There have been numerous lattice-matched MOFs with identical crystallographic parameters, while it is a challenge to integrate lattice-mismatched MOFs together by a conventional synthesis method, because the high interface energy caused by topology mismatch of different MOFs would generally lead to phases of MOF individuals [20, 21]. To address this challenge, we developed a coprecipitation strategy for the synthesis of lattice-mismatched dual-MOFs, consisting of homogeneously distributed MOF individuals. HKUST-1 and ZIF-8 are two well-known MOFs, which are built from paddle-wheel [Cu₂(COO)₄] secondary building units (SBUs) and 1,3,5-benzene-tricarboxylate (BTC) and tetrahedrally coordinated zinc cations connected by 2-methylimidazolate (MIM) linkers, respectively [22, 23]. It has been demonstrated that the formation enthalpies of the Cu-O bond in [Cu₂(COO)₄] SBUs and the Zn-N bond in Zn(MIM)₂ are about −186 and −145 kJ/mol, which are higher than those of Zn-O and Cu-N bonds (about −180 and −90 kJ/mol), respectively [24, 25]. Therefore, nucleations of HKUST-1 and ZIF-8 are thermodynamically favourable, indicating that one-pot stoichiometric reaction of Cu(NO₃)₂·3H₂O, Zn(NO₃)₂·6H₂O and H₃BTC in the presence of excess 2-methylimidazole would result in a dual-MOF composite, denoted as HKUST-1/ZIF-8.

The power X-ray diffraction (PXRD) pattern of HKUST-1/ZIF-8 clearly shows the character diffraction peaks ascribed to ZIF-8 and HKUST-1, including the overlapped peaks, while the diffraction intensity is variable depending on the contents of HKUST-1 and ZIF-8 (Figure 1(a), Figures S1 and S2). The individual HKUST-1 and ZIF-8 in the dual-MOF could be easily etched with acetic acid (HAc) or ammonium hydroxide (NH₃·H₂O) to result in hierarchical MOFs HKUST-1 and ZIF-8, based on their distinct acid and alkali stability, denoted as HKUST-1/ZIF-8-A and HKUST-1/ZIF-8-B, respectively. The PXRD peaks of HKUST-1/ZIF-8-A and HKUST-1/ZIF-8-B match well with those of HKUST-1 and ZIF-8, respectively, indicating successful production of dual-MOF HKUST-1/ZIF-8.

The FT-IR spectrum of HKUST-1/ZIF-8 shows the characteristic signals of HKUST-1 appearing at 1620, 1439, and 1363 cm^-1 that are assigned to the COO–Cu₂ stretching vibration, and the absorption band at 1110 cm^-1 is ascribed to the C–O stretching vibration (Figure 1(b) and Figure S5) [26]. The peaks for ZIF-8 appear at 1148 cm^-1 for the C–N stretching vibration and at 421 cm^-1 for the Zn–N stretching vibration [27]. These results further confirmed the combination of HKUST-1 and ZIF-8 in the dual-MOF. There is also a very weak peak at 470 cm^-1 ascribed to the Zn–O stretching vibration, which becomes evident when increasing the ratio of HKUST-1 to ZIF-8, indicating that HKUST-1 and ZIF-8 in HKUST-1/ZIF-8 are coordinatively interconnected in the dual-MOF (Figure S5). The FT-IR patterns of HKUST-1/ZIF-8-A and HKUST-1/ZIF-8-B match well with those of HKUST-1 and ZIF-8, respectively, demonstrating that the dual-MOF is majorly built from HKUST-1 and ZIF-8.

The textual characters of HKUST-1/ZIF-8 were studied by N₂ sorption experiments (Figures 1(c) and 1(d), Figures S6 and S7, Table S1). HKUST-1/ZIF-8 exhibits the typical type IV isotherms, resulting in the total pore volume of 1.057 cm³/g and Brunauer-Emmett-Teller (BET) surface area of 968.3 m²/g, which are lower than those of HKUST-1 (V_{\text{pore}}=1.068\text{c}{\text{m}}^3/\text{g}, S_{\text{BET}}=1517.4{\text{m}}^2/\text{g}) and ZIF-8 (V_{\text{pore}}=1.531\text{c}{\text{m}}^3/\text{g}, S_{\text{BET}}=1670.7{\text{m}}^2/\text{g}), respectively. The surface area shrinkage may originate from interwrapping of the two MOFs. Nonlocal density functional theory (NLDFT) analysis results showed that the dominated pore sizes are ~9 and ~11 Å in HKUST-1/ZIF-8, which are ascribed to the pores in HKUST-1 and ZIF-8, respectively. In contrast, the etched samples of HKUST-1/ZIF-8-A and HKUST-1/ZIF-8-B exhibit the type IV isotherms with major pore sizes of ~9 and ~11 Å, respectively. The above results further proved that the dual-MOF is composed of HKUST-1 and ZIF-8.

The morphology of HKUST-1/ZIF-8 was monitored by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) (Figure 2 and Figures S8–S15). The SEM image of HKUST-1/ZIF-8 presents stacked irregular small particles with diameters of 40-60 nm, while the etched samples of HKUST-1/ZIF-8-A and HKUST-1/ZIF-8-B show the almost identical particle shape and sizes. TEM images clearly show the hierarchical porous structures in both etched samples, which originated from intertemplating between HKUST-1 and ZIF-8 in the dual-MOF. Energy-dispersive X-ray (EDX) spectroscopy elemental mapping images show that the Cu, Zn, O, and N elements in HKUST-1/ZIF-8 are uniformly distributed, indicating that HKUST-1 and ZIF-8 components are homogeneously distributed, which could be, respectively, etched under different conditions. The ratio of HKUST-1 to ZIF-8 is about 0.56 in the sample of HKUST-1/ZIF-8, as revealed by inductively coupled plasma optical emission spectroscopy (ICP-OES) and ¹H NMR spectrum (Figure S16, Table S2).

The thermal behaviors of HKUST-1/ZIF-8 were monitored by TG-MS measurement under N₂ atmosphere (Figure 3(a) and Figures S17–S21). Upon raising the annealing temperature, a weight loss is observed in the temperature range of 200-450°C, which is similar to that of HKUST-1, indicating that the weight loss is derived from decarboxylation of the carboxylate linkers in HKUST-1 [28]. When the temperature was raised above 550°C, the TG curve sharply decayed, which is similar to the thermal behaviors of phase-pure ZIF-8 [29]. The thermal behaviors of the dual-MOF clearly demonstrate that the HKUST-1 component could be easily pyrolyzed while the ZIF-8 component remains intact under suitable pyrolytic conditions. We therefore pyrolyzed HKUST-1/ZIF-8 under N₂ atmosphere at different temperatures for 2 h to fabricate Cu@ZIF-8 composites. The FT-IR spectrum indicates that the carboxylate moieties in HKUST-1 were almost completely decarboxylated for a solid sample of HKUST-1/ZIF-8 pyrolyzed at 350°C (Figure 3(b) and Figures S22–S24). In contrast, the characteristic ring stretching of 2-methylimidazolate ligands at 1460 and 1380 cm^-1 became predominant, and the characteristic absorption peaks of C–N stretching vibration at 1148 cm^-1 and Zn–N stretching vibration at 421 cm^-1 remain intact, indicating that ZIF-8 is stable under the pyrolytic conditions. In the PXRD pattern, there appeared broad metallic Cu peaks (JCPDS, Card No. 04-0836, 43.3 and 50.4°) in the annealed sample of HKUST-1/ZIF-8, while the peaks ascribed to ZIF-8 retained, indicating successful decomposition of the HKUST-1 component, which was further confirmed by XPS spectra (Figure 3(c) and Figures S25–S28).

The textual characters of the 350°C treated sample of dual-MOF HKUST-1/ZIF-8 were analyzed by N₂ sorption measurements (Figure 3(d), Figures S29–S31, Table S1). Cu@ZIF-8 exhibits the typical type IV isotherms with a BET surface area of 1014.1 m²/g. The relatively high nitrogen uptakes at low pressures were contributed by the extensive micropores, while the type H4 hysteresis loops at higher pressures suggest that there are mesopores in the composite material, indicating that Cu@ZIF-8 has hierarchical porous structure, which could facilitate mass transport during catalysis. The pore size distribution analysis revealed that the micropores in ZIF-8 were well reserved while those in HKUST-1 were wiped out in the annealed product, further certifying the decomposition of the HKUST-1 skeleton and retaining of the ZIF-8 skeleton under the pyrolytic conditions.

TEM and SEM images show that there are small aggregated NPs with average diameters of 10\pm 1.6\text{nm} inside ZIF-8 (Figure 4 and Figures S32–S38). A high-resolution TEM (HRTEM) image shows that the lattice-fringe spacing of the NPs is 0.181 nm, ascribed to the metallic Cu (200) facet. Consistent with the conclusions from PXRD patterns, TEM images showed that the diameters of Cu NPs are dependent on the proportion of HKUST-1 and ZIF-8. Upon increasing the ZIF-8 content, the size of Cu NPs gradually decreases under the identical pyrolytic conditions, indicating that ZIF-8 in the dual-MOF plays important roles in tuning the sizes of Cu NPs by acting as physical barriers to prevent the aggregation of Cu NPs under annealing conditions (Table S3). EDX images of Cu@ZIF-8 shows that Cu, Zn, O, and N elements are uniformly distributed, indicating that the Cu NPs are well dispersed in hierarchical porous ZIF-8, which might serve as the homogeneous active sites in catalysis.

To investigate the roles of organic ligands in the formation of the HKUST-1/ZIF-8 dual-MOF and Cu@ZIF-8 composite, we studied the PXRD patterns of the pristine and annealed samples of HKUST-1/ZIF-8 with different proportions of organic ligands (Figures S39–S41). All pristine samples show the similar PXRD profile of HKUST-1/ZIF-8. Excess dosage of H₃BTC would prevent the formation of ZIF-8, owing to the increased acidity. When the feed ratio of 2-methylimidazole to H₃BTC was raised from 3 to 9, there appear the characteristic diffraction peaks ascribed to ZIF-8 and metallic Cu in the annealed samples. It is worth noting that the diffraction peaks of metallic Cu became broaden, upon increasing the content of 2-methylimidazole. By further increasing the dosage of 2-methylimidazole, no obvious diffraction peak could be assigned to the copper species, indicating the small sizes of in situ-formed Cu NPs. These experimental results revealed that the feed ratio of the two organic ligands plays crucial roles in the production of the dual-MOF and finally tuning the size of in situ-formed Cu NPs. These conclusions were further confirmed by FT-IR, TEM, ICP-OES, and ¹H NMR (Figures S44–S63, Tables S2 and S3).

The annual world output of styrene to produce synthetic rubbers and plastics is more than 30 million tons, while the trace phenylacetylene would poison the polymerization catalysts and subsequently decrease the product quality [30, 31]. Therefore, removal of trace phenylacetylene from styrene feedstock is of high industrial importance [32]. Because there is no efficient way that could separate phenylacetylene from styrene in the industry, semihydrogenation represents the most practically applicable pathway to eliminate phenylacetylene from styrene [33]. Among numerous developed semihydrogenation catalysts, noble metal-based catalysts inherently exhibited high catalytic activity but low selectivity due to easy overhydrogenation, while earth abundant nonprecious metal-based catalysts showed high selectivity but intrinsic low catalytic activity [34–37]. For example, metallic Cu demonstrates high selectivity in semihydrogenation reaction, attributed to its excellent capability of easy desorption of the alkene species [38–40]; however, the catalytic activity of Cu-based catalysts is very low, while overhydrogenation still occurred [41].

It is interesting that the annealed product Cu@ZIF-8 demonstrates excellent catalytic properties, which could quantitatively transform phenylacetylene into styrene (>99.9% conversion and >99.9% selectivity) under the catalytic conditions of 1.5 MPa H₂ and 130°C (Figure 5(a), Figures S64–S69, Table S4). It is worth noting that no overhydrogenation occurred when prolonging the reaction time, even phenylacetylene was fully hydrogenated, which surpassed most of the literature work (Table S5). In contrast, the pristine ZIF-8, HKUST-1, and HKUST-1/ZIF-8 are almost inactive under the identical conditions (Table S4). The phenylacetylene conversion (9.0%) is dramatically decreased for the composite material derived from core-shell nHKUST-1@ZIF-8, because most of the Cu NPs are aggregated in the inner core with low dispersion (Figure 5(b), Figures S70–S72, Table S4) [42]. The excellent catalytic performance of Cu@ZIF-8 should be attributed to the homogeneous distribution and easy access of Cu NPs inside ZIF-8 with excellent molecular sieving properties and synergistic effect, which could highly improve the efficiency and selectivity in the catalysis [43]. In addition, it has been demonstrated that the structural heterogeneity would lead to low catalytic selectivity by undergoing different reaction pathways [44]. Therefore, the nature of homogenized active sites in Cu@ZIF-8 plays a vital role in improving the selectivity, attributed to their identical spatial and electronic interaction with reactant molecules. We also synthesized a composite Cu/ZIF-8 catalyst for comparison by a conventional solution impregnation method. Because there are substantial numbers of agglomerated Cu NPs on the outside of ZIF-8 with a broad size distribution, the styrene selectivity is very low (45.7%) (Figure 5(b), Figures S73–S75, Table S4). The catalytic properties are very sensitive to the Cu content in the composite materials, because the size of encapsulated Cu NPs is dependent on the Cu content, while the catalytic selectivity is related to the size of Cu NPs (Table S6). High Cu content resulted in low substrate conversion, while low Cu content led to overhydrogenation of styrene, due to the formation of large- and small-sized Cu NPs, respectively, revealing the structure-dependent characteristic of this catalytic reaction (Figure S76, Table S4) [36, 45].

To comprehensively understand the unique catalytic properties of Cu@ZIF-8 in the semihydrogenation reaction, the ZIF-8 component in Cu@ZIF-8 was digested in acetic acid (HAc) aqueous solution (Figures S77 and S78). The collected solid (denoted as Cu@ZIF-8-A) was used to catalyze the semihydrogenation reaction. Cu@ZIF-8-A exhibits relatively low activity (12.2% substrate conversion) under the same conditions, which was also observed for the annealed sample of HKUST-1 (26.2% conversion), indicating that the ZIF-8 matrix plays important roles in improving and stabilizing the catalytic efficiency of the encapsulated Cu NPs by confining and synergistic effect (Table S4) [46]. The ZIF-8 skeleton could be easily decomposed to form ZnO at 550°C (Figures S79–S82), which also resulted in decreased substrate conversion (85.6%) and styrene selectivity (48.1%) under the identical conditions, demonstrating that the microenvironment and molecular sieving effect offered by porous ZIF-8 are beneficial for improving the catalytic activity and selectivity (Table S4).

The very high selectivity catalyzed by Cu@ZIF-8 should also be ascribed to the unique molecular sieving function, which could selectively accumulate phenylacetylene molecules in the presence of styrene (Table S7). Cu@ZIF-8 could selectively absorb a large amount of phenylacetylene (3.46 mmol/g) from a solution of styrene and phenylacetylene in toluene (volume ratio of 1 : 1 : 100) at room temperature, while the adsorbed styrene is negligible. In contrast, the pristine HKUST-1/ZIF-8 exhibits low selectivity between phenylacetylene (3.58 mmol/g) and styrene (3.26 mmol/g). These results revealed that Cu@ZIF-8 could selectively enrich phenylacetylene and release styrene to improve the selectivity by avoiding overhydrogenation reaction during catalysis [38–40]. When Cu@ZIF-8 was used to hydrogenate styrene under the same catalytic conditions, no converted styrene could be detected, thus further proving its substrate specificity for phenylacetylene (Table S8). Cu@ZIF-8 was further evaluated in hydrogenation of a mixture of phenylacetylene and styrene (molar ratios of 1 : 1, 1 : 10, and 1 : 100), aimed at assessing its practical application for eliminating phenylacetylene from styrene (Figure 5(c), Table S9). Phenylacetylene is completely removed in the presence of excess styrene without occurring overhydrogenation, indicating that the molecular sieving function plays very important roles in the semihydrogenation reaction.

To understand the kinetic behaviors of Cu@ZIF-8 in the semihydrogenation reaction, we studied its catalytic properties at different temperatures. The semihydrogenation reaction catalyzed by Cu@ZIF-8 could be initiated at 80°C, which resulted in 20.1% styrene yield for 10 h and >99.9% styrene yield for 72 h, indicating that Cu@ZIF-8 is a highly active catalyst that could initiate the catalytic reaction under very mild conditions (Figure S83). The kinetic behaviors of Cu@ZIF-8 were investigated in the temperature ranges of 100-130°C by reducing the reaction time to 3 h, which would keep low conversions of phenylacetylene for the calculation of rational initial reaction rates (r_0) (Figure 5(d)). The apparent activation energy (E_{\text{a}}) for Cu@ZIF-8, calculated on the basis of the Arrhenius plot, is 43.3 kJ/mol, which is very close to 30–55 kJ/mol for Pd/C and much lower than that for Cu/Al₂O₃ [41, 47]. The low E_{\text{a}} value indicates that the special microenvironment provided by the porous ZIF-8 support would highly improve the hydrogenation activity of Cu NPs, which was proven by the high E_{\text{a}} value of Cu@ZIF-8-A (79.6 kJ/mol) without the porous ZIF-8 support. The highly improved catalytic activity of Cu@ZIF-8 could also be partially ascribed to the H₂ trapping capability of the pores in ZIF-8, which serves as nanoreactors to speed up the hydrogenation process [48].

Cu@ZIF-8 also demonstrates high selectivity for semihydrogenation of a series of terminal alkynes (Figure 6). Ether, amine, pyridyl, halogen, and vinyl groups could be well tolerated, which resulted in the full conversion of the corresponding terminal alkynes with excellent selectivity of alkenes (>99.9%) (Figures S84–S103). It is worthy of note that Cu@ZIF-8 can also be extended to semihydrogenation of biorelated molecules, such as vitamin E, a class of biologically essential fat-soluble antioxidants. Vitamin E was traditionally synthesized from trimethylhydroquinone and isophytol, while isophytol was mainly derived from alkenols [49]. It is interesting that alkynol could be quantitatively semihydrogenated by Cu@ZIF-8 to yield the alkenol product (11) without overhydrogenation, showing great promise for practical applications.

X-ray photoelectron spectroscopy (XPS) was utilized to study the interfacial interaction between Cu NPs and ZIF-8 support (Figure S104). The Cu 2p_3/2 binding energy for Cu⁰ species in Cu@ZIF-8 positively shifts to 933.0 eV, compared with the characteristic Cu 2p_3/2 binding energy (932.6 eV) for Cu⁰ species without the ZIF-8 supporting matrix in pyrolyzed HKUST-1 at 350°C, while the binding energy of N 1s negatively shifts to 398.9 eV, compared with 399.4 eV for ZIF-8. The shift degree is also sensitive to the ratio of HKUST-1 to ZIF-8 in the dual-MOFs. These results indicate that there occurred electron transfer from encapsulated Cu NPs to the ZIF-8 supporting matrix, leading to positively charged Cu^δ+ species located on the surface of Cu NPs inside the negatively charged ZIF-8 supporting matrix. The unique interfacial interaction between Cu NPs and ZIF-8 support would weaken its combination with styrene molecules to improve the styrene selectivity in catalysis, which has been confirmed by the selective sorption experiment between alkyne and alkene molecules (Table S7) [41, 45, 50].

Based on the results obtained in this work and the literature, we proposed a possible catalytic pathway for semihydrogenation of terminal alkyne over Cu@ZIF-8 (Scheme 2) [37, 40, 51]. It has been known that terminal alkyne is firstly absorbed on the surface of the Cu active site to form an intermediate. Meanwhile, the Cu species in Cu@ZIF-8 is able to adsorb, activate, and dissociate molecular H₂, which would react with the C≡C bond in alkyne. As a result, alkyne is converted into alkene, which is subsequently desorbed and released to avoid overhydrogenation of the alkene product on the surface of Cu NPs, owing to the weak sorption capability of positively charged Cu^δ+ species located on the surface of Cu NPs to alkene moiety.

The composite catalyst Cu@ZIF-8 can be simply recovered by centrifugation and subsequently reused for catalyzing semihydrogenation of phenylacetylene with retained high catalytic efficiency and selectivity for six successive runs (Figure S105). The PXRD pattern and ICP-OES analysis results revealed that the catalyst is very stable (Figure S106, Table S3). The sharpened diffraction peaks of the recovered ZIF-8 after catalysis indicate that there occurred self-healing of the ZIF-8 skeleton under the catalytic conditions. The TEM image shows no obvious particle agglomeration in the recovered catalyst, demonstrating the important roles of the ZIF-8 matrix in stabilizing the Cu NPs by serving as the physical barriers (Figure S107). Compared with those reported catalysts, Cu@ZIF-8 exhibits high stability and unprecedented catalytic properties for quantitative transformation of alkynes into alkenes, which is able to simplify the postprocesses and save energy to meet the standard of practical applications (Table S5).

All reagents and solvents were commercially available, which were directly used without purification, unless otherwise indicated. Zinc(II) nitrate hexahydrate (Zn(NO₃)₂·6H₂O) (99%), copper(II) nitrate trihydrate (Cu(NO₃)₂·3H₂O) (99%), acetone (≥99.5%), isopropanol (≥99.7%), methanol (≥99.5%), ethanol (≥99.7%), triethylamine (TEA) (≥99.5%), toluene (≥99.5%), N,N-dimethylformamide (DMF) (≥99.5%), ammonium hydroxide (NH₃·H₂O) (25–28%), and acetic acid (HAc) (≥99.8%) were purchased from Sinopharm. 1,3,5-Benzenetricarboxylic acid (H₃BTC) (98%), styrene (99.5%), phenylacetylene (98%), and 2-methylimidazole (98%) were obtained from Energy Chemical.

Powder X-ray diffraction (PXRD) patterns were collected on a RIGAKU D/MAX 2550/PC for Cu-Kα radiation (\lambda =1.5406\text{Å}) in the 2\theta range of 5–55° with an exposure time of 1 s and step size of 0.02°. Thermogravimetric-mass spectrometry (TG-MS) experiments were performed on a TGA/DSC1 1100S (Mettler Toledo) coupled with a ThermoStar GSD320 mass spectrometry (Pfeiffer Vacuum) under 1 atm N₂ atmosphere. ¹H NMR spectra were collected on a 400 MHz Bruker spectrometer. Fourier transform infrared (FT-IR) spectra were recorded from KBr pellets on a Nicolet NEXUS 470 spectrometer in the range of 400–4000 cm^-1. X-ray Auger electron spectroscopy (XAES) and X-ray photoelectron spectroscopy (XPS) were measured on a Thermo ESCALAB 250Xi with Al-Kα irradiation (1486.6 eV), and the C1s peak at 284.8 eV was used to calibrate the binding energies. The morphology of the composite materials was studied by field emission scanning electron microscopy (FESEM) on a Hitachi SU-8010, transmission electron microscopy (TEM) on a Hitachi HT-7700, and high-resolution TEM (HRTEM) on a JEOL 2100F system equipped with energy-dispersive X-ray (EDX) spectroscopy. The N₂ adsorption/desorption isotherms were performed on a Micromeritics ASAP 2020 surface area analyzer. The solid samples were degassed in vacuum at 80°C for 6 h prior to analysis. Copper and zinc contents were determined by using a Varian 730-ES ICP-OES Spectrometer. The catalytic results for semihydrogenation of phenylacetylene were analyzed by GC (FULI, 9790 II) equipped with an FID detector and a PEG-20M capillary column using N₂ as carrier gas. GC-MS spectra were recorded on a SHIMADZU GCMS-QP 2010.

Cu(NO₃)₂·3H₂O (1.21 g, 5.0 mmol) and Zn(NO₃)₂·6H₂O (1.49 g, 5.0 mmol) were added into 10 mL methanol under stirring, resulting in a homogeneous blue solution. 2-Methylimidazole (1.64 g, 20.0 mmol) in 5 mL methanol was subsequently added into the above solution under stirring at room temperature, which was further reacted for 10 min to form a homogeneous dark blue solution. A solution of H₃BTC (700 mg, 3.33 mmol) and TEA (3.0 mL) in 10 mL methanol was added into the above solution at room temperature, which was further reacted for 3 h under stirring. The green solid of HKUST-1/ZIF-8 was collected by centrifuging and washing with methanol several times, which was subsequently activated at 80°C in an oven overnight.

In a typical experimental procedure, 200 mg activated HKUST-1/ZIF-8 and HAc (0.5 mL) were added into 5 mL methanol, which was stirred for 30 min at room temperature. The blue solid was collected by centrifuging, washing with methanol several times, and drying at 80°C overnight, denoted as HKUST-1/ZIF-8-A, where A represents the acetic acid.

200 mg activated HKUST-1/ZIF-8 and NH₃·H₂O (0.5 mL) were immersed in 5 mL methanol, which was stirred for 30 min at room temperature. The white solid was collected by centrifuging, washing with methanol several times, and drying at 80°C overnight, denoted as HKUST-1/ZIF-8-B, where B represents the ammonium hydroxide.

In a typical experimental procedure, 200 mg activated HKUST-1/ZIF-8 solid sample was put in a furnace, which was pyrolyzed at 350°C for 2 h under continuous nitrogen flow. The annealed product was denoted as Cu@ZIF-8.

200 mg Cu@ZIF-8 and HAc (0.5 mL) were added into 5 mL methanol, and the mixture was reacted for 30 min under stirring at room temperature. The solid was collected by centrifugation and washed with methanol several times, which was dispersed in 2 mL toluene for the catalytic test, denoted as Cu@ZIF-8-A, where A represents the acetic acid.

Semihydrogenation of alkynes was performed in a 50 mL stainless high-pressure autoclave. Alkyne (0.2 mmol), solvent (2 mL), and catalyst (10 mol% based on Cu) were loaded to the autoclave, which was sealed, flushed five times with H₂ at 1 MPa, and subsequently pressurized with H₂ to the desired pressure. The reaction mixture was heated to the targeted temperature, reacted at the temperature for a certain time, and cooled down to room temperature. The reaction mixture was centrifuged, and the liquid phase was subjected to GC-MS analysis by using DMF as an internal standard.