X-ray diffraction (XRD)
Figure 1 shows the X-ray diffraction patterns of commercial Mg(OH)2, Mg5 and Mg5H supports. Mg5 exhibits reflection peaks at 36.90°, 42.82°, and 62.27° (2θ) resulting from to the (111), (200) and (220) planes of the cubic phase of MgO (according to JCPDS 89-7746). In the case of commercial Mg(OH)2, reflection peaks at 18.48°, 37.95°, 50.86°, 58.57°, and 68.38° (2θ), are observed, corresponding to the (001), (101), (102), (110) and (103) planes of the hexagonal phase of Mg(OH)2 (according to JCPDS 07-0239) (Paul, et al. 2012). Aditionally, planes of cubic MgO are also exhibited, it consists of a mixture of two phases with small traces of MgO. The re-hydroxylated MgO (Mg5H, Fig. 1) exhibited reflection peaks corresponding to the (001), (101), (102), (110) and (103) planes of the hexagonal phase of Mg(OH)2 with small peaks of the cubic-MgO phase, indicating that the MgO is almost hydroxylated to Mg(OH)2 during the ammonia treatment.
Figure 2A shows the CS/Mg5 and CS/Mg5H X-ray patterns. In the case of CS/Mg5, Mg(OH)2 reflection peaks are observed, denoting a re-hydroxylation of MgO after the one pot synthesis. The phase composition for CS/Mg5 is around 11.0% of MgO and 89.0% of MgOH)2. An additional peak can be observed in the low-angle region between 5–15° (2θ), suggesting a laminar structure in the formed re-hydroxylated Mg5 support (Paul, et al. 2012). This indicates that during CdS precipitation, the hydroxylation of MgO toward Mg(OH)2 is carried out and the formation of the layer structure of Mg(OH)2 occurs at the same time, mainly due to the alkaline conditions generated by the use of ethylenediamine and also the aqueous conditions. In addition, the low intensity of the broadened peaks was attributed to the small crystallite size (5.3 nm) and also to the intercalation process.
For CS/Mg5H, no peaks corresponding to the layer structure are observed. The absence of an intercalated structure suggests that the formation of Mg(OH)2 is not affected by the presence of ethylenediamine and that the hydroxylation process is stabilized during the ammonia treatment. Although a small MgO peak was also detected (Table 1), the high intensity of corresponding Mg(OH)2 peaks advises that this support has a high crystallinity. The phase composition presented in this supported material is mostly of Mg(OH)2 (86.4%).
Table 1
Data of crystallite size, band gap energy and specific surface area for CdS(en) supported materials.
Sample | Crystallite size (nm) | Eg (eV)) | Surface area (m2/g) | CdS wt. % |
Mg(OH)2 | MgO |
(001) | (101) | (200) | (220) |
CdS(en) | ---- | ---- | ---- | ---- | 2.60 | ---- | ---- |
Mg5 | ---- | ---- | 09.6 | 07.9 | 3.53 | ---- | ---- |
Mg5H | 10.7 | 8.2 | 15.9 | 8.4 | 4.56 | 45.81 | ---- |
CdS/Mg5 | 5.3 | ≤ 4 | ---- | 9.6 | 2.78 | 77.56 | 3.6 |
CdS/Mg5H | 7.0 | 8.2 | ---- | 7.5 | 2.65 | 26.51 | 4.2 |
In both cases, supported materials [from 20 to 40º (2θ), Fig. 2B] display diffraction peaks at 25.10°, 26.66°, and 28.48° corresponding to the (100), (002) and (101) planes of hexagonal phase of CdS (according to JCPDS 41-1049), however the crystal sizes cannot be determined since these reflection peaks are very broad and small.
Thermogravimetric analysis
TGA and TA measurements were used to analyze the thermal stability of materials in static air (25°C–800°C. Figure 3 shows the TGA curve of commercial Mg(OH)2.
The TGA curve shows two steps of weight losses. The first one appears below 250°C and is related to the loss of free physisorbed water (4.5wt.%), the heat flow indicates that this process is endothermic; the second and major weight loss is displayed in an interval of 250–410°C. The strong endothermic peak observed around 356°C in the TA curve, is related to the decomposition of Mg(OH)2 and crystallization of MgO particles (Wang, et al. 2007). However, the observed weight loss of ∼26.7wt.% at 600°C, is slightly lower than the theoretical value for the transformation of Mg(OH)2 to MgO (30.8wt.%) (Kumari, et al. 2009). This is probably the result of an incomplete dehydroxylation of Mg(OH)2 in a short time during this temperature interval and to the presence of small traces of MgO as it was suggested previously by XRD.
Figure 4 displays TGA and TA thermograms of the supported CdS materials. In the case of the CS/Mg5, the TGA curve show three weight loss steps. The first one appears below 200°C with a weight loss of 10–15 wt.%, because of the loss of physisorbed water, the TA indicates that this process is endothermic. However, the high weight loss and the prominent endothermic peak indicate that large amounts of physisorbed species intercalated in the interlayer structures of formed Mg(OH)2 are eliminated. The second and higher weight loss observed in the process occurs in an interval temperature of 200–450°C and matches with the strong endothermic peak observed around 363°C in the TA. These processes are related to the dehydroxylation of Mg(OH)2 and the subsequent crystallization of the MgO particles. The observed weight loss for this process is around 25.6 wt.% for CS/Mg5, which is also slightly lower than the theoretical value of the transformation of Mg(OH)2 to MgO (30.8wt.%). This result confirms that this annealed MgO sample used as support is hydroxylated during the CdS(en) precipitation caused by using an ethylenediamine aqueous solution. A third weight loss appears in the interval from 450°C to 660°C and is around ∼1.2% of the total weight and probably corresponds to residual compounds. Finally, a slight weight loss (1%) corresponding to the transformation of CdS to CdO, is observed around 664 and 684°C (Xiao, et al. 2007). For CS/Mg5H material similar results were obtained. In both cases, two weight losses are observed from 220 to 350ºC and 410 to 450 ºC accompanying by two exothermic signals detected around 280°C and 420°C, this can be related with the loss of ethylenediamine anchored to the CdS surface that is carried out in two steps (2%), a similar behavior was described before for related compounds (Feng, et al. 2009).
FTIR spectra
The FTIR spectrum of CS/Mg5 (Fig. 5.) exhibits small bands at 3624 cm− 1 associated to the stretching vibrations of the hydroxyl (O-H) group, and also at 1500 cm− 1 characteristic of the Mg(OH)2 support (Zhu, et al. 2011), however the low transmittance of these bands can be attributed to the low crystallinity of Mg(OH)2 (see Fig. 1.). In contrast, for the CS/Mg5H sample, the presence of sharp bands associated to hydroxyl (O-H) group can be attributed to a better crystallization of the Mg(OH)2 support.
In addition, a band at 3337 cm− 1 is observed corresponding to the stretching vibrations of N-H moiety, probably due to the formation of hydrogen interactions between the N atom of the CdS(en) and the environmental moisture. The bands at 1577 and 1318 cm− 1 are associated to the stretching vibrations of -NH and C-N bonds respectively, confirming the coordination of the ethylenediamine to the CdS surface (Hernández-Gordillo, et al. 2015). These results corroborate the precipitation of the CdS(en) hybrid material on the formed Mg(OH)2 support. For CS/Mg5, the bands expected for ethylenediamine molecule are not as evident as in the case of CS/Mg5H, this is probably for the high amount of CdS(en) formed on Mg5H surface, in contrast to the CdS(en) resultant on Mg5 support. Considering that the Mg5 support is hydroxylated in the CdS formation (Fig. 2), during the arrangement of the Mg(OH)2 lamellar structure, a part of CdS(en) could be formed inside of the layers, decreasing the amount of CdS(en) on Mg5 surface. The band displayed at 873 cm− 1 assigned to the Mg-O bond was not detected (Meshkani, et al. 2009).
Scanning Electron Microscope (SEM) Analysis
Results obtained by SEM for the supported CdS samples are shown in Fig. 6A and 6B. The morphology observed in the materials by SEM micrographs agrees with the results found in the X-ray section (3.1). It is clearly observed that after the one pot synthesis, a lamellar morphology of Mg(OH)2-MgO support is obtained. For supported CS/Mg5H materials, large particles of Mg(OH)2 of lamellar morphology were observed and stabilized after ammonia treatment. In both cases, CdS particles could not be detected because of the resulting small particle sizes.
Surface area
The calculated specific surface for the CS/Mg5 and CS/Mg5H materials is particularly different (78 m2/g and 27 m2/g) and the lower surface area for the last sample can be attributed to the high crystallinity of the Mg(OH)2 support (Zhang, et al. 2017).
Elemental analysis
Elemental analysis by EDS shows the amounts of CdS supported on the resultant materials CS/Mg5 and CS/Mg5H (Fig. 7). For CS/Mg5, the observed amount of supported sulfide was very low regarding to the expected amount (3.6%). Considering the MgO hydroxylation process during the CdS precipitation, it is important to think about that this oxide gains around 42 wt. % [physisorbed H2O (15%), hydroxylation (26%) and impurities (1%)], and as consequence the amount of CdS(en) seems to decrease, resulting in 2.9 wt.% of CdS(en). In the case of CS/Mg5H, the experimental quantity is closer to the estimated amount of CdS(en) (4.2%), due to the Mg(OH)2 support did not undergo any important change. This experiment also shows the excellent dispersion of CdS(en) particles on support, obtained by the synthetic method employed (Fig. 8).
UV–Vis spectroscopy
UV-vis diffuse reflectance spectra of the supported CdS materials are shown in Fig. 9(A-B). Both supported materials exhibit absorption in the UV region between 190 and 230 nm (Low UV) corresponding to the Mg(OH)2 electronic transitions (not shown) (Kumaria, et al. 2009), however these transitions cannot be clearly observed (in these supported materials) because they are overlapped with the absorption edge of CdS(en) dispersed particles. This absorption edge is observed in a wide interval of the UV-vis light region close to 360–470 nm, mainly at 450 nm, attributed to the intrinsic band-gap transition of electrons from the valence band to the conduction band of the CdS semiconductor, matching with the emission spectrum (line) of the blue LED lamps used for the photocatalytic test. CS/Mg5 displays low absorption ability in the blue region (Fig. 9a), while CS/Mg5H material can absorb until 480 nm, mainly at 450 nm.
The optical bandgap energies of these materials were calculated using the basic relationship between reflectance and incident photon energy (E = hν) given by: (FR×hv)2vs (hv) and the plot of (FR×E)2 vs (E) is shown Fig. 9(b). So, the band gap energy was determined by extrapolating the straight-line portion to the abscissa at zero absorption co-efficient. The bandgap energies of the resultant materials are given in Table 1. The high band gap energies of the obtained semiconductors indicate that the quantum size confinement effect caused by the small crystallite size of CdS(en) is very high in these materials.
Photocatalytic H2 production
Blank tests for the H2 production without photocatalyst (photolysis) and using either Mg5 or Mg5H supports without CdS(en) were completely negligible (not shown) because any H2 production wasn´t observed. Evaluation of CS/Mg5, CS/Mg5H and CdS(en) materials as photocatalysts in the hydrogen evolution reaction (HER) was carried out using an aqueous solution of 50 Vol.% of methanol, under blue light irradiation during 7 h. Figure 10 displays the profiles of H2 production for the photocatalytic reaction of synthesized supported materials, and it is possible to determine that CS/Mg5H compound presented a higher photocatalytic activity. This material exhibited the highest performance at 5 h of reaction, however, the reaction rate decreases after this time. The rate of H2 production (Fig. 10B) was standardized considering the specific amount of CdS (2–4 wt.%) deposited on the Mg(OH)2 support. CS/Mg5H is 4 times more active than the CdS(en) unsupported and 3 times more active than the CS/Mg5, despite that CS/Mg5H material displays a lower superficial area, as is illustrated in Table 1. This efficient photoactivity would be attributed to the next reasons: a) the high dispersion of CdS(en) small particles on the support compared to CdS bulk, and b) in the case of CS/Mg5H, the Mg(OH)2 support crystallizes better in contrast to CS/Mg5 photocatalyst, this is possible to support in the broad and short peaks detected by the XRD patterns for this material (Fig. 2.). This better crystallization of the support allows for a better well-organized dispersion of the CdS(en) nanoparticles on the Mg(OH)2 network, leading to a more reactive structure for this photocatalyst. It is important to take into account that CdS/Mg5H material displays a higher absorption of light in the wavelength (450 nm) at which photocatalytic runs were carried out. The photocatalytic stability of the CdS/Mg5H material in the HER is shown in the Fig. 10C. In this case, high rates of H2 production are reached in the first cycle, but it was decreased in the second and the third cycle. After the second cycle, the solution color was turned orange and the color of the supported material was turned from yellow to dark. This suggests that CdS was possibly lixiviated to the solution by the phototocorrotion process.
As a mechanistic approach, we can conclude that the high activity of the CdS(en) supported samples is influenced not only by the quantum confinement effect of the CdS(en) semiconductor, but also by the morphology and crystalline structure of the supports. Considering that the Mg(OH)2/MgO mixture support cannot be activated by blue light irradiation, the H2 evolution is only photogenerated on the CdS(en) nanoparticles. The role of Mg(OH)2/MgO mixture as support seems like to facilitate the separation of photoelectron/holes at the interface and thus enhances the phocatalytic ability of CdS(en) to abstract a proton from the medium. In this case, CdS conduction band position is favorably negative and is able to photogenerate e− that can be quickly transferred to the protons (H+) to produce H2. While methanol is oxidized by the photogenerated h+ to formaldehyde or formic acid (Schneider, et al. 2013). The surface coating provides more stability to hybrid CdS(en) in water for the photocatalytic process, and at the same time, has to conserve a good accessibility to the active surface sites, where the charge carriers need to be transferred for the reaction (Ben-Shahar, et al. 2015). In this case, features conferred to hybrid CdS(en) are able to satisfy these characteristics, due to the electropositivity of the resulting Cd active sites, because of the cordination of ethylenediamine since the polarization existing between the atoms of N and Cd, in contrast to the non-hybrid CdS semiconductor, and as a result of that, CdS(en) hybrid surface increases the number of active sites for the photocatalytic reaction (Wei, et al, 2013; Hernández-Gordillo, et al. 2015; Ramírez-Rave, et al. 2015; Ramírez-Rave, et al. 2020). The proposed mechanism for H2 production on CdS(en) supported photocatalysts is shown in Fig. 11. The use of methanol improved the H2 production in this reaction, acting as holes scavenger and avoiding both the electron/hole recombination. Additionally, the higher activity for the CdS/Mg5H under blue light irradiation can be attributed to the interaction of CdS(en) with crystalline Mg(OH)2/MgO mixture, where CdS(en) hybrid Organic-Inorganic material is very well dispersed.