Effect of various alkaline agents on the size and morphology of nano-sized HKUST-1 for CO₂ adsorption

Abstract

Two type of modulators, including sodium acetate and triethylamine were used to synthesize nanosized HKUST-1 crystals with tailored size and morphology using a coordination modulation method. The results of X-ray diffraction, scanning electron microscopy, transmission electron microscopy, and N₂ sorption isotherms confirm that the size of the HKUST-1 crystals can be tuned from the micron scale to the nanoscale through precise modulation of the amounts of the modulators. Moreover, hierarchical porous structures are obtained with the addition of modulators. It is suggested that a moderate acid–base environment and capping groups play an important role in controlling the size and morphology of HKUST-1. Dynamic CO₂ breakthrough experiments illustrate that the CO₂ capacity of nanosized HKUST-1 crystals is superior to that of micron-sized HKUST-1 crystals under low CO₂ partial pressure conditions (0.1 bar). Besides, the CO₂ adsorption capacity of nanosized HKUST-1 crystals was maintained after ten adsorption/desorption cycles.

---

1. Introduction

Fast industrial development and human activities are considered to be the major reasons leading to the sharp increase in atmospheric carbon dioxide concentration. As a result, controlling carbon dioxide emissions that leads to global warming is a tough issue facing researchers.^1–7 Flue gases are the major sources of CO₂ emission which account for almost 47% of the total carbon emission globally.⁸ It is a priority to find inexpensive, effective and stable materials and technologies to capture CO₂ from flue gases. Physisorption between CO₂ molecules and certain adsorbents is assumed to be a convenient and reversible process to capture CO₂. Besides, the process also leads to much less parasitic energy compared to the traditional aqueous amine absorbents and amine-functionalized solids based on chemisorption.⁹ Recently, porous materials such as zeolites,¹⁰ activated carbon,¹¹ mesoporous materials,^12,13 carbon molecular sieves,¹⁴ etc. that act as adsorbents to capture CO₂ have been extensively studied. However, the materials still suffered from several problems, such as low CO₂ adsorption capacity or difficult for regeneration.

Metal–organic frameworks (MOFs) are a new class of functional hybrid materials that attract much attention owing to their potential application in gas adsorption and separation, catalysis and drug delivery, etc.^7,15–18 The structures of MOFs are composed of metal nodes (single metal ions or clusters) bridged of organic ligands to form porous 3-D reticular materials through supermolecule self-assembly. Moreover, these compounds possess extraordinary high surface areas, low density, and well-defined pore structures.^19–21 Therefore, the developed 3-D pore structures of MOFs can be used for gas molecules such as CO₂ adsorption.

HKUST-1 is a promising candidate in CO₂ capture owing to its open metal sites, high surface area and narrow pore size distribution.²² Mavrandonakis et al. reported the interaction between CO₂ molecules and HKUST-1 by means of density functional theory.²³ Staudt et al. studied the adsorption of H₂, N₂, CO₂, CH₄ on the HKUST-1 under high pressure and the CO₂ adsorption capacity reached 42.8 wt% at 300 bar.²⁴ Despite many studies related to the CO₂ capture under high pressure have been done, few reports were found about the CO₂ adsorption over HKUST-1 at a lower CO₂ partial pressure (for post-combustion CO₂ capture).²⁵ Besides, previous works were related to the synthesis, characterization, and application of bulk HKUST-1. The synthesis of nanosized MOFs and applications in CO₂ capture were ignored. Moreover, it had been reported that nanosized MOFs can greatly accelerate the CO₂ adsorption kinetics as compared to their micro-sized counterparts e.g. nano-scaled Al(OH)(NDC) showed higher H₂ and CO₂ uptake capacities as well as excellent selectivity for CO₂ over N₂ and O₂.²⁶ Therefore, it is necessary to evaluate the CO₂ adsorption capacities of nanosized HKUST-1 from flue gases.

The methods for the synthesis of nano-sized MOFs, including water-in-oil microemulsions method,²⁷ ultrasonic,²⁸ microwave-assisted solvothermal method,^29,30 direct mixing method,³¹ room temperature synthesis^32–35 and solvothermal synthesis with the addition of polymer and surfactant³⁶ have been developed. However, the methods still suffered from several problems such as the control of the size and the morphology as well as the aggregation of the nanocrystals.³¹ The coordination modulation using modulators with or without the same functionality of organic ligands is the most commonly used method, which enables the impeding of coordination interactions between metal ions and organic ligands to fabricate nanosized MOFs.^37–40 Kitagawa et al. used dodecanoic acid to influence the coordination equilibria to further control the crystal growth of nanosized crystals.⁴¹ On the basis of Kitagawa's work, Zhang et al. combined coordination modulation with acid–base adjustment to precisely control the size and morphology of MOFs.³⁷

In this work, nanosized HKUST-1 was synthesized by controlling coordination modulation using different modulators. The influence of modulator type and concentration on the textural property and size of HKUST-1 were discussed. The CO₂ adsorption properties from flue gas (10 vol% CO₂, 90 vol% N₂) of nanosized and microsized HKUST-1 were measured over a fixed-bed reactor. The relationship between the CO₂ adsorption performance and the structural properties of HKUST-1 was also investigated. In addition, the regeneration and the stability of nanosized HKUST-1 were also evaluated for ten consecutive cycles.

2. Experimental section

2.1 Synthesis of HKUST-1 compound

2.1.1 Chemicals. Copper nitrate trihydrate (Cu(NO₃)₂·3H₂O, 99%), copper acetate monohydrate (Cu(OAc)₂·H₂O 99%), 1,3,5-benzenetricarboxylic acid (H₃BTC, 98%), N,N-dimethylformamide (DMF, 99.5%), triethylamine (TEA, 99%), sodium acetate (NaAc, 99%), ethanol (EtOH, 99.7%) were purchased from Aladdin Industrial Corporation.

2.1.2 Synthesis of bulk HKUST-1 crystal without any additive. 0.50 g of H₃BTC was dissolved in 12 mL of 1 : 1 : 1 mixture of DMF/EtOH/H₂O, and 1.04 g of Cu(NO₃)₂ dissolved in the same mixture. Then the two mixtures were mixed and sealed in a 50 mL of vial, then heated at 80 °C for 24 h. The mixture was defined as the standard solution. After crystallization, the obtained blue powders were filtered and washed two times with 25 mL of DMF, and then dried at 200 °C under vacuum overnight. The yield is about 96%.

2.1.3 Synthesis of nanosized HKUST-1 crystal with TEA and NaAc. 1–4 equiv. of TEA and 1–8 equiv. of NaAc were added to the standard solution, and the blue precipitations were instantly generated. Subsequently, the mixture was heated at 80 °C for 24 h. The washing and drying processes were the same as before. The obtained samples synthesized with TEA and NaAc were denoted as HKUST-1-Tn and HKUST-1-Nn, respectively. “n” was represented the equiv. number of modulators. For all of HKUST-1-Tn and HKUST-1-Nn samples, high yield of ∼92% was obtained.

2.1.4 Synthesis of HKUST-1 crystal with Cu(OAc)₂. 0.50 g of H₃BTC was dissolved in 12 mL of 1 : 1 : 1 mixture of DMF/EtOH/H₂O, and 0.86 g of Cu(OAc)₂ dissolved in the same mixture. The two mixtures were mixed to form blue suspension in one minute. Then 0–4 equiv. TEA was added to the mixture. Subsequently, the mixture was heated at 80 °C for 24 h. The washing and drying steps were similar to the previous steps. The obtained samples were denoted as HKUST-1-Cn and weighted (∼90%).

2.2 CO₂ dynamic breakthrough experiment

CO₂ adsorption and desorption processes were conducted in a fixed-bed reactor. Moderate adsorbents were added in a U type quartz tube with an inner diameter of 8 mm and plugged by quartz wool in the side of gas outlet. The HKUST-1 powders were pressed into a disk under 5 MPa for 3 min, and then the disk was broken to particles and the particles were sieved for 20–40 mesh particles. The sorbents were activated at 200 °C for 6 h in argon atmosphere with a flow of 60 mL min⁻¹, then cooled to adsorption temperature under argon atmosphere. The simulated flue gas (CO₂ 10 vol%, N₂ 90 vol%) were introduced into the reactor. The CO₂ concentration of the outlet gas of the quartz tube was determined by a gas analyzer (Vaisala, Finland) for every 10 s. The CO₂ adsorption capacity of HKUST-1 can be determined from the breakthrough curves as reported in previous report.⁴²

2.3 Characterization

Powder X-ray diffraction (PXRD) patterns were collected with a D8 Advanced diffractometer operated at 40 kV and 40 mA with monochromated Cu Kα radiation (λ = 1.5406 Å) and with a scan speed of 3° min⁻¹. The simulated PXRD patterns were calculated from modeled crystal data using the Diamond 3.2i software suite. Nitrogen sorption isotherms were measured at 77 K on a Micromeritic ASAP 2020 system. The samples were outgassed at 473 K overnight before the measurement. Scanning electron microscopy (SEM) images were obtained on a JEOL at 10 kV. Before measurement, the samples suspended in the ethanol were dropped onto a tin-foil plate and dried at room temperature. Transmission electron microscopy (TEM) investigations were performed with a JEM-2100F (operated at 10 kV). Before the measurement, the material was deposited onto a holey carbon foil supported on a copper grid. The Fourier transform infrared spectrum (FT-IR) was collected on a Nicolet Magna-II550 using the KBr technique. Thermal gravimetric analysis (TGA) was performed on a Rigaku TG thermal gravimetric analyzer in the temperature range of 30–700 °C under nitrogen atmosphere at a heating rate of 10 °C min⁻¹. The basic strength distributions of samples were examined by CO₂ temperature-programmed desorption (CO₂-TPD). The experiments were conducted on BALZA Q-Mass spectrometer. 150 mg (20–40 mesh) of HKUST-1 sample was put in a quartz microreactor and pretreated by argon flow of 50 mL min⁻¹ at 473 K for 1 h. After it was cooled to 30 °C, CO₂ was introduced to the sample for 1 h, then the sample was flushed with argon at 30 °C for 1.5 h to remove the physisorbed CO₂. After the purge of argon, the sample was heated at a heating rate of 10 °C min⁻¹ under argon flow of 60 mL min⁻¹, CO₂ desorption curve was recorded by mass spectrometer.

3. Results and discussion

3.1 Influence of modulators on the structure, size and morphology of HKUST-1

In the Cu(NO₃)₂ system, the effects of NaAc and TEA concentration on the size and the morphology of HKUST-1 were investigated in detail. The standard solution without additives kept clear for 6 h at room temperature (r.t.). While with the addition of additives, the solution became cloudy simultaneously at r.t. which indicated that the basicity of NaAc and TEA can accelerate the deprotonation of H₃BTC and promote the nucleation of particles.³⁷ The structure of the samples were characterized by XRD and the results were shown in Fig. 1. As illustrated in Fig. 1(a), the samples prepared with 1–6 equiv. of NaAc were very similar and were in good agreement with the simulated pattern, indicating that the structure of HKUST-1 was well preserved. With the amount of NaAc increased up to 7 equiv., all the peaks ascribed to HKUST-1 still existed, but new diffraction peaks located at 2θ of 7.06, 8.26, 9.94 and 12.08 were observed. Further increasing the NaAc amount to 8 equiv., new diffraction peaks appeared which indicated the formation of new phases. Similarly, the samples showed the same patterns with the bulk HKUST-1 with the addition of 1–3 equiv. of TEA, indicating that the structures of HKUST-1 were well kept (Fig. 1(b)). Upon introduction of 4 equiv. of TEA, new diffraction peaks at 2θ of 9.58 and 13.50 were observed, suggesting the formation of new phases. Hence, a moderate amount of NaAc (≤6 equiv.) or TEA (≤3 equiv.) did not have an effect on the structure of HKUST-1.

Fig. 1 XRD patterns of HKUST-1 synthesized with different amounts of NaAc and TEA (given as equivalents with respect to H₃BTC).

Five new peaks appeared in the HKUST-1-N7 pattern which still existed in the HKUST-1-N8, suggesting that it was not a pure phase. Through retrieving ICDD PDF database, it was convinced that the new peaks were not attributed to NaAc. Hence it rules out the possibility of residual NaAc in the sample. At the same time, three new peaks appeared in the HKUST-1-T4. In the synthesis process of HKUST-1, the mother liquid was blue because the amount of Cu(NO₃)₂ was excess. However, when 4 equiv. of TEA was added to the solution, the mother liquid was colorless after reaction. This phenomena also implied that the residual Cu(NO₃)₂ was not exist in the form of copper ions. They may form some coordination complex with excess amounts of TEA. In order to confirm the new phase, FTIR analysis was carried out (shown in Fig. S1†). It can be found that the IR absorption bands in the 1700–1500 and 1500–1300 cm⁻¹ ranges are attributed to ν_asym(C–O₂) and ν_sym(C–O₂) stretching modes, respectively. The IR absorption bands were existed in the HKUST-1-N8, which suggested that the structure of HKUST-1 was retained. However, new IR absorption bands appeared at 1051, 1023, 825, 719, 685, 625, 525 cm⁻¹. 1051 and 1023 cm⁻¹ absorption bands were assigned to the residual NaAc. According to the previous interpretation, the IR absorption bands at around 500 cm⁻¹ were assigned to Cu–O stretching modes.⁴³ Hence, it may be caused by coordination interaction between the excess NaAc and Cu(II) ions of [Cu₂C₄O₈] framework cage. Furthermore, some new IR adsorption bands was observed at 1613, 1540, 1177, 1157, 1033, 1010, 872, 836, 589, and 546 cm⁻¹. It can be deduced that new coordination complex is formed by the interaction of TEA and Cu(II).

Compared with the pattern of bulk HKUST-1, peak broadening of HKUST-1-Nn and HKUST-1-Tn were attributed to the reduced particle size of HKUST-1. The sizes of all the HKUST-1 samples were calculated using the Scherrer equation, and the results were listed in Table S1.† In the process of calculating particle size, the deconvolution parameter is set to 1.5. The particle size was calculated according to the Scherrer equation as follows:

where D is the particle size (nm); K is Scherrer constant, here K = 1; B is FWHM (rad); θ is the diffraction angle (rad). The value of FWHM was calculated according to the most intense (222) reflection on each spectrum.⁴⁴ It can be found that NaAc can effectively decrease the particle size of HKUST-1 to ∼40 nm, and the particle size of HKUST-1-Nn was smaller than that of HKUST-1-Tn, suggesting that NaAc was much effective than TEA in reducing the particle size of HKUST-1. The good agreement between the patterns of nanosized HKUST-1 and bulk HKUST-1 revealed that the nanosized HKUST-1 kept high crystallinity. Furthermore, (111) diffraction peak appeared in the XRD pattern, while it disappeared along with the addition of NaAc and TEA. It is difficult to define the (111) diffraction peak in the simulated pattern of HKUST-1 (Fig. S2†), whose intensity depends on the oriented direction of crystal growth. For the bulk HKUST-1, small amounts of HKUST-1 crystals are oriented along the [111] direction which lead to the appearance of obvious (111) diffraction peak. The result is in accordance with the previous report.⁴⁵ At the same time, nanosized HKUST-1 crystals lose the anisotropy with the addition of TEA or NaAc. Therefore, the (111) diffraction peak disappeared.

Fig. 2 and 3 clearly showed the size and the morphology change of HKUST-1 particles. Regular octahedral HKUST-1 particles with an average diameter of 15 μm were formed without any additives (Fig. 3(a)). However, the particles decreased from 15 μm to 40 nm with the introduction of NaAc (Fig. 2(c)). Interestingly, the particle size of HKUST-1-Nn samples presented a first decrease and then increased with the addition of NaAc. This tendency was also reflected from the data of Table S1† e.g. with addition of 1 equiv. of NaAc, the average size of HKUST-1-N1 decreased to ∼80 nm. Further increasing the amount of NaAc to 3 equiv., the size of HKUST-1 particles decreased to ∼40 nm. However, the aggregation was serious when the addition of NaAc exceeded the stoichiometric ratio (>3 equiv.), which was due to the high surface energy of nanoparticles (Fig. 2(d)).³³ As a result, it was not easy to measure the particle size distribution from SEM images. From the embed image of Fig. 2(d), it can be seen that a hierarchically porous materials were formed by interconnection of particles with diameter of 50 nm, leading to the formation of inter-particle voids. As shown in Fig. 2(e) and (f), the particle size of HKUST-1 increased slightly from 50 to 70 nm with the addition of 5–6 equiv. of NaAc, which might result from the competitive situation caused by excess capping agents for the complexation of copper cations to decrease the oversaturation of solvent.⁴¹ In addition, the particle size of HKUST-1-N7–N8 further increased. The appearance of rod particles (Fig. 2(g) and (h)) also demonstrated the formation of the new phase, which was in agreement with XRD results. Likewise, a moderate amount of TEA can also decrease HKUST-1 particles size from 15 μm to ∼65 nm in agreement with the result of XRD (Fig. 3). Excess TEA (TEA/H₃BTC = 4 equiv.) resulted in the formation of plate-like particles. With addition of 1 equiv. of TEA, the size of bulk HKUST-1 decreased significantly from 15 μm to 175 nm. Further adding the TEA, the particle size decreased continually e.g. the particle size of HKUST-1-T3 reached the minimum value (∼65 nm). However, the particle size distribution obtained from SEM images was larger than the results calculated from PXRD which indicated the obtained structure was constructed from aggregates of small crystallites.

Fig. 2 SEM images of HKUST-1-Nn samples produced with different amounts of NaAc: (a) 1, (b) 2, (c) 3, (d) 4, (e) 5, (f) 6, (g) 7, (h) 8 equiv.

Fig. 3 SEM images of HKUST-1-Tn samples produced with different amounts of TEA. (a) 0, (b) 1, (c) 1.5, (d) 2, (e) 3, (f) 4 equiv.

The initial pH value of precursors of HKUST-1 were measured after different amounts of additives were added to the precursor for 5 min (Table S2†). It can be found that the addition of NaAc and TEA can promote the basicity of system. Before crystallization, the solution had a pH value of 2.35, which was difficult for the deprotonation of H₃BTC and the formation of HKUST-1 crystals. With the addition of NaAc and TEA, higher pH value was obtained which lead to more deprotonated BTC ions for coordination with copper ions, resulting in the increase of nucleation rate and the formation of nanosized HKUST-1 crystals. The size of HKUST-1-T3 (Fig. 3(e)) was slightly larger than that of HKUST-1-N3 (Fig. 2(c)) which meant that NaAc seemed to be slightly favorable than TEA to decrease the particle size of HKUST-1. This might because that NaAc not only provided a moderate basic environment, but also contained the functional group for facile synthesis of HKUST-1 with copper source.⁴¹ That is to say, NaAc can be used as a capping group to compete with the organic ligand for the formation of spherical nanosized HKUST-1 particles with sizes in the range of 40–70 nm (particle size distribution in Fig. 2).

Hierarchical porosity of HKUST-1-N5 was confirmed by N₂ adsorption isotherms. As shown in Fig. 4, a combination of the type I microporous sorption at very relative pressure and the type IV mesoporous sorption with type H4 hysteresis loop was observed, indicating the presence of micropores and mesopores.⁴⁶ The pore size distribution of HKUST-1-N5 showed a narrow peak centered at 10.7 nm and a large amounts of macropores larger than 50 nm. It also can be seen from TEM images (Fig. 5(c)), many inter-particle voids were formed. The above results suggested that the addition of NaAc could reduce the particle sized and create larger grain voids. What' more, the high magnitude TEM image of HKUST-1-N4 and N5 showed that the 40–50 nm sized HKUST-1 crystals were aggregated with 2–3 nm crystals (Fig. 5(b) and (d)). It was due to that these smaller crystals were unstable as isolated one, so they form steady and robust aggregates. Accordingly, the size distribution of nanosized HKUST-1 from the SEM images was measured according to the size of steady aggregates (Fig. 2 and 3).

Fig. 4 N₂ adsorption isotherm and pore size distribution of HKUST-1-N5.

Fig. 5 TEM images of HKUST-1 samples obtained with different amounts of modulators. (a) and (b) HKUST-1-N4; (c) and (d) HKUST-1-N5; (e) HKUST-1-T1.5; (f) HKUST-1-T3.

It can be found that the total pore volume and the BET surface area of bulk HKUST-1 are 0.74 cm³ g⁻¹ and 1748 m² g⁻¹, respectively, within the range of values reported in the literature.³³ The type I adsorption isotherm suggests the exist of micropores. With the addition of NaAc and TEA, the BET surface area and total pore volume of the samples decreased, whereas the CO₂ adsorption capacities of nanosized HKUST-1 increased a lot. In order to analyze the relationship between textural property of nanosized HKUST-1 samples and CO₂ adsorption capacity, all the HKUST-Nn samples were separated into two groups. Through comparing the textural properties of all the HKUST-1-Nn samples, it can be found that the CO₂ adsorption capacity decrease with the decreasing of the BET surface area and the micropore volume percentage for N1–N3. Interestingly, hierarchical structure was formed in N4–N6 as confirmed by N₂ adsorption isotherm (Fig. 4(a)). The pore size distribution of HKUST-1-N5 showed a narrow peak centered at 10.7 nm and a large amounts of macropores larger than 50 nm (Fig. 4(b)). Furthermore, the total pore volume of HKUST-1-N5 reached 0.73 cm³ g⁻¹, almost the same with that of bulk HKUST-1, which was caused by the interconnected nanosized HKUST-1 network. The increment of CO₂ adsorption capacity in N4–N5 may be due to the formation of hierarchical porosity. In the case of HKUST-1-N6, the particle size increased again, leading to the decrease of hierarchical porosity and CO₂ adsorption capacity.

In the case of HKUST-1-Tn, BET analysis showed that the specific surface area of samples synthesized with TEA addition decreased obviously e.g. the BET surface area reduced from 1748 m² g⁻¹ for bulk HKUST-1 to 336 m² g⁻¹ for HKUST-1-T3 (Fig. S3†). The percentage of micropore for HKUST-1-Tn samples also decreased significantly as the amount of TEA increased, suggesting that the reduction in surface area was attributed to the decrease in micropore volume percentage. Meanwhile, the average pore size gradually increased with the addition of TEA in which the inter-particle voids produced by decreased particle size of HKUST-1 are larger (Fig. 5(e and f)).⁴¹ The above results suggested that only by adjusting the basicity was not effective to obtain the nanosized crystals with high specific surface area and pore volume. What's more, the relationship between the CO₂ adsorption capacity of HKUST-1-Tn and textural property was the same with HKUST-1-N1–N3, which is determined by the BET surface area. However, there was an exception, HKUST-1-T3 showed a 1.04 mmol g⁻¹ of CO₂ adsorption capacity with 336 m² g⁻¹ of BET surface area, which was larger than that of HKUST-1-T2. Based on this results, it can be deduced that the CO₂ adsorption capacity of HKUST-1 is not only determined by the textural properties of adsorbents. Hence, CO₂-TPD analysis of HKUST-1 samples were conducted to study the basicity of HKUST-1. The specific analysis of CO₂-TPD can be found in 3.4.

According to the above results, capping groups are still necessary to stabilize the nanosized HKUST-1 with high surface area and pore volume. In the Cu(NO₃)₂ system, NaAc is more favorable than TEA in the synthesis of uniform nanosized HKUST-1 with high specific surface area and micropore volume which might due to the competition effect of capping agent of NaAc with organic ligands leading to the formation of well dispersed nanosized particles.

3.2 Influence of copper resource on the structure and morphology of HKUST-1

When mixed Cu(OAc)₂ with H₃BTC solution at r.t., a blue precipitate of HKUST-1 was formed which suggested it was a fast reaction, because Cu(OAc)₂ can provide copper ions and the second building units (SBUs) for facile HKUST-1 synthesis via ligand exchange at the same time. On the other hand, the increased nucleation rate may be promoted by different relative basicity of the anions and the ability to deprotonate the H₃BTC.³³

It was noted that the structure of HKUST-1 was well retained until 4 equiv. of TEA was added according to XRD patterns of HKUST-1-Cn (Fig. 6), and the crystallinity increased with the increase of TEA. The diffraction peak of (111) began to strengthen, which suggested the large amounts of crystals enclosed by (111) crystallographic facets were formed.³⁸ As shown in Fig. 7, the addition of TEA resulted in the formation of well-separated nanoparticles with high crystallinity. However, the samples obtained from Cu(OAc)₂ without any additives showed ill-defined gel-like morphology as shown in Fig. 7(a). Moreover, according to the results of PXRD and SEM, high concentration of TEA lead to the increased mean size of particles. This may be due to the fact that TEA binding to Cu²⁺ in competition with the deprotonated ligands which was more remarkable in a fast reaction and led to decreasing of the oversaturation of the precursor.⁴⁷ In the present case, high concentration of additive could provide a slow nucleation and form larger regular particles during the heating process. In addition, the textural properties of HKUST-1 (equiv. ≥ 3) also showed obvious improvement. As shown in Table 2, the BET surface area also presented a first decrease and then increase tendency. In the case of high concentration of TEA (TEA/H₃BTC = 3 equiv.), HKUST-1-C3 had the largest surface area and microporous volume percentage as well as the smallest average pore size with regular well-separated particles. The increase of specific surface area in the HKUST-1-Cn samples may due to the increase of crystallinity. Furthermore, the aggregation of nanosized HKUST-1 crystals is not as serious as HKUST-1-C0 i.e. the HKUST-1-C3 particles are well dispersed and different from the ill-defined gel-like HKUST-1-C0. Based on the above reasons, a remarkable increase was found in the specific surface area of HKUST-1-C3. The above results also suggested that the coordination modulation was more suitable in size-control for a faster reaction. With lower TEA concentrations, a large amount of nuclei were formed and rapidly grow at the same time. While the available reagents were quickly depleted, affording smaller crystals with lower crystallinity. However, with the increase of modulator concentration, the nucleation rate became slow and large particles were formed. These results were in good agreement with the work of Kitagawa.⁴¹ On the base of above results, a mechanism for the nanosized HKUST-1 was proposed. As shown in Scheme 1, it can be found that the size-control effect of modulators in the Cu(NO₃)₂ and Cu(OAc)₂ system played an opposite effect in the synthesis of nano-sized HKUST-1 crystals. In the case of Cu(NO₃)₂ system, the basicity of TEA and NaAc can improve the pH of HKUST-1 precursor to adjust the deprotonation and nucleation rate. Moreover, the capping group of NaAc can compete with the organic ligands for the coordination of Cu²⁺. Compared to TEA, the effect of NaAc can obtain nanosized HKUST-1 with high specific surface area and large pore volume. Remarkably, nanosized HKUST-1 with sizes in the range of 40–85 nm (according to the results of SEM images) over a wide range of NaAc and H₃BTC ratios (1–6) without change of H₃BTC structure. While for Cu(OAc)₂ system, TEA played an opposite role in the fast reaction. Low concentration of modulator can largely promote deprotonation and nucleation rate to obtain nanosized HKUST-1 with a serious aggregation and lower specific surface area and small total pore volume. With the increase of modulator concentration, the reaction rate became slow and larger particles (>100 nm) were obtained (Fig. 7(c)). Smaller crystals are indeed growing in line with the persistent nucleation during the heating process, leading to larger crystals with greater size polydispersity (Fig. 7(d)).

Fig. 6 XRD patterns of HKUST-1-Cn samples synthesized with different amounts of TEA and Cu(OAc)₂.

Fig. 7 SEM images of HKUST-1-Cn samples synthesized with different amount of TEA and Cu(OAc)₂, (a) 0, (b) 1, (c) 2, (d) 3, (e) 4 equiv.

Scheme 1 Proposed formation mechanism of nanosized HKUST-1.

3.3 The stability of nanosized and bulk HKUST-1 crystals

Typical TGA/DTG curves of bulk and nanosized HKUST-1 samples were shown in Fig. S4.† The two samples have the similar TG curves from room temperature to 600 °C. The first weight loss peak between 30 °C to 280 °C was due to the desorption of physisorbed water and decomposition of organic molecules in the frameworks. The second weight loss peak between 300 and 500 °C might be attributed to the decomposition of the frameworks. It can be seen that the nanosized HKUST-1 crystals decomposed began at 300 °C and ending at 360 °C, while the bulk HKUST-1 crystals decomposed began at 270 °C and ending at 350 °C which showed that nanosized HKUST-1 was more stable than bulk HKUST-1 crystals.

3.4 Influence of particle size on CO₂ adsorption capacity

To have a better understanding of the CO₂ adsorption performance of the HKUST-1 synthesized with different modulators and copper resources in the lower CO₂ pressure. CO₂ adsorption capacities were calculated from breakthrough curves obtained in a fixed-bed reactor with simulated flue gas (CO₂, 10 vol%). Fig. S5† showed the breakthrough curves of all HKUST-1 samples. The CO₂ adsorption capacities were listed in Tables 1 and 2. The CO₂ adsorption capacity of bulk HKUST-1 was 0.96 mmol g⁻¹ at 30 °C and 0.1 bar. While the nano-sized HKUST-1 showed better CO₂ adsorption capacity than that of micro-sized HKUST-1. For the Cu(NO₃)₂ system, HKUST-1-N5 possessed the largest CO₂ adsorption capacity of 1.34 mmol g⁻¹. It was believed that the CO₂ adsorption capacity of nanosized HKUST-1 samples mainly depend on the porous structure.²² For the Cu(NO₃)₂ system, the enhancement of CO₂ adsorption capacity of HKUST-1-N5 was attributed to the maximum total pore volume (0.73 cm³ g⁻¹), resulted from the formation of hierarchical porous structure and inter-particle cavity. While for the Cu(OAc) system, the CO₂ adsorption capacity of HKUST-1-C1.5 and HKUST-1-C3 had the same CO₂ adsorption capacity (1.16 mmol g⁻¹) in spite of greatly different specific surface area and microporous volume percentage, which was slight higher than that of HKUST-1-C0. Therefore, it was reasonable to believe that the CO₂ adsorption capacity of HKUST-1 sample was determined not only by the porous structure, but also by some sort of interaction similar to chemical adsorption. To prove our assumption, CO₂-temperature programmed desorption (CO₂-TPD) analysis was carried out to measure the basicity of the samples.

Table 1 Textural structure of all HKUST-1 samples synthesized with Cu(NO₃)₂

Samples	SBET (m^2 g^−1)	Total pore volume (cm^3 g^−1)	Micropore volume% (t-plot)	Average pore size (nm)	CO2 adsorption capacity (mmol g^−1)
N0	1748	0.74	83.78	1.69	0.96
N1	1269	0.61	70.49	1.94	1.14
N2	1229	0.69	59.42	2.25	0.93
N3	811	0.61	45.90	2.99	0.91
N4	848	0.42	71.43	1.98	1.17
N5	1031	0.73	47.22	2.81	1.34
N6	798	0.47	59.57	2.36	1.02
T1	754	0.40	65.00	2.11	0.77
T1.5	945	0.50	64.00	2.10	1.14
T2	729	0.43	55.81	2.37	0.98
T3	336	0.37	21.62	4.36	1.04

---

Table 2 Textual structure of HKUST-1 synthesized with copper acetate and different amount of TEA

Sample	SBET (m^2 g^−1)	Total pore volume (cm^3 g^−1)	Microporous volume% (cm^3 g^−1)	Average pore size (nm)	CO2 adsorption capacity (mmol g^−1)
HKUST-1-C0	935.41	0.57	56.14	2.43	1.04
HKUST-1-C1	598.62	0.49	44.90	3.29	1.02
HKUST-1-C1.5	572.21	0.55	36.36	3.82	1.16
HKUST-1-C3	1400.30	0.65	72.30	1.85	1.16
HKUST-1-C4	1392.87	0.66	71.21	1.90	1.02

---

Fig. 8 showed the CO₂-TPD curves of HKUST-1 samples. HKUST-1 possessed a weak basic sites with weak intensity (peak α), which was similar to the previous report.⁴⁸ Therefore, CO₂ adsorption on the open metal sites was stronger than a typical van der Waals force. The open metal sites possess strong enough affinity toward CO₂ such that it is capable of capture CO₂ at ambient conditions or a lower CO₂ partial pressure. It can be observed that the medium strength basicity of nanosized HKUST-1 crystals was weaker than that of micron-sized HKUST-1 crystals, which meant the desorption of nanosized HKUST-1 was relatively easy. The second peak (peak β) was attributed to the decomposition of organic ligands, which was accordance with the results of TG results (Fig. S3†). Previous study by means of density functional theory illustrated that the open metal sites provided electrostatic interaction between the open metal charge and the CO₂ multipole moments.⁴⁹ The results further verified that the interaction between CO₂ and open metal sites of HKUST-1 exceeded the physisorption. At the same time, the electrostatic interaction is not stronger than the chemical bond such that CO₂ molecules can be easily desorbed.

Fig. 8 CO₂-TPD profiles of nanosized HKUST-1-N1, HKUST-1-T3, and microsized HKUST-1.

Based on the CO₂-TPD results, there might be two primary adsorption sites in the HKUST-1. One is the small pore cage composed of a ring of six metal dimers and six trimesate groups, while the other is the open Cu site. The CO₂ adsorption sites was shown in Fig. 9. The triangular-shaped windows that adsorb CO₂ molecules through van der Waals force was marked as no. 1 site, which could be easily desorbed. For the CO₂ adsorption on the open Cu sites, the interaction is attributed to the electrostatic force stronger than the van der Waals force (marked as no. 2 site).

Fig. 9 Possible CO₂ adsorption mechanism on the HKUST-1 adsorbents.

3.5 Regenerability of nanosized HKUST-1 in multiple cycles

Excellent adsorbents must possess a stable regenerability in practical application. Therefore, the regenerability of nanosized HKUST-1 (HKUST-1-N5) was measured. The adsorption/desorption cycles were carried out as followed: firstly, samples were adsorbed in 30 °C until saturation, then increasing the temperature to 200 °C in a argon flow to remove the adsorbed CO₂ until no CO₂ was detected in the outlet. The results illustrated that the desorption process finished in 15 min under a mild condition and the adsorption capacity was almost unchanged after ten adsorption/desorption cycles (Fig. 10), which revealed the good stability of nanosized HKUST-1 adsorbents.

Fig. 10 CO₂ adsorption capacity of nanosized HKUST-1-N5 samples for multicycle of adsorption/desorption.

4. Conclusion

In summary, the nanosized HKUST-1 was synthesized by coordination modulation method and the effects of different modulators including NaAc and TEA as well as the type of copper source on the structure and morphology of HKUST-1 crystals were discussed. The results suggest that the effect of modulator was opposite for different copper system which might due to the different nucleation rate e.g. for the Cu(NO₃)₂ system, both moderate acid–base environment and capping groups play an important role in the synthesis of uniform nanosized HKUST-1. However, for the Cu(OAc)₂ system, high concentration of TEA provide a slow nucleation rate of HKUST-1 leading to the formation of larger crystals.

The CO₂ adsorption experiments in simulated flue gases prove that the nanosized HKUST-1 had a higher CO₂ adsorption capacity than micron-sized counterparts. In particular, the CO₂ adsorption capacity of HKUST-1-N5 synthesized with Cu(NO₃)₂ and NaAc reached 1.34 mmol g⁻¹ at 30 °C and 0.1 bar, which was enhanced by 39.6% compared to that of bulk HKUST-1. Hierarchical porous structure and uniformed nanosized HKUST-1 particles possess high surface area and total pore volume, and expose more open metal sites for CO₂ adsorption. After ten successive adsorption/desorption cycles, the adsorbents keep an excellent adsorption property. Moreover, the regeneration process is simple and quick.

Acknowledgements

The authors acknowledge supports from National Natural Science Foundation of China (21306217), “Strategic Priority Research Program-Climate Change: Carbon Budget and Related Issues” of the Chinese Academy of Sciences (XDA05010109, 05010110), the Instrument Developing Project of the Chinese Academy of Sciences (YZ201139), the Chinese Academy of Sciences the strategic pilot science and technology projects-carbon dioxide capture, use and storage key technology and engineering demonstration (XDA070401) and the Key Science and Technology Program of Shanxi Province, China (MD2014-09).

References

1. D. Britt, H. Furukawa, B. Wang, T. G. Glover and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 20637.
2. H. Choi, Y. C. Park, Y. H. Kim and Y. S. Lee, J. Am. Chem. Soc., 2011, 133, 2084.
3. S. Choi, T. Watanabe, T.-H. Bae, D. S. Sholl and C. W. Jones, J. Phys. Chem. Lett., 2012, 3, 1136.
4. X. Jiao, L. Li, N. Zhao, F. Xiao and W. Wei, Energy Fuels, 2013, 27, 5407.
5. R. Krishna and J. M. van Baten, Sep. Purif. Technol., 2012, 87, 120.
6. J. Liu, J. Tian, P. K. Thallapally and B. P. McGrail, J. Phys. Chem. C, 2012, 116, 9575.
7. J. Yu, Y. Ma and P. B. Balbuena, Langmuir, 2012, 28, 8064.
8. F. Schroeder and R. A. Fischer, in Functional Metal–Organic Frameworks: Gas Storage, Separation and Catalysis, ed. M. Schroder, 2010, vol. 293, p. 77.
9. J. Liu, P. K. Thallapally, B. P. McGrail, D. R. Brown and J. Liu, Chem. Soc. Rev., 2012, 41, 2308.
10. T.-H. Bae, M. R. Hudson, J. A. Mason, W. L. Queen, J. J. Dutton, K. Sumida, K. J. Micklash, S. S. Kaye, C. M. Brown and J. R. Long, Energy Environ. Sci., 2013, 6, 128.
11. R. Thiruvenkatachari, S. Su, H. An and X. X. Yu, Prog. Energy Combust. Sci., 2009, 35, 438.
12. X. Wang, X. Ma, C. Song, D. R. Locke, S. Siefert, R. E. Winans, J. Möllmer, M. Lange, A. Möller and R. Gläser, Microporous Mesoporous Mater., 2013, 169, 103.
13. X. Yan, L. Zhang, Y. Zhang, K. Qiao, Z. Yan and S. Komarneni, Chem. Eng. J., 2011, 168, 918.
14. R. Haghpanah, A. Rajendran, S. Farooq and I. A. Karimi, Ind. Eng. Chem. Res., 2014, 53, 9186.
15. K. Sumida, D. L. Rogow, J. A. Mason, T. M. McDonald, E. D. Bloch, Z. R. Herm, T. H. Bae and J. R. Long, Chem. Rev., 2012, 112, 724.
16. A. Corma, H. García, F. X. Llabrés and I. Xamena, Chem. Rev., 2010, 110, 4606.
17. J. L. Liu, S. Bashir and H. C. Zhou, Abstr. Pap. Am. Chem. Soc., 2012, 244, 1.
18. A.-H. Lu and G.-P. Hao, Annu. Rep. Prog. Chem., Sect. A: Inorg. Chem., 2013, 109, 484.
19. H. Furukawa, N. Ko, Y. B. Go, N. Aratani, S. B. Choi, E. Choi, A. O. Yazaydin, R. Q. Snurr, M. O'Keeffe, J. Kim and O. M. Yaghi, Science, 2010, 329, 424.
20. H. Furukawa, K. E. Cordova, M. O'Keeffe and O. M. Yaghi, Science, 2013, 341, 1230444.
21. N. Stock and S. Biswas, Chem. Rev., 2012, 112, 933.
22. S. Liu, L. Sun, F. Xu, J. Zhang, C. Jiao, F. Li, Z. Li, S. Wang, Z. Wang, X. Jiang, H. Zhou, L. Yang and C. Schick, Energy Environ. Sci., 2013, 6, 818.
23. B. Supronowicz, A. Mavrandonakis and T. Heine, J. Phys. Chem. C, 2013, 117, 14570.
24. J. Moellmer, A. Moeller, F. Dreisbach, R. Glaeser and R. Staudt, Microporous Mesoporous Mater., 2011, 138, 140.
25. J. Liu, Y. Wang, A. I. Benin, P. Jakubczak, R. R. Willis and M. D. LeVan, Langmuir, 2010, 26, 14301.
26. J. Zhang, L. Sun, F. Xu, F. Li, H.-Y. Zhou, Y.-L. Liu, Z. Gabelica and C. Schick, Chem. Commun., 2012, 48, 759.
27. W. Shang, X. Kang, H. Ning, J. Zhang, X. Zhang, Z. Wu, G. Mo, X. Xing and B. Han, Langmuir, 2013, 29, 13168.
28. Z. Li, Y. Tai, M. Zhang and L. Qiu, Chem. Res. Appl., 2014, 26, 13.
29. Z. Ni and R. I. Masel, J. Am. Chem. Soc., 2006, 128, 12394.
30. J.-S. Choi, W.-J. Son, J. Kim and W.-S. Ahn, Microporous Mesoporous Mater., 2008, 116, 727.
31. L. Huang, H. Wang, J. Chen, Z. Wang, J. Sun, D. Zhao and Y. Yan, Microporous Mesoporous Mater., 2003, 58, 105.
32. D. J. Tranchemontagne, J. R. Hunt and O. M. Yaghi, Tetrahedron, 2008, 64, 8553.
33. J. Huo, M. Brightwell, S. El Hankari, A. Garai and D. Bradshaw, J. Mater. Chem. A, 2013, 1, 15220.
34. G. Majano and J. Pérez-Ramírez, Helv. Chim. Acta, 2012, 95, 2278.
35. D. Zacher, J. Liu, K. Huber and R. A. Fischer, Chem. Commun., 2009, 1031,  10.1039/b819580c.
36. A. Ranft, S. B. Betzler, F. Haase and B. V. Lotsch, CrystEngComm, 2013, 15, 9296.
37. H. Guo, Y. Zhu, S. Wang, S. Su, L. Zhou and H. Zhang, Chem. Mater., 2012, 24, 444.
38. F. Wang, H. Guo, Y. Chai, Y. Li and C. Liu, Microporous Mesoporous Mater., 2013, 173, 181.
39. Y. Zhou and B. Yan, Inorg. Chem., 2014, 53, 3456.
40. T. Tsuruoka, S. Furukawa, Y. Takashima, K. Yoshida, S. Isoda and S. Kitagawa, Angew. Chem., 2009, 48, 4739.
41. S. Diring, S. Furukawa, Y. Takashima, T. Tsuruoka and S. Kitagawa, Chem. Mater., 2010, 22, 4531.
42. L. Li, N. Zhao, W. Wei and Y. Sun, Fuel, 2013, 108, 112.
43. E. Borfecchia, S. Maurelli, D. Gianolio, E. Groppo, M. Chiesa, F. Bonino and C. Lamberti, J. Phys. Chem. C, 2012, 116, 19839.
44. J. Gorka, P. F. Fulvio, S. Pikus and M. Jaroniec, Chem. Commun., 2010, 46, 6798.
45. E. Biemmi, C. Scherb and T. Bein, J. Am. Chem. Soc., 2007, 129, 8054.
46. H. R. Abid, H. M. Ang and S. Wang, Nanoscale, 2012, 4, 3089.
47. D.-A. Yang, H.-Y. Cho, J. Kim, S.-T. Yang and W.-S. Ahn, Energy Environ. Sci., 2012, 5, 6465.
48. J. Kim, S.-N. Kim, H.-G. Jang, G. Seo and W.-S. Ahn, Appl. Catal., A, 2013, 453, 175.
49. H. Wu, J. M. Simmons, G. Srinivas, W. Zhou and T. Yildirim, J. Phys. Chem. Lett., 2010, 1, 1946.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra03986j

---