Amine-Functionalized ZnO Nanosheets for Efficient CO2 Capture and Photoreduction

Amine-functionalized ZnO nanosheets were prepared through a one-step hydrothermal method by using monoethanolamine, which has a hydroxyl group, for covalent attachment on ZnO and a primary amine group to supply the amine-functionalization. We demonstrate that the terminal amine groups on ZnO surfaces substantially increase the capability of CO2 capture via chemisorption, resulting in effective CO2 activation. As a result, the photogenerated electrons from excited ZnO can more readily reduce the surface-activated CO2, which thereby enhances the activity for photocatalytic CO2 reduction.


Introduction
Nowadays, the increasing level of CO2 concentration from the use of fossil fuels has become the most serious environmental concern related to global warming and climate change. Researchers have realized that photocatalytic conversion of CO2 into hydrocarbons by means of solar energy would be an ideal way to lower down the CO2 concentration in the atmosphere and generate sustainable chemical fuels in the meantime. A pioneering study on the reduction of CO2 was demonstrated by Inoue et al. in 1979 by using semiconductor photocatalysts [1]. Afterwards, many scientific studies have been devoted to the development of efficient photocatalysts for CO2 photoreduction [2][3][4][5][6][7][8][9][10][11][12][13][14]. Among them, ZnO-based OPEN ACCESS nanostructures with different sizes and morphologies have been extensively reported as photocatalysts since they are stable, non-toxic, and low cost [15][16][17][18][19][20]. However, the efficiencies for CO2 photoreduction of these ZnO-based photocatalysts were usually unsatisfactory [21,22], which might be ascribed to the low affinity between CO2 and ZnO surfaces [23].
As is well known, the amine functionalization can effectively assist immobilization of CO2 and has been extensively used for CO2 capture in industry [24,25]. Tremendous efforts have been made to introduce amine groups onto solid material surfaces to promote CO2 capture [26,27]. The chemical interactions between amine groups and CO2 lead to the formation of carbamate (or bicarbamate) that can transform into carbonate upon hydrolysis [28]. In this work, we report a straightforward method to prepare amine-functionalized ZnO nanosheets by using monoethanolamine (MEA) that possesses a hydroxyl (-OH) group for covalent attachment on ZnO and a primary amine (-NH2) group to endow an amine-functionalized surface. We demonstrate that the terminal amine groups on ZnO surfaces substantially improve the CO2 adsorption, which consequently results in significantly enhanced photocatalytic activity for CO2 reduction as compared to the clean ZnO without surface amine groups. Figure 1a shows the X-ray diffraction (XRD) patterns of the prepared MEA-ZnO sample and clean ZnO sample. These two samples exhibit almost the same diffraction lines which can be indexed to wurtzite ZnO. This indicates that the involvement of MEA into the reaction did not alter the crystal structure of ZnO. The absorption spectra of both samples also appear to have almost no difference, as shown in Figure 1b [29]. In comparison, the clean ZnO sample prepared without MEA appears as irregular particles, as shown in Figure 2b. The sheet-like growth of ZnO is probably due to attached MEA molecules that suppress the intrinsically anisotropic growth of ZnO along the (0001) direction. As such, the ZnO crystal growth would intend to proceed sideways towards a nanosheet [15].   The effect of amine functionalization by MEA is examined through the CO2 adsorption test. As shown in Figure 4, the MEA-ZnO exhibits much higher (approximately six times) CO2 uptake capability (0.87 cm 3 ·g −1 ) than the clean ZnO (0.13 cm 3 ·g −1 ), while in comparison, the difference in their surface area is quite minor (15.1 m 2 ·g −1 for MEA-ZnO vs. 9.6 m 2 ·g −1 for clean ZnO). This result proves that the amine-functionalized ZnO surface can greatly enhance CO2 adsorption. Moreover, the MEA-ZnO shows a dramatic rise in CO2 uptake along with elevated CO2 pressure, while the clean ZnO shows almost linearly increased CO2 adsorption. This observation suggests that specific interactions between CO2 and MEA-ZnO occur during the CO2 adsorption process. We further carried out photocatalytic CO2 reduction by irradiating the prepared sample (MEA-ZnO or clean ZnO) with UV-Vis light in the presence of CO2 and water vapor. As shown in Figure 5, the clean ZnO sample exhibits very low generation rates of methane (2.1 μmol/g) and CO (14.8 μmol/g). In comparison, the amine-functionalized sample (MEA-ZnO) showed a much higher production rate for methane (4.4 μmol/g) and CO (25.3 μmol/g). This comparison indicates that the amine functionalization by MEA can not only enhance the capture of CO2 molecules but also facilitates their photoreduction into CO and CH4. Control experiments were carried out in the absence of light or CO2 (in O2 atmosphere), which did not lead to generation of CO or CH4. This suggests that excitation of ZnO is required to obtain the products (CO and CH4) which come from CO2 photoreduction rather than the oxidation of MEA by the photogenerated holes of ZnO.

Results and Discussion
To measure the amount of O2 formed during the photocatalytic reaction, we performed a 12 h photocatalytic reaction for the MEA-ZnO catalyst under the reaction conditions described above. The detected amount of H2, O2, CO, and CH4 were 81 ppm, 1145 ppm, 1520 ppm, and 146 ppm, respectively. According to the stoichiometry of the involved reactions (1) to (4) and the amount of gaseous products (H2/CO/CH4), the O2 evolution amount should be around 1092 ppm, which is slightly lower than the actual detected O2 amount (1145 ppm). This might be due to the possible generation of liquid hydrocarbon products (e.g., CH3OH, HCOOH) from the reactions (5) to (7), which could not be effectively analyzed by the TCD (thermal conductivity detector) of GC (Gas Chromatography  We have also evaluated the long-term stability of MEA-ZnO for CO2 photoreduction through a cycling test. After each cycle of the four-hour test, the reactor was degassed by CO2 and the sample was re-used for the next test. As shown in Figure 6, no obvious decrease of the yield of CH4 and CO was observed after three cycles, suggesting that the MEA-ZnO sample is very stable during the process of CO2 photoreduction. After the cycling test, the MEA-ZnO sample was examined again by FT-IR spectroscopy, and no change was observed in the IR spectrum, which further confirmed the good stability of MEA-ZnO during the reaction. Figure 6. The cycling test of CO2 photoreduction over 12 h by using the MEA-ZnO sample.
Based on the above results, we conclude that the amine functionalization of ZnO can effectively facilitate the capture of CO2 as well as its further photoconversion into CH4 and CO. The principle can be illustrated in Figure 7. The terminal amine groups enable the creation of "C-N" bonding with CO2 by forming carbamate, which acts like an activation process of CO2 due to higher reactivity of carbamate than that of linear CO2 [31][32][33][34]. Further, these active carbamates are close to the ZnO surface and tend to establish direct interactions with Zn 2+ , which allows for receiving electrons from excited ZnO to implement reduction reactions towards CO and CH4 production. By this method, the overall CO2 photoreduction process can be promoted.

Preparation of Amine-Functionalized ZnO (MEA-ZnO)
The amine-functionalized ZnO sample was prepared by a hydrothermal method. In a typical procedure, a mixture aqueous solution (20 mL) containing 0.1 M Zn(Ac)2, 0.2 M NaOH, and 0.005 M MEA was put in a Teflon-lined autoclave and heated at 90 °C for 12 h. Then the solid product was collected and annealed at 250 °C under Ar atmosphere for 3 h to remove physisorbed MEA on ZnO surfaces. For comparison, the clean ZnO sample was also prepared as the control sample under the same condition without using MEA.

Photoreduction of CO2
The photocatalytic CO2 reduction was performed in a 100 mL gastight reactor with a quartz window and two side-sampling ports. In a typical process, the MEA-ZnO sample (20 mg) was suspended in an ethanol solution and the suspension was dispersed on a 6.25 cm 2 glass slide upon drying. The glass slide containing MEA-ZnO was then put into the gastight reactor. Prior to the irradiation, the reactor was purged with high purity CO2 for 30 min to remove the residual air and then 0.1 mL ultrapure water was injected into the reaction system. A xenon lamp (MAX-302, Asahi Spectra Co. Ltd., Tokyo, Japan) was used as the light source for the photocatalytic reaction. During the reaction, the gas product was analyzed periodically through a gas chromatograph (GC-7890A, Agilent Technologies Inc., Santa Clara, CA, USA) with a TCD detector.

Conclusions
In summary, we have successfully prepared amine-functionalized ZnO nanosheets through a hydrothermal process. The amine functionalization on ZnO surfaces substantially increases the capability of CO2 capture via chemisorption with effective CO2 activation. As such, amine-functionalized ZnO nanosheets exhibit significantly enhanced photocatalytic activity for CO2 reduction into CH4 and CO as compared to the clean ZnO.