoms of H₂S became almost parallel to the surface having angles H–S–Ti_5c of 91.76° for both the H atoms, while H₂S is not tilted at all. The H–S–H angle is 91.4° and H–S distance of 1.36 Å for both H atoms. The distance O–Ti_5c and O–H is 1.87 and 0.97 Å, respectively. The O–H–Ti_5c angle is measured to be 113.2°. S-atom is not tilted in this case and S–O_2c distance is 1.63 Å.

It can be seen that in this system, the rotation of H₂S by 90° had a negative effect on the H₂S adsorption energies as it decreased from −0.33 to −0.09 eV without S and from −1.54 to −1.07 eV in the presence of S. It can be understood that H₂S rotation is not favored in this system while the previously mentioned configuration is the most stable one. In addition, the lesser distance between S of H₂S and the Ti_5c in the former case can be attributed to a stronger adsorption. Finally, the most important is the OH orientation, which allows us to conclude that pointing of H of H₂S towards O of OH plays key role in the system stability.

3.2 Defect (in TiO₂ surface) creation effect on adsorption energies of various adsorbents

In order to explore the effect of vacancies on H₂S adsorption or dissociation, a defect was created in the lattice of (2 × 2) TiO₂ unit cell by removing a two way coordinated O_2c atom, as shown in Fig. 7(a). Different calculations were carried out for various systems as discussed below.

Fig. 7 Different geometries of H₂S adsorption in the presence of vacancy; (a) slab, (b) H₂S on O_3c, (c) H₂S on Ti_5c and (d) H₂S on O_2c (vacancy site).

Effect on H₂S adsorption. H₂S was placed on three (O_3c, Ti_5c and O_2c) positions in the presence of nearby vacancy, as shown in Fig. 7. The E_ads for Ti_5c, O_2c-vac and O_3c were found to be −0.20, −0.24 and −0.31 eV, respectively. Relative to perfect system, the adsorption energy increased from −0.13 to −0.20 eV for Ti_5c site and from −0.17 to −0.31 eV at O_3c site, while the system became exothermic (+0.04 to −0.24 eV) from endothermic when H₂S is placed on O_2c. Hence, the defected surface, to some extent, is found more reactive towards H₂S adsorption. The distance S–Ti_5c is found to be 2.69 Å, angle H–S–H is 96° and H₂S is slightly tilted from the geometry when H₂S is placed on Ti_5c. The H–S–H angle is found to be 92.7° for O_2c-vac position and S–O_3c distance is found to be 3.0 Å. H–S–H angle is 93.6° for H₂S on O_3c position. The H–S bond length of 1.35 Å remains the same for all the H₂S positions.

Effect on H₂S adsorption in the presence of S. As presented in Fig. 8, the S is placed on O_2c while H₂S is kept on O_3c, Ti_5c and O_2c-vac in three different calculations. The calculated adsorption energies were measured as −0.21, −0.24 and −0.96 eV on O_2c-vac, Ti_5c and O_3c sites, respectively. As discussed above these values of E_ads on the respective locations were −0.20, −0.24 and −0.31 eV in the absence of S. Thus, the presence of S gave a very minute effect on the H₂S adsorption energy except for the configuration where H₂S binds with O_3c (i.e. E_ads increased from −0.31 to −0.96 eV). However, the trend was almost similar when compared to the perfect system explained in previous section. The geometrical data in this regard were presented in Fig. 8.

Fig. 8 Different geometries in presence of vacancy and S, (a) H₂S on O_3c, (b) H₂S on Ti_5c and (c) H₂S on O_2c (vacancy site).

Effect on H₂S adsorption in the presence of OH and S. Calculations were performed by adding OH in the system discussed in the previous subsection for the cases depicted in Fig. 9(a) and (b). The purpose is to compare perfect and defected systems towards H₂S interaction and destruction. Initially, OH and H₂S were occupying adjacent Ti_5c sites with O_2c vacancy in between them while S absorbs on O_2c. A huge increase in adsorption energy of H₂S (−3.52 eV) was noticed whereas in the latter configuration shown in Fig. 9(b), this value was found to be −2.67 eV. This difference in the E_ads of H₂S for the two configurations can again be attributed to the H₂S molecule and OH orientations; more stabilization resides with lesser O–H (of H₂S) distance.

Fig. 9 Adsorption properties of H₂S on defected surface in the presence of S and OH depicted in part (a) and (b). In (b) H₂S stands rotated at 90° relative to (a), while in (c) S-atom has been removed.

Moreover, substantial stability of the system on OH addition in the presence of vacancy agrees with our previous findings on perfect surface. Here the H-atom gets tilted towards OH as it was noted in the case of perfect surface. To evaluate the actual cause of increase in the H₂S adsorption energy, further computations were carried by removing OH from the system, as depicted in Fig. 9(b). It was found that the H₂S adsorption energy decreased to almost fifty percent i.e., −1.31 eV as compared to −2.67 eV when OH was present in this system. It can therefore be deduced from these findings that OH addition has a substantial effect towards the adsorption energy of H₂S. The effect becomes quite pronounced in the presence of vacancy. In agreement with our findings on perfect surface, orientation of the molecule and OH is of crucial importance.

Just to further ascertain these findings, calculations were also carried out without S-atom. H₂S molecule rotated at 90° relative to the system given in Fig. 9(a) (as shown in Fig. 9(c)) yielded lower E_ads of H₂S (decreased from −2.67 to −2.25 eV). However, this decrease is in line as observed on the perfect surface. In addition, it highlights the significant rise in the E_ads of H₂S which was caused by addition of OH. This very high E_ads of H₂S may cause poisoning of the catalytic surface. Hence, OH radicals must be avoided on the defected surface to save the catalyst from H₂S poisoning. Detailed geometries have been shown in Fig. 9.

3.3 Dissociation of H₂S

Looking at the affinity of OH towards H-atom of H₂S as found in these computations, it seems to be more likely that when H₂S will dissociate in the presence of OH, one H-atom of H₂S will join OH to form water. The investigations of this hypothesis were carried out on perfect and defected surface by placing HS with OH in the system in the presence of S.

H₂S dissociation on perfect surface. Initially, H₂S dissociation on the perfect surface was computed. The energy of the co-adsorbed HS + H fragments was evaluated and found that the system was slightly destabilized by 0.12 eV relative to the H₂S adsorption level. Both lower value of E_ads of H₂S and endothermicity on the perfect surface lead us to conclude that the H₂S molecule may desorb before heading towards disintegration. It has been established in the previous sections that the adsorption energy of H₂S substantially enhances when OH was co-adsorbed with the molecule. It was also confirmed in previous investigations,^22,26 that the presence of water on the anatase TiO₂ (001) spontaneously lead to its dissociation and thus made OH radical available on the surface. Therefore, NEB calculations were performed on those systems where adsorption energy of H₂S was sufficient e.g., where OH was co-adsorbed in the system. The NEB calculation of the co-adsorbed system of OH and H₂S predicted a negligibly small activation barrier of 0.03 eV. The transition state is visible in the energy profile for the dissociation of H₂S, as shown in Fig. 10. In the transition state, the H₂S has moved towards OH from its normal adsorbed position. Here, S–H bond activation takes place by stretching the bond length up to 1.45 Å as compared to normal S–H bond distance of 1.36 Å computed for gas phase H₂S. On the other hand tilted OH remains almost stable upon its position. The approaching H-atom is 1.4 Å distant from O-atom in the transition state. The transition state has been confirmed by the unique imaginary frequency of 433 cm⁻¹. The formation of H₂O and HS is thermodynamically favorable by −0.67 eV. Such a small activation barrier and exothermicity suggest that H₂S destruction on anatase (001) is quite facile in the presence of OH species and in agreement to the findings of our experimental studies.⁴⁵

Fig. 10 Energy profile for the activation of H₂S on anatase (001).

H₂S dissociation on defected surface. The defected system containing OH and S co-adsorbed with H₂S was also investigated. The H₂S was placed on O_2c-vac location and OH on Ti_5c. It is worth mentioning that H₂S is chemisorbed with sufficient adsorption energy. During the optimization process, it was found that one H-atom was detached from H₂S and made a bond with OH forming H₂O. Initially H-atom of H₂S was 1.78 Å away from the O atom of OH radical. In the final situation OH bond length (H atom which detached from H₂S) was 1.04 Å.

Additionally HS fragment of H₂S moved to stay on Ti_5c with HS (H atom which has been detached from H₂S) distance of 1.89 Å. The resulting rotated HS is tilted making an angle of H–S–Ti_5c of 94.2°. The decomposition of H₂S and formation of H₂O also lead to some structural changes in the co-adsorbed S and associated O_2c on the surface. The detailed adsorption energies along with all the geometries are presented in Table 1S (ESI†). Thus, in the presence of vacancy, H₂S decomposition is a spontaneous process. The negligible activation barriers of 0.05 eV were found in our experimental observations.⁴⁸ Thus, our DFT results are in complete agreement with the experimental findings for the S-doped nanofibers with highest H₂S destruction efficiency.⁴⁵

4 Conclusions

Interaction of H₂S, its dissociated products and OH radical were studied on anatase phase of TiO₂ (001) surface implying DFT calculations at GGA level of approximation. The aim of the present work is to optimize H₂S adsorption and study its destruction on the surface inspired by the experimental observations carried out in our previous work.⁴⁵ The perfect surface weakly interacts with H₂S. However, the presence of S on the surface led to a strong H₂S binding on a particular position O_3c by giving an adsorption energy of −0.91 eV. Increase in H₂S coverage also helped to stabilize the molecules to some extent. Defect creation in the surface further stabilized the molecule slightly. Nevertheless, the most significant impact was experienced due to the presence of OH radicals in the vicinity, which caused the adsorption energy of H₂S to enhance for both the perfect and defected surfaces. S, HS, OH species interact sufficiently strongly with this surface. The presence of OH radicals on the surface help to decompose H₂S into HS and H₂O (H₂S + OH → HS + H₂O) instantaneously. Thus anatase (001) surface can readily destroy H₂S species if moisture is present in the system. On the other hand, it was reported in a recent study that hydroxylated anatase (001) leads to Ti³⁺ formation from the Ti–O bond cleavage. This is beneficial for reactions of H₂O and O₂ in addition to its prime favorability for ethanol dissociation.^41,49 Therefore, further studies should be carried out to explore the effects of Ti³⁺ for H₂S adsorption as an extension to this study.

Conflict of interest

No competing interests in the publication of this paper.

Author's contribution

NS carried out the experimental studies and the computational studies as well. He was involved in preparing the draft and final manuscript. AH conceived the study, and participated in its basic understanding by the author to carry out this study. He also participated in helping the author to draft this manuscript and final vetting of the manuscript. NM, MBK and SGS helped the authors for the preparation of draft manuscript. All authors read and approved the final manuscript.

Acknowledgements

Authors from Pakistan gratefully thank the National Centre for Physics (NCP), Islamabad, Pakistan, Nanoscience and Catalysis division and DNE, PIEAS, Islamabad labs and facilities for carrying out this study.

References

1. M. J. Pearson, Ind. Eng. Chem. Prod. Res. Dev., 1977, 16, 154–158.
2. Z. M. George, Phosphorus Sulfur Relat. Elem., 1976, 1, 315–322.
3. E. Laperdrix, G. Costentin, N. Nguyen, F. Studer and J. C. Lavalley, Catal. Today, 2000, 61, 149–155.
4. A. Fujishima, T. N. Rao and T. A. Tryk, J. Photochem. Photobiol., C, 2000, 1, 1–21.
5. A. G. Agrios and P. Pichat, J. Appl. Electrochem., 2005, 35, 655–663.
6. M. Lewandowski and D. F. Ollis, in Seminoconductor photochemistry and photophysics, ed. V. Ramamurthy and K. S. Schanze, Marcel Dekker, New York, 2004, pp. 249–282.
7. P. Pichat, in Chemical degradation methods for wastes and pollutants: environmental and industrial applications, ed. M. A. Tarr, Marcel Dekker, New York, 2003, pp. 77–119.
8. H. Ibrahim and H. Lasa, Appl. Catal., B, 2002, 38, 201–213.
9. J. Zhang, A. Terukazu, M. Minagawa, K. Kinugawa, H. Yamashita, M. Matsuoka and M. Anpo, J. Catal., 2001, 198, 1–8.
10. A. Mills and S. Le Hunte, J. Photochem. Photobiol., A, 1997, 108, 1–8.
11. A. L. Linsebigler, G. Lu and J. T. Yates Jr, Chem. Rev., 1995, 95, 735–743.
12. A. Fujishima and X. Zhang, C. R. Chimie, 2006, 9, 750–760.
13. M. C. Canela, M. A. Rosana and W. F. Jardim, J. Photochem. Photobiol., A, 1998, 12, 73–80.
14. N. Shahzad, S. T. Hussain, A. Siddiqua and M. A. Baig, J. Nanosci. Nanotechnol., 2012, 12(6), 5061–5065.
15. N. Shahzad, S. T. Hussain and A. Nisar, Chalcogenide Lett., 2013, 10(1), 19–26.
16. N. Shahzad and C. Qiao, Mater. Sci. Forum, 2013, 756, 225–230.
17. L. Min, Z. Jun-ying, Z. Yue, Z. Guo-feng and W. Tian-min, Appl. Surf. Sci., 2011, 257, 10710–10714.
18. C. Shenghui, S. Shuangqing, L. Bingjie, M. Yunfei, Y. Youguo and H. Songqing, Surf. Sci., 2014, 620, 51–58.
19. L. Lixia, Z. Riguang, H. Peide and W. Baojun, Fuel Process. Technol., 2013, 106, 222–230.
20. D. G. Masoud, S. Narges, A. Mahdi and G. A. Morteza, Appl. Surf. Sci., 2012, 261, 697–704.
21. T. V. Reshetenko, S. R. Khairulin, Z. R. Ismagilov and V. V. Kuznetsov, Int. J. Hydrogen Energy, 2002, 27, 387–394.
22. W. F. Huang, H. T. Chen and M. C. Lin, J. Phys. Chem. C, 2009, 113, 20411–20420.
23. M. R. Ranade, A. Navrotsky, H. Z. Zhang, J. F. Banfield, S. H. Elder, A. Zaban, P. H. Borse, S. K. Kulkarni, G. S. Doran and H. J. Whitfield, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 6476–6481.
24. A. S. Barnard and P. Zapol, Phys. Rev. B: Condens. Matter Mater. Phys., 2004, 70, 235403.
25. A. S. Barnard, P. Zapol and L. Curtiss, Surf. Sci., 2005, 582, 173–188.
26. C. Arrouvel, M. Digne, M. Breysse, H. Toulhoat and P. Raybaud, J. Catal., 2004, 222, 152–166.
27. M. Lazzeri, A. Vittadini and A. Selloni, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 63, 155409.
28. H. Akhtar, PhD thesis, Eindhoven University of Technology, Netherlands, 2010.
29. X. Q. Gong and A. Selloni, J. Phys. Chem. B, 2005, 109, 19560–19562.
30. A. Vittadini, M. Casarin and A. Selloni, Theor. Chem. Acc., 2007, 117, 663.
31. G. Kresse and J. Hafner, Phys. Rev. B: Condens. Matter Mater. Phys., 1993, 47(1), 558–561.
32. J. P. Perdew and Y. Wang, Phys. Rev. B: Condens. Matter Mater. Phys., 1992, 45(23), 13244–13249.
33. P. E. Blöchl, Phys. Rev. B: Condens. Matter Mater. Phys., 1994, 50(24), 17953–17979.
34. G. Kresse and D. Joubert, Phys. Rev. B: Condens. Matter Mater. Phys., 1999, 59(3), 1758–1775.
35. H. J. Monkhorst and J. D. Pack, Phys. Rev. B: Condens. Matter Mater. Phys., 1976, 13(12), 5188–5192.
36. M. Horn, C. F. Schwerdt and E. P. Meagher, Z. Kristallogr., 1972, 136(3–4), 273–281.
37. M. Methfessel and A. T. Paxton, Phys. Rev. B: Condens. Matter Mater. Phys., 1989, 40(6), 3616–3621.
38. P. Pulay, Chem. Phys. Lett., 1980, 73(2), 393–398.
39. J. D. Head, Int. J. Quantum Chem., 1997, 65(5), 827–838.
40. G. Henkelman, B. P. Uberuaga and H. Jonsson, J. Chem. Phys., 2000, 113(22), 9901–9904.
41. Z. Riguang, L. Zhixue, L. Lixia and W. Baojun, Appl. Surf. Sci., 2015, 353, 150–157.
42. F. Lin, G. Zhou, Z. Li, J. Li, J. Wu and W. Duan, Chem. Phys. Lett., 2009, 475, 82–85.
43. A. Hussain, J. Gracia, B. E. Nieuwenhuys and J. W. Niemantsverdriet, ChemPhysChem, 2010, 11(11), 2375–2382.
44. M. Menetrey, A. Markovits and C. Minot, Surf. Sci., 2003, 524(1–3), 49–62.
45. N. Shahzad, PhD thesis, National University of Sciences and Technology (NUST), Pakistan, 2010.
46. M. E. Amani, J. Jacek and J. B. Teresa, J. Mater. Chem. A, 2015, 3, 8194–8204.
47. O. Mabayoje, M. Seredych and T. J. Bandosz, ACS Appl. Mater. Interfaces, 2012, 4(6), 3316–3324.
48. N. Shahzad and A. Nisar, Chalcogenide Lett., 2015, 12, 129–136.
49. Y. Wang, H. J. Sun, S. J. Tan, H. Feng, Z. W. Cheng, J. Zhao, A. D. Zhao, B. Wang, Y. Luo, J. L. Yang and J. G. Hou, Nat. Commun., 2013, 4, 2214.

---

Footnotes

† Electronic supplementary information (ESI) available: Additional data file is a table giving detailed adsorption energies along with all the geometries and presented as Table 1S. See DOI: 10.1039/c5ra20875k

‡ NS is working as Assistant Professor at National University of Sciences and Technology, Islamabad, Pakistan.

---