Interfacial Action of Co-Doped MoS₂ Nanosheets on Directional Piezoelectric Catalytic Generation of Reactive Oxygen Species ^†

Abstract

Molybdenum disulfide (MoS₂) with single- and odd-numbered layers is a novel piezocatalyst, and its piezocatalytic molecular oxygen activation is considered a promising and low-cost strategy for environmental remediation. In this study, the odd-numbered layers of Co-doped MoS₂ ultrathin nanosheets were successfully fabricated, which decomposed tetracycline by 99.8% in 15 min through shaking vibration. Moreover, to verify the enhanced piezoelectric catalytic activity of MoS₂ via the doping effect, molecular oxygen activation properties were predicted through DFT calculation and monitored by generated reactive oxygen species (ROS) evolution. In addition, the primary reactive species responsible for the degradation of tetracycline pollutants were also investigated in detail.

1. Introduction

The activation of molecular oxygen to generate ROS such as hydroxyl radicals (•OH), superoxide radicals (•O₂⁻), singlet oxygen (¹O₂), and hydrogen peroxide (H₂O₂) is an effective method for resistant organic pollutant breakdown and sterilization. Although it is a feasible strategy, its spin-forbidden nature hinders its activation and applications [1]. Recently, piezocatalytic molecular oxygen activation has been proposed. In this process, electrons and holes migrate toward the two sides of the catalyst surface and accumulate on the active sites due to a piezoelectric field induced by vibration energy, and then, the piezocharges react with molecular oxygen from dissolved oxygen or water to produce ROS [2]. Two-dimensional MoS₂ piezoelectric catalysts have emerged as novel participants in the piezoelectric catalytic molecular oxygen activation reaction. Due to their non-centrosymmetric structure, single- and odd-numbered layers of MoS₂ exhibit a high piezoelectric effect and can provide spatially distinct orientations of spontaneous polarization, as initially reported in 2014 [3]. However, molecular oxygen does not conduct through MoS₂ because of its high band gap, poor inter-domain electron transport, and restricted catalytic active sites on the basal plane [4]. Foreign-element doping has proven to be a successful method of controlling the electronic structure of semiconductors, leading to higher catalytic activity in MoS₂ as a result of the lattice distortion of atomic configurations [5]. Herein, because of a finite piezoelectric response during the piezocatalysis process, Co doping into odd layers of the MoS₂ structure produced an additional catalytic active center in the Mo-S network and allowed for higher conductivity, which, in turn, enabled increased molecular oxygen adsorption energy and accelerated charge transfer. Moreover, a molecular oxygen activation mechanism is also proposed using DFT calculation and ROS detection experiments, and then, ultrafast piezoelectric catalytic antibiotic pollutant degradation activities are investigated as well.

2. Experimental

2.1. Preparation of MoS₂ and Co-Doped MoS₂ Nanosheets

In a typical synthesis, 0.20-mole ((NH₄)₆Mo₇O₂₄·4H₂O), 0.05-mole, 0.10-mole, and 0.15-mole Co(OAc)₂·4H₂O and 0.40-mole CS(NH₂)₂ were dissolved into 10 mL of deionized water. The solution was transferred to a Teflon-lined autoclave with a 25 mL capacity and heated at 180 °C for 12 h. The resultant black products were washed several times with distilled water and ethanol and finally dried at 60 °C for 12 h. The samples obtained with different Co dopant amounts are referred to as Co-MoS₂-0.05, Co-MoS₂-0.10, and Co-MoS₂-0.15.

2.2. Piezoelectric Catalytic Tetracycline Degradation and Electrical Energy Consumption

The hydromechanical stress initiated by shaking vibration might be a feasible strategy to apply a sensitive force to piezoelectric materials. Thus, in the present study, a shaker was introduced to simulate hydromechanical stress on a piezoelectric catalyst through mechanical vibration, which triggered the piezoelectric effect in the degradation of tetracycline pollutants. The details of the experimental design of the shaking-vibration-induced piezoelectric catalytic tetracycline pollutant degradation are set up and described in the Supplementary Materials. In addition, the input energy consumption per order (EE/O_in) was used to compute the EE/O values of various catalysts and is also described in the Supplementary Materials.

2.3. Quantitation of Radical Concentration

The steady-state concentration of ROS ([ROS]_ss), which represents the radical generation efficiency (ROS), was computed indirectly by monitoring the degradation of a radical probe. [•O₂⁻]_ss and [•OH]_ss are the steady-state concentrations of •O₂⁻ and •OH, which can be calculated from the degradation kinetics of nitrotetrazolium blue chloride (NBT) and terephthalic acid (TA) in a system. The detailed information is described in the Supplementary Materials.

3. Results and Discussion

3.1. Material Characterization

Morphology images of MoS₂ and different Co-doped MoS₂ samples are shown in Figure 1a–d. The samples were composed of numerous netlike curled ultrathin sheets. Notably, no other particles were detected in low-content amounts of Co dopant (e.g., Co-MoS₂-0.05 and 0.10) (Figure 1b,c). However, with an increasing Co doping amount in MoS₂ (e.g., Co-MoS₂-0.15), the morphology slightly changed into particulate aggregates, shown as red arrows in Figure 1d. In Figure 1e, the Co-doped samples also exhibited the 2H phase of MoS₂, whose XRD patterns agreed well with that of the reference MoS₂. The patterns were basically unchanged during doping, suggesting that the 2H phase was not affected by the low amount of Co incorporation. However, except for the diffraction peaks of MoS₂, some other diffraction peaks in the XRD patterns of the as-obtained Co-MoS₂-0.15 matched well with CoS₂ (JCPDS No. 041-1471). According to the solid solution, Co atoms did not tend to bond to Mo atoms; then, separate CoS₂ and MoS₂ phases formed with limited miscibility. As an example, Li and coworkers discovered that Sr doping in a LaGaO₃ crystal lattice produced more Sr-rich secondary phases, showing that the formation of secondary phases in LaGaO₃ strongly depends on the number of dopants [6]. The calculated thicknesses of Co-MoS₂-0.05, Co-MoS₂-0.10, and Co-MoS₂-0.15 were ca. 3.2 nm, 3.5 nm, and 3.7 nm according to Debye–Scherrer’s equation. Considering that the c parameters of 2H-MoS₂ are 6.4 Å, the calculated atomic number layers of the Co-doped MoS₂ samples were 5.4, 5.1, and 5.6 for Co-MoS₂-0.05, Co-MoS₂-0.10, and Co-MoS₂-0.15, as shown in

Table 1. Different thicknesses and layer numbers of MoS₂ and Co-MoS₂.

No.	Sample Name	Temp (°C)	Time (h)	Thickness (nm)	Layer No.
1.	MoS2	180	12	3.2	~5
2.	Co-MoS2-0.05	180	12	3.2	~5
3.	Co-MoS2-0.10	180	12	3.5	~5
4.	Co-MoS2-0.15	180	12	3.7	~5

. Similar phenomena have been reported by Yan et al. [7]. As shown in EDX mapping images of the Co-MoS₂-0.10 catalyst, a homogeneous distribution of Mo, S, and Co elements indicated that Co atoms were uniformly distributed throughout the whole MoS₂ matrix (Figure 1f). The SEM images and XRD patterns of the bulk MoS₂ are shown in Figure S1. X-ray photoelectron spectroscopy (XPS) was used to examine the elemental composition and oxidation states of MoS₂ and Co-MoS₂-0.10, and the results are displayed in Figure S2 and described in the Supplementary Materials.

In the 2H MoS₂ with a single- or odd-numbered layer, the centrosymmetric structure is broken, resulting in in-plane intrinsic piezoelectricity [8]. The origin of the layer-dependent piezoelectric effect in this as-prepared Co-MoS₂-0.10 was also investigated using the HR-TEM technique. The TEM images showed that Co-MoS₂-0.10 was composed of numerous ripples, thin nanosheets, and active edge sites (Figure 2a,b). As can be seen from the side view of the HRTEM image in Figure 2c, the crystal fringes along the nanosheet edge were not perfect and apparently nonconsecutive, which was ascribed to the fact that the more exposed interior atoms inevitably induced the formation of crystallographic defects on the surfaces. On the basis of the crystal cells of MoS₂ and the atomic arrangement of the (002) plane, it was calculated that Co-MoS₂-0.10 had a 5-crystal-cell-unit atomic thickness corresponding to the c parameter of MoS₂ (6.4 Å), and the results were consistent with the calculated thickness and approximated layer numbers, as described Table 1.

In addition, to further evaluate the thickness of the Co-MoS₂-0.10 ultrathin nanosheets, an image of the tapping mode of AFM (Figure 2d) was taken. Considering that the c parameters of MoS₂ are 6.4 Å, 3.4 nm, and 4.5 nm in thickness, this showed that the Co-MoS₂-0.10 nanosheets were composed of five and seven numbered layers. Notably, a broader and shifted intensity in the Raman band in a range of 360–430 cm⁻¹ with a shorter distance was observed in Co-MoS₂-0.10 compared with the bulk MoS₂, as shown in Figure 2e (asterisks shown Figure 2e represent for the intensity in a range of 360–430 cm⁻¹ in Co-MoS₂-0.10). Doping-induced S vacancy in the abundant single- and few-layered characteristics of MoS₂ was seen in Co-MoS₂-0.10 [5,7].

As illustrated in Figure 2f, the Mott–Schottky plots of the bare MoS₂ and Co-MoS₂-0.10 samples showed a positive slope, which represents a typical n-type semiconductor. The flat-band potential of Co-MoS₂-0.10 was calculated to be −0.41 V versus the Ag/AgCl electrode and was equal to −0.17 V versus the normal hydrogen electrode (NHE). Compared with bare MoS₂ (inset Figure 2f), increased flat potential, i.e., Fermi level, was observed in Co-MoS₂-0.10 because the impurity level and Fermi level were both closer to the conduction band [8]. As explained earlier, the increased flat band potential leads to the increased separation of free charge carried in the piezocatalysts, which actively participate in the redox reactions under the piezoelectric potential exerted by the mechanical stress [9].

3.2. Piezoelectric Catalytic Tetracycline Degradation and Electrical Energy Consumption

According to the catalytic degradation experiments, there were no obvious changes in the absence of a catalyst, showing that the self-degradation of tetracycline was negligible, as shown in Figure 3a. Firstly, under ultrasonic vibrations in the dark, Co-MoS₂-0.10 showed 98.23% piezocatalytic degradation efficiency, while Co-MoS₂-0.05 and Co-MoS₂-0.15 attained 75.28% and 55.28% piezocatalytic degradation efficiency in 15 min (Figure 3a). As a result, suitable Co-doping amounts (Co-0.10 mole) provided the best performance. The pseudo-first-order kinetic characteristics of the piezoelectric catalytic process are illustrated in Figure 3b. Furthermore, the rate constants of only the shaking process (k = 0.004 min⁻¹) and only the catalysts (k = 0.009 min⁻¹) were lower than that of the bulk MoS₂/shaking (k = 0.011 min⁻¹), Co-MoS₂-0.15/shaking (0.038 min⁻¹), bare MoS₂/shaking (0.07 min⁻¹), Co-MoS₂-0.05/shaking (0.12 min⁻¹), Co-MoS₂-0.10/shaking (0.18 min⁻¹). The pseudo-first-order rate constant of the Co-MoS₂-0.10/shaking system was 0.18 min⁻¹, 16.36 times, 4.74 times, 2.57 times, and 1.5 times higher than the value of bulk MoS₂/shaking, Co-MoS₂-0.15/shaking, MoS₂/shaking, and Co-MoS₂-0.05/shaking systems. In addition, removal efficiencies of up to 91.81% for COD and 82.11% for TOC were achieved in the Co-MoS₂-0.10/shaking system. As already mentioned, the bare MoS₂ and Co-doped MoS₂ samples exhibited the same odd-numbered layer. As a result, a suitable Co-doping amount (Co-0.10 mole) provided the best performance, indicating that the increased catalytic active sites of MoS₂ generated piezoelectric current to promote the participation of piezopolarization in the catalytic performance. In addition, a suitable dopant amount tends to increase electron–hole separation, resulting in the highest catalytic activity. Obviously, increased cobalt loading in MoS₂ leads to detrimental results. A large number of CoS₂ nanoparticles aggregated each other and tended to form the morphology of crumbled bulk balls on the surface of MoS₂ (shown in Figure 1d), which reduced the catalytically active sites of MoS₂. Therefore, Co-MoS₂-0.15 showed the lowest catalytic activity compared with that of Co-MoS₂-0.05 and Co-MoS₂-0.10. Increased cobalt loading in MoS₂ leads to detrimental results.

Table 2. The EE/O values of different catalysts during shaking vibrations.

No	Catalyst/Shaking	k (min−1)	P (kW) × 10−6	V (L)	EE/O (kWhm−3 order−1)
1.	Bulk MoS2	0.011	177.02	0.01	851.01
2.	Co-MoS2-0.15	0.038	177.02	0.01	28.16
3.	Bare MoS2	0.070	177.02	0.01	11.78
4.	Co-MoS2-0.05	0.120	177.02	0.01	3.85
5.	Co-MoS2-0.10	0.180	177.02	0.01	0.63

shows the EE/O values of the bulk MoS₂/shaking (851.01 kWhm⁻³ order⁻¹), Co-MoS₂-0.15/shaking (28.16 kWhm⁻³ order⁻¹), bare MoS₂/shaking (11.78 kWhm⁻³ order⁻¹), Co-MoS₂-0.05/shaking (3.05 kWhm⁻³ order⁻¹), Co-MoS₂-0.10/shaking (0.63 kWhm⁻³ order⁻¹) systems. The Co-MoS₂-0.10/shaking system produced the lowest EE/O value (0.63 kWhm⁻³ order⁻¹), which would save energy by over 18.69 times compared with that of the bare MoS₂, making it the most energy-efficient system with reasonably high tetracycline removal efficiency. As a result, piezocatalysis is a promising new technology for wastewater treatment that uses low energy consumption and is a highly efficient method for tetracycline degradation.

3.3. Theoretical and Experimental Insights into Molecular Oxygen Activation Reaction

The underlying mechanism of Co-dopant-modulated molecular oxygen activation was investigated using DFT calculation (Figure 4a). The calculated adsorption energy of O₂ molecules on pristine MoS₂ was −0.102 eV. However, Co atoms embedded in MoS₂ increased the larger adsorption energy to −1.635 eV between MoS₂ and O₂, indicating that the O₂ adsorption capacity could be significantly enhanced. As illustrated in Figure 4a(iii), O₂ molecules tend to adsorb on the central Co site on MoS₂ nanosheets and are energetically favorable to the lying-on configuration in Co dopant sites, resulting in the formation of two Co–O bonds with bond lengths (d_Co–O) of around 1.88 Å. The amount of charge transfer, as shown by a Bader charge analysis, was 0.553 e⁻ from Co-MoS₂ to adsorbed O₂ molecules. Because of the strong chemisorption of O₂ molecules on the Co dopant in MoS₂ and the effective charge transfer, the O–O bond (d_O–O) of the adsorbed O₂ molecule was elongated from that of free O₂ molecules, 1.233 Å to 1.357 Å. The elongated O–O bond length was very close to the typical value of ^•O₂⁻ radicals [9], which implies that ROS can be produced when O₂ is adsorbed on the surface of Co-MoS₂.

The piezocatalyzed molecular oxygen activation over Co-MoS₂-0.10 was evaluated by quantitative active species generation analysis. The steady-state concentration of ROS ([ROS]_ss), which represents the radical generation efficiency (ROS), was computed indirectly by monitoring the degradation of a radical probe. In this way, scavengers such as nitrotetrazolium blue chloride (NBT) and terephthalic acid (TA) were used as probes to assess [•O₂⁻]_ss and [•OH]_ss. As illustrated in Figure S3, the observed rate constants for Black T dye, TA, and NBT were 0.180, 0.055, and 0.034 min⁻¹, respectively. NBT is a chemical scavenger that can be used to scavenge •O₂⁻ radicals, and its second-order-rate constant is 5.88 × 10⁴ M⁻¹s⁻¹. Based on Equations (S7) and (S8), described in the Supplementary Materials, [•O₂⁻]_ss was calculated to be 34.41 × 10⁻¹⁰ M in the piezoelectric catalytic process. Terephthalic acid (TA) reactivates toward •OH with a second-order-rate constant of (3.3 × 10⁹ M⁻¹ s⁻¹). Furthermore, [•OH]_ss was calculated to be 15.38 × 10⁻¹² M (the detailed calculation is described in Text S5 of the Supplementary Materials). It was observed that although ^•O₂⁻ radicals were the dominant active species, ^•OH radicals took the second most important role in tetracycline degradation.

A periodic force is applied to Co-doped MoS₂ under mechanical vibration, leading to the bending of Co-doped MoS₂ nanosheets. A shaking-vibration-assisted piezoelectric field in odd- and few-numbered layers of Co-MoS₂ ultrathin nanosheets can be created, and across its width, it has negative piezopotential on the compressive strain and positive piezopotential on the tensile strain region, driving the generated electrons and holes to the opposite sites, as illustrated in Figure 4b. The piezoelectrically induced internal electric field can tilt the band structure and enable a reduction of O₂ with electrons [10]. Benefiting from this feature, a direct one-electron-reduction of O₂ would be the main path for ROS generation in the current experimental work. On the other hand, the CB potential of Co-doped MoS₂ is more negative than the reduction potential of O₂/H₂O₂ (0.695 V vs. NHE) via the two-electron-reaction pathway [11]. Importantly, the redox couples, Co²⁺/Co³⁺ and Mo⁴⁺/Mo⁶⁺, generated by Co-doped MoS₂, played a distinct role in improving the production of ROS via the one-electron-reaction and the two-electron-reaction pathways. The standard oxidation potential of Co²⁺/Co³⁺ is within a range of −1.40 to −1.53 V [12], which is higher than the one-electron reduction potential of O₂ (e⁻ + O₂ = ^•O₂⁻, −0.33 V vs. NHE) [13]. The oxygen is reduced by e⁻ to generate ^•O₂⁻, as described in Equations (2) and (3).

$$Co-Mo{S}_2 \stackrel{mechanical vibration}{\to } e^{-}\left(positive surface\right)+h^{+}\left(negative surface\right)$$

(1)

$${Co}^{2+}+O_2\to {Co}^{3+}+O_2^{•-}$$

(2)

$${Co}^{3+}+e^{-}\to {Co}^{2+}$$

(3)

On the other hand, the Mo atom on the surface of the Co-doped MoS₂ serves as a Mo⁴⁺/Mo⁶⁺ redox couple with a standard oxidation potential of (0.43–0.50 V) [14] that could promote the generation of H₂O₂ through the two-electron-reduction of O₂ (Equations (4) and (5)) [15].

$${2H}^{+}+O_2+{Mo}^{4+}\to {Mo}^{6+}+H_2{O}_2$$

(4)

$${Mo}^{6+}+{2e}^{-}\to {Mo}^{4+}$$

(5)

$$H_{2}O+h^{+}\to {HO}^{•}+H^{+}$$

(6)

$$H_2{O}_2+e^{-}\to {HO}^{•}+O{H}^{-}$$

(7)

The generated electrons from the VB of the Co-doped MoS₂ are promoted to Co²⁺- accommodated basal and edge sites of 2H MoS₂. Then, the electrons further transfer from the Co atom to the Mo atom because the electronegativity of Mo (2.16) is larger than that of Co (1.88) [16]. However, due to the strong interaction between Co–O bonds, it is feasible to transfer the electron from the Co atom to the adsorbed O₂ molecule, resulting in enhanced ROS generation, and the adsorbed O₂ acts as an electron acceptor. In other words, the redox couple of Co²⁺/Co³⁺ serves as an electron reserve and aids in the enhanced ^•O₂⁻, as described in Equations (2) and (3). According to theoretical calculations, the experimental active radical detection results showed that piezocatalyzed molecular oxygen activation via the one-electron-reaction pathway was dominant in the present study. Thus, one of the primary ROS, ^•OH, can also be generated through the oxidation of water (Equation (6)) and the electron reduction of H₂O₂ (Equation (7)). Ultimately, in the present case, tetracycline was successfully degraded within 15 min into nontoxic decomposed molecules such as CO₂ and H₂O under a piezoelectric catalytic process.

4. Conclusions

In summary, the piezoelectric catalytic properties of Co-doped MoS₂ ultrathin nanosheets successfully decomposed tetracycline by 99.8% in 15 min under shaking vibrations. Moreover, to verify the origin of the enhanced piezoelectric catalytic activity of MoS₂ via the doping effect, molecular oxygen activation properties were predicted using DFT calculation and monitored by generated ROS evolution. In addition, the mechanism of ROS generation that is responsible for the degradation of pollutant tetracycline was also investigated in detail.