In recent years, industrial activity has generated large amounts of wastewater, which often contains high concentrations of organic pollutants and inorganic heavy metal ions. Among the heavy metal ions, one of the most common contaminants is Cr(Ⅵ), which has been classified as a carcinogen and mutagen because of its high toxicity and harmfulness to most organisms, including humans [1-3]. Although Cr(Ⅵ) can be removed from aquatic environments by adsorption [4], it still poses a risk through secondary pollution. Cr(Ⅲ), however, is low-toxic and even beneficial to living organisms, and is also easily precipitated or adsorbed on inorganic/organic substrates at neutral or alkaline pH [5, 6]. Therefore, it is environmentally beneficial to deoxidize Cr(Ⅵ) to Cr(Ⅲ). Another class of aquatic contaminants is organic dye molecules, which usually contain benzene rings, making them hard to degrade by conventional chemical and biological methods [7]. Discharging of dyes into aquatic environments, even in small amounts, can pose severe risks to human health [8].

Photocatalysis is a potential technology for carrying out Cr(Ⅵ) reduction [9], which is more effective and lower in cost compared with other methods like electro-reduction [10-12] and chemical reduction [13-15]. Meanwhile, photocatalysis can also be used to decompose organic pollutants into biodegradable or less toxic compounds, or even into inorganic CO₂, H₂O, NO₃^- and halide ions [7, 16]. TiO₂ has been reported as an effective photocatalyst for these reactions because of its high activity, low toxicity, low cost and good stability and durability [17-19], although it suffers from low photocurrent quantum yield, difficult post-separation and the tendency toward agglomeration [20, 21]. CdS, SnS₂ and WO₃ have been used as photocatalysts under visible light irradiation [22-30], but their effectiveness was limited by their poor efficiency and slow rate [31]. Consequently, it has become urgent to find new efficient photocatalysts for the reduction of Cr(Ⅵ) and degradation of organic pollutants.

Coordination polymers (CPs) have gained much attention for their fascinating architectures and potential applications in photocatalysis [7, 32], conventional catalysis [33, 34], separation [35, 36], CO₂ capture [37, 38], gas storage [39, 40], pollutant adsorption [41, 42] and other processes [43-45]. Because of the catalytically active metal sites and the synergistic effects of the metals with functional organic ligands, along with the large surface area and permanent pores, CPs show great potential opportunities for heterogeneous photocatalysis, including H₂ production, Cr(Ⅵ) reduction and organic-pollutant degradation [46-50]. Zhao et al. [51] reported the reduction of Cr(Ⅵ) with high efficiency by NNU-36 ([Zn₂(BPEA)(BPDC)₂]·2DMF). Shen et al. [52] demonstrated that the Pd@UiO-66(NH ₂) nanocomposite exhibited excellent photocatalytic reduction of Cr(Ⅵ), and even showed enhanced activity in a binary system of Cr(Ⅵ) and organic dyes. Wang et al. [53] developed ZnO@ZIF-8 core-shell heterostructures, which were utilized to perform selective photoreduction of Cr(Ⅵ) from mixed Cr(Ⅵ)/MB solution (MB = methylene blue). Besides these 3D materials, 2D layered materials and their composite photocatalysts also have many unique properties, such as tunable band gaps, which can enhance their photocatalytic performance [54-56]. In summary, these previous reports revealed the potential application of photocatalysts based on CPs (including metal-organic frameworks, MOFs) to reduce Cr(Ⅵ) to Cr(Ⅲ) and degrade organic pollutants in wastewater.

Herein, the primary ligand 1, 3-dibenzyl-2-imidazolidone- 4, 5-dicarboxylic acid (H₂L) and the secondary ligand rigid 4, 4′-bipyridine (bpy) were selected to construct a new CP, Zn(bpy)L (named BUC-21). To the best of our knowledge, this is the first use of H₂L as a linker to build a CP. The current work demonstrates that BUC-21 exhibits an excellent reusability and much higher Cr(Ⅵ) reduction efficiency and reduction rate than those of P25 under the same conditions. Meanwhile, BUC-21 shows high photocatalytic activities for degradation of organic dyes, including cationic methylene blue (MB) and rhodamine B (RhB) and anionic methyl orange (MO) and reactive red X-3B (X-3B). In a matrix system containing both Cr(Ⅵ) and an organic dye, however, both the Cr(Ⅵ) reduction rate and the organic-pollutant degradation rate are decreased. A possible mechanism for this effect is proposed.

All commercially available chemicals were reagent grade, and used as received without further purification. Fourier transform infrared (FTIR) spectra were recorded in the region ranging from 4000 to 400 cm^-1 on a Nicolet 6700 infrared spectrophotometer with KBr pellets. Thermogravimetric analyses (TGA) were performed from 80 to 800 ℃ in an air stream at a heating rate of 10 ℃ min^-1 on a DTU-3c thermal analyzer using α-Al₂O₃ as reference. Ultraviolet-visible (UV-vis) diffuse reflectance spectra of solid samples were measured by a PerkinElmer Lambda 650S spectrophotometer, in which BaSO₄ was used as the standard with 100% reflectance.

A mixture of ZnCl₂ (0.3 mmol, 40.89 mg), bpy (0.3 mmol, 46.85 mg) and H₂L (0.3 mmol, 106.31 mg) was sealed in a 25 mL Teflon-lined stainless steel Parr bomb containing deionized H₂O (18 mL), heated at 160 ℃ for 72 h, and cooled down slowly to room temperature. White block-like crystals of C₂₉H₂₄N₄O₅Zn (BUC-21, yield 82% based on ZnCl₂) were isolated and washed with deionized water. Anal. Calcd. for BUC-21, C₂₉H₂₄N₄O₅Zn: C, 60.6%; N, 9.8%; H, 4.2%. Found: C, 60.8%; N, 9.9%; H, 4.3%. FTIR (KBr)/cm^-1: 3575, 3434, 3057, 2920, 2861, 2463, 1961, 1645, 1536, 1493, 1417, 1281, 1220, 1141, 1070, 1047, 1033, 1015, 965, 949, 915, 860, 813, 762, 745, 726, 703, 673, 643, 619, 570, 519, 484.

X-ray single-crystal data collection for BUC-21 was carried out with a Bruker CCD area detector diffractometer with graphite-monochromatized Mo K_α radiation (λ = 0.71073 Å ) using the φ-ω mode at 298(2) K. The SMART software [57] was used for data collection and the SAINT software [58] for data extraction. Empirical absorption correction was performed with the SADABS program [59]. The structure was solved by direct methods (SHELXS-97) [60] and refined by full-matrix-least squares techniques on F² with anisotropic thermal parameters for all of the non-hydrogen atoms (SHELEL-97) [60]. All hydrogen atoms were located by Fourier difference synthesis and geometrical analysis. These hydrogen atoms were allowed to ride on their respective parent atoms. All structural calculations were carried out using the SHELX-97 program package [60]. Crystallographic data and structural refinement for BUC-21 are summarized in Table S1. Selected bond lengths and angles are listed in Table S2.

The working electrode was prepared on a fluorine-doped tin oxide (FTO) glass, which was cleaned by sonication in ethanol for 15 min, and then dried under air. The FTO slide was dip-coated with 10 μL of slurry, which was obtained by mixing 5.0 mg BUC-21 powder (with particle sizes less than 0.074 mm) and 0.3 mL ethanol under sonication for 2 h. During dip-coating, the side part of the FTO slide was protected using Scotch tape. A copper wire was connected to the side part of the FTO glass using conductive tape. The uncoated parts of the electrode were isolated with an epoxy resin, and the exposed area of the electrode was about 0.25 cm². The electrochemical measurements were performed in a conventional three-electrode cell, using a Pt plate as counter electrode and a saturated Ag/AgCl electrode as reference electrode. Before measurement, the working electrodes were immersed in a Na₂SO₄ aqueous solution (0.2 mol L^-1) for 30 s. The photocurrents were measured with a Metrohm PGSTAT204 workstation. A UV lamp with optical power efficiency of 142 mW cm^-2 (Beijing Perfectlight Co. Ltd.) was used as a light source.

The photocatalytic experiments for Cr(Ⅵ) reduction from K₂Cr₂O₇ aqueous solution were conducted at ambient conditions in a quartz reactor containing 15.0 mg photocatalyst and 200 mL of 10 mg L^-1 Cr(Ⅵ) solution. After stirring for 30 min to reach adsorption-desorption equilibrium, the suspensions were irradiated by a 500-W Hg lamp (Beijing Aulight Co., Ltd). During the photocatalytic degradation experiments, 1.5 mL aliquots were extracted at 3 min intervals for analysis. The Cr(Ⅵ) content in the supernatant was determined colorimetrically at 540 nm using the diphenylcarbazide (DPC) method. Additionally, two cationic organic dyes, MB (10 mg L^-1) and RhB (10 mg L^-1), and two anionic organic dyes, MO (10 mg L^-1) and X-3B (50 mg L^-1), were selected as organic pollutant models to evaluate the photocatalytic performances of BUC-21 under the same conditions. A Laspec Alpha-1860 spectrometer was used to monitor the concentration changes determined at the maximum absorbance of 664, 554, 463 and 540 nm for MB, RhB, MO and X-3B, respectively.

The crystal structure analysis revealed that the asymmetric unit of Zn(bpy)L is built up of one Zn(Ⅱ) atom, two bpy ligands and two monodentate bridging carboxylate groups from two different L^2- groups (Fig. 1(a)). The Zn-N and Zn-O distances are in the range of 2.049-2.050 Å and 1.928(4)-1.959(5) Å , respectively, which are comparable to previously reported values [61]. The Zn(Ⅱ) atoms, adopting tetrahedral geometries as illustrated in Fig. 1(b), are linked into 1D chains by completely deprotonated L^2- groups along the a-axis, which are further joined into a 2D framework by rigid bpy ligands, as shown in Fig. 1(c) and (d).

BUC-21 was stable under air, and stable and insoluble in water and common organic solvents, including ethanol, methylbenzene, chloroform, ether, DMSO and DMF. The FTIR spectrum of BUC-21 was recorded, and as depicted in Fig. S1, the band at 2861 cm^-1 can be assigned to v(-CH₂-) vibrations [62]. The strong and broad bands at 1645 and 1417 cm^-1 are ascribed to the asymmetric and symmetric vibrations, respectively, of carboxyl groups. The bands at 1220 and 1141 cm^-1 can be assigned to the v(C-N) vibrations of the phenyl rings. The thermal stability of BUC-21 was examined by TGA, which established that the framework of BUC-21 was stable up to 340 ℃ (Fig. S2). The optical properties of BUC-21 were investigated by UV-vis diffuse reflectance spectrophotometry (UV-vis DRS). As shown in Fig. 2, the sample absorbed light in the UV region. The band gap (E_g) value was estimated using the Kubelka-Munk function, F = (1 − R)²/2R [63], where R is the reflectance of an infinitely thick layer at a given wavelength. The F versus E plot is shown in Fig. 2 (inset), where a steep absorption edge is displayed and the E_g value of the sample can be assessed at 3.4 eV, implying that BUC-21 is a potential wide-gap photocatalyst [64, 65].

Previous studies have shown that the reduction rate of aqueous Cr(Ⅵ) over photocatalysts is greatly influenced by the pH values [2, 52]. Theoretically, the Cr₂O₇^2- anion is predominant in solution under acidic conditions. To obtain the optimum reaction conditions, controlled experiments were first carried out with the addition of different amounts of H₂SO₄ solution (1:1, v/v). As shown in Fig. 3, the pH greatly influenced the photocatalytic activity of BUC-21. As the pH was successively decreased from 4.0 to 2.0, the reduction rate of Cr(Ⅵ) markedly increased. At pH = 1, however, the photocatalytic activity was significantly lower than at pH = 2 or 3, possibly because of destruction of the framework of BUC-21 (Fig. S3). The efficiency of photocatalytic Cr(Ⅵ) reduction was even poorer under alkaline conditions (not shown). As mentioned above, the predominating species of Cr at pH 2-4 is Cr₂O₇^2- [52, 66]. Therefore, the proposed reactions may be described by Eqs. (1) and (2):

(1)

(2)

The Cr(Ⅵ) reduction efficiencies of H₂L, BUC-21 and commercial P25 under UV light irradiation are shown in Fig. 4. About 39% and 59% of the initial content of Cr(Ⅵ) was deoxidized after 30 min in the presence of P25 and H₂L as photocatalysts, respectively. In comparison, the reduction rate was significantly increased in the presence of BUC-21, which achieved a reduction efficiency of 96% within 30 min. The extent of Cr(Ⅵ) reduction was minimal in the absence of photocatalyst or light, implying that the removal of Cr(Ⅵ) could be classified as a photocatalytic process.

To further characterize the photocatalytic behavior of BUC-21, four organic dyes (MB, RhB, MO and X-3B) were selected to perform degradation experiments under UV light irradiation. For comparison, control experiments were conducted in two different systems, under identical conditions to the main experiments except for the absence of light (but presence of photocatalysts) in one system, and the absence of photocatalyst (but presence of UV irradiation) in the other. The adsorptive removal rates of the four selected organic dyes in the presence of BUC-21 in the dark were negligible, as shown in Fig. 5. In the absence of any photocatalyst, ca. 20% MB, 16% RhB, 15% MO and 53% X-3B were decomposed after 30 min UV irradiation, and with the H₂L ligand as photocatalyst, 61% X-3B was decomposed. In comparison, BUC-21 exhibited outstanding UV-photocatalytic activities, degrading 93% MB, 97% MO and 100% X-3B in 30 min. These photocatalytic activities of BUC-21 are much higher than those in previous reports [67-71]. Besides, it should be noted that X-3B was completely removed by BUC-21 within only 9 min, which was faster than the time required by P25 (24 min) under the same conditions. A comparison of the performance of BUC-21 with those of previous 2D catalysts is listed in Table 1.

Furthermore, the photocatalytic activity of BUC-21 for simultaneous Cr(Ⅵ) reduction and dye degradation in a Cr(Ⅵ)/dye mixed system was investigated at pH = 2.0. As displayed in Fig. 6, for the mono-component systems, the Cr(Ⅵ) reduction and X-3B degradation efficiencies were 96% and 100%, respectively, within only 15 min. In the Cr(Ⅵ)/dye mixed system, the efficiency of Cr(Ⅵ) reduction was still ca. 96%. However, the removal percentage of X-3B was significantly decreased to 68%. It is well known that both holes (h⁺) and electrons (e^-) are produced in a photocatalyst under light irradiation. Cr(Ⅵ) is reduced to Cr(Ⅲ) by e^-, while many dyes can be oxidized by h⁺ [52, 78]. On this basis, one would expect both the reduction of Cr(Ⅵ) and degradation of X-3B to be enhanced in the mixed system. Instead, the photocatalytic degradation of X-3B was decreased, and the reduction of Cr(Ⅵ) was not improved. To explain this observation, we propose that the X-3B was not oxidized by the holes but by ·O₂^- radical anions, which were generated by the reaction of e^- and O₂. In the mixed system, the e^- were consumed by Cr(Ⅵ), resulting in the decreased rate of X-3B degradation.

To prove the assumption that the X-3B was photocatalytically degraded by ·O₂^-, benzoquinone (BQ) was introduced as a scavenging agent to trap ·O₂^-. In a separate experiment, the potential role of ·OH radicals was also investigated by adding isopropyl alcohol (IPA). The concentrations of both BQ and IPA were 0.2 mmol L^-1. As shown in Fig. 7, the X-3B degradation rate was significantly restrained in the presence of BQ, which suppresses ·O₂^- activity. Although the degradation proceeded at a steady rate, the removal percentage was down to 51%. However, when IPA was added instead, as an ·OH scavenger, X-3B was completely degraded within 24 min, suggesting that ·OH was not the dominant active species involved in the degradation. It can be concluded that ·O₂^- was the main active species, and ·OH was secondary, while the holes played no significant role in this photocatalytic reaction.

It can be seen from the above experiments that BUC-21 exhibited high photocatalytic activities towards Cr(Ⅵ) and organic pollutants in acidic and neutral solutions. In a further set of experiments, a solution of X-3B (initial pH = 4.3) was selected to study the photocatalytic activity of BUC-21 in alkaline conditions. As the pH value was successively increased from 6 to 12, the photocatalytic efficiency of X-3B degradation was markedly improved (Fig. 8). This relationship between the degradation rate of X-3B and the pH of the solution may be related to either the protonation/deprotonation of the basic sites present in the dyes, or the formation and stability of active intermediates responsible for decomposition, both of which are pH-dependent [79]. An alternative explanation is the greater concentration of hydroxide ions (OH^-) in the solution at higher pH, which induced the generation of hydroxyl free radicals (·OH) [80]. Under acidic conditions, ·OH played only a secondary role in decomposing X-3B (Fig. 7), probably because of its low concentration. However, at higher pH values, the degradation of X-3B was effectively promoted by the greater concentration of ·OH. In conclusion, BUC-21 is stable (Fig. S3), and is especially effective towards organic pollutants across a wide range of pH from 2 to 12.

The recyclability and stability of BUC-21 were tested by performing repeated usage cycles under identical reaction conditions. As shown in Fig. 9(a), the reduction rate of Cr(Ⅵ) over BUC-21 stayed above 95%, and the removal of the four selected organic dyes showed only a small decrease, after three runs' photocatalytic reactions, indicating that BUC-21 was highly stable and reusable. Moreover, the PXRD pattern of the used photocatalyst matched well with the simulated pattern based on the single-crystal data. It can be seen from Fig. 9(b) that the crystallographic structure and morphology of the catalyst underwent no noticeable change during the experiments, implying the good reusability and high stability of BUC-21 for Cr(Ⅵ) reduction and dye decomposition.

The transient photocurrent responses of BUC-21 under intermittent UV light irradiation are illustrated in Fig. 10, in which it is clear that BUC-21 coated on FTO glass showed strong photocurrent responses, which rapidly decayed as soon as the light was switched off. These results indicate that BUC-21 can be photoexcited to generate electron-hole pairs, which can then be separated efficiently, implying the good potential of BUC-21 for practical use as a photocatalyst.

Most CPs and MOFs, with the exception of MOF-5 [81], belong to the class of molecular photocatalysts, whose mechanisms are usually explained in terms of frontier molecular orbital theory [7]. A possible mechanism for the photocatalytic reduction of Cr(Ⅵ) and degradation of organic dyes over BUC-21 is proposed [7, 82, 83]. As illustrated in Fig. 4(b), under light irradiation, an electron (e^-) is excited from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO), producing a hole (h⁺) in the former. The photo-induced electron in the LUMO is usually easily lost, and either takes part in the reduction of Cr(Ⅵ) to Cr(Ⅲ), or is transferred to dissolved O₂ to generate ·O₂^-, which can then oxidize a dye molecule. The HOMO, however, strongly demands one electron to return to its stable state. Therefore, one electron is captured from a water molecule, which is oxygenated into ·OH. The ·OH radicals can further decompose organic dyes.

In summary, the synthesis of a new coordination polymer, BUC-21, has been accomplished. BUC-21 exhibited excellent reusability, and much higher performance for photocatalytic reduction of Cr(Ⅵ) and degradation of dyes than P25 under UV light illumination. Moreover, the photocatalytic behavior of BUC-21 in a mixed system of Cr(Ⅵ) and dye was also studied. In the mixed system, the reduction of Cr(Ⅵ) was not promoted, while the degradation of X-3B was significantly reduced, relative to the mono-component systems. The lower degradation rate of X-3B resulted from the consumption of photo-excited electrons by Cr(Ⅵ), limiting the generation of ·O₂^-, which was the main active species needed to decompose X-3B.