Effect of additives on the growth of HKUST-1 crystals synthesized by microfluidic chips with concentration gradient

Abstract

Metal-organic frameworks (MOFs) have attracted considerable attention as novel nanoporous materials that combine the properties of organic and inorganic porous materials. HKUST-1 is one of the most well-developed and representative MOFs with wide applications in gas storage and separation, adsorption, and capture. In this study, we used microfluidics, an advanced technique of manipulation of small fluid volumes in microscale or even nanoscale channels, to investigate the effect of sodium dodecyl sulfate (SDS) on the growth of HKUST-1 crystals. We directly observed the morphological evolution of HKUST-1 crystals through droplet arrays with the SDS concentration gradient. The morphology of HKUST-1 evolved from cubic to cuboctahedron and finally to octahedron with increasing SDS concentration. The study results demonstrated the important role played by anions in solution in the structural regulation of HKUST-1.

INTRODUCTION

Metal-organic frameworks (MOFs), also called porous coordination polymers (PCPs), are formed by coordination of metal ions and organic linkers, producing various structures and topologies.^1,2 Thus, MOFs combine the properties of organic and inorganic porous materials. MOFs have attracted increasing interest owing to their extremely high specific surface area, uniform nanoporosity, flexible structure that can be well designed, and wide range of applications in various fields,^3,4 including gas storage and separation,^5,6 catalysis,⁷ sensors,⁸ electronics and optics,⁹ energy storage,¹⁰ and diagnostics and controlled drug release.¹¹

HKUST-1,¹² also called MOF-199 [Cu₃(BTC)₂, BTC = 1,3,5-benzenetricarboxylate], is one of the well-developed MOFs with reasonable thermal stability. It has a cubic crystal structure with a squared channel pore size of 1 nm and is composed of Cu paddlewheel [Cu₂(CO₂)₄] secondary building units (SBUs).¹² In the last two decades, HKUST-1 has generated considerable excitement as a representative porous material having wide applications in gas storage and separation,¹³ adsorption,^14,15 and capture.¹⁶ Many methods have been used to fabricate this material, including the solvothermal method,¹⁷ microwave-assisted strategy,¹⁸ aerosol-assisted method,¹⁹ and mechanochemical method.²⁰ However, currently, there is a need to develop an advanced synthesis method of HKUST-1 for its extended application by controlling its shape and size and for realizing higher efficiency and environmental friendliness. Additives and capping reagents have been used to synthesize nanosized MOFs with controlled shape and size because these materials can affect the nucleation and growth process of crystals. For example, Sun et al.²¹ controlled the morphology of HKUST-1 crystals by using the surfactant cetyltrimethylammonium bromide (CTAB) as a structure director based on solution-based direct precipitation at room temperature. They also found that the addition of different inorganic salts, including sodium and potassium, could simply tune and control the particle morphology and size of HKUST-1 nanocrystals.²² Wee et al.²³ found that the synthesis procedure involving either reactant mixing and freeze drying or mixing of cooled reagents efficiently tunes the HKUST-1 crystal size in the sub-micrometer range. However, the crystals prepared by these methods have poor uniformity of crystal size distribution because of drastic temperature changes during synthesis. Although Hwang et al.²⁴ achieved the size and morphological control of HKUST-1 crystals by changing the solvent and coordination modulator, the crystal size was nonuniform because the microwave-assisted reaction processing results in rapid crystallization. Moreover, the solvent used in this method was diethylene glycol, which is toxic, making the method environmentally malignant.

The synthesis of MOFs using microfluidics, as an alternative method to traditional batch reactions, has attracted considerable research attention in recent years. Microfluidics,²⁵ also called lab-on-chip, can manipulate the small volume of fluid in micrometer-scale channels, supporting processes such as mixing, separation, merging, and diluting. Droplet microfluidics²⁵ is an essential branch of the microfluidics technique. Microdroplets provide a confined environment analogous to cells; droplet microfluidics is ideally suited to compartmentalize and isolate reactants. It has made splendid progress in chemistry and biology^26–29 with compelling physical advantages, including reduced reagent consumption, efficient mass and heat transfer, rapid reaction, and high-throughput and large-scale production. Microfluidics has also been applied to fabricate versatile MOFs with various topologies and functions,³⁰ including HKUST-1,^31,32 ZIF-8,^33,34 UiO-66,³⁵ MIL-88A,³⁶ MOF-5, and IRMOF-3.³⁷

Microfluidics provides a new platform for crystal engineering owing to its various advantages such as simplicity, versatility, controllability, and efficiency. Inspired by the control strategy of biomineralization in organisms, we achieved control over calcium carbonate crystallization using a microfluidic chip called a “crystal hotel.” The crystal hotel device was designed to provide a series of confined reaction volumes of well-defined shape, size, and internal patterning into which a controlled flow of reactant ions and additives could be achieved. In addition, by utilizing the optical transparency of poly(dimethylsiloxane) (PDMS), we presented the definitive real-time growth of calcium carbonate crystals.³⁸ Because crystallization is a rapid process in bulk preparation, it is difficult to obtain information about crystal nucleation and growth.³⁹ Besides, using another crystal hotel microfluidic chip with a different structure, we studied the effects of additives on the early stages of growth of calcite single crystals.

In this paper, we report a facile and economical method to prepare HKUST-1 crystals using a microfluidic chip with 48 defined “crystal hotel rooms” (Fig. 1). The morphology of HKUST-1 microcrystals can be easily controlled by the addition of sodium dodecyl sulfate (SDS, C₁₂H₂₅SO₃Na). With increasing SDS concentration, the morphology of HKUST-1 evolves from cubic to cuboctahedron and finally to octahedron via an array of droplets with SDS concentration gradients produced by the microfluidic chip. The results demonstrate that metal sodium ions and ${\mathrm{C}}_{12}{\mathrm{H}}_{25}{\mathrm{SO}}_3^{-}$ anions have a combined effect on the structure of HKUST-1 crystals.

FIG. 1.

The Crystal Hotel microfluidic device. (a) Photograph of a PDMS device filled with a solution of red dye. (b) Chip design of a Crystal Hotel with 48 “rooms.” (c) Crystallization in a single room (i). An aqueous solution (light blue) is introduced through inlet 2 to fill the channel and rooms (ii). Subsequently, air (white) is introduced through inlet 1 to push the solution out of the channel and isolate the solution contained in each room (iii). The additives (yellow) are then pumped through inlet 2 (iv). Finally, the air is introduced through inlet 1 to push the additives out of the channel.

EXPERIMENTAL

Sample preparation

All reagents were purchased commercially and used without purification. The HKUST-1 precursor solution was prepared by adding Cu(NO₃)2 ⋅ 3H₂O (5 mmol, 1.22 g, Aladdin Industrial Corporation, Shanghai) and trimesic acid (2.5 mmol, 0.58 g, Aladdin Industrial Corporation, Shanghai) to dimethyl sulfoxide (DMSO; 5 g, Aladdin Industrial Corporation, Shanghai) at room temperature. NaCl (0–0.4676 g, 0–8 mmol, Sinopharm Chemical Reagent Co., Ltd., Shanghai) was dissolved in 20 ml of the precursor solution and treated with ultrasound for 15 min at room temperature to form a transparent solution. The concentration of NaCl was set to be 0, 0.1, 0.2, or 0.4M. The surfactant SDS (0.2724 g, 1 mmol, Aladdin Industrial Corporation, Shanghai) was added in 20 ml of the precursor solution and treated with ultrasound for 15 min at room temperature until a transparent solution with a concentration of 50 mM was obtained. Then, precursor solutions with SDS concentrations 1, 5, and 10 mM were obtained by diluting the 50 mM solution. The 10 mM SDS/DMSO solution was prepared by dissolving 0.0544 g of SDS (0.2 mmol) in 20 ml of DMSO treating it with ultrasound for 30 min at room temperature.

Fabrication of microfluidic chips

The optic microfluidic chips were built with PDMS by typical soft lithography.⁴⁰ First, 4-in. silicon wafers were coated with photoresist of SU-8 2075 (MicroChem Co., USA) and were exposed to UV radiation for 8 s before they were placed in the SU-8 developer for developing a template structure. Next, a PDMS precursor mixture with 10:1 of the prepolymer and curing agent was poured over the patterned silicon wafer; subsequently, the silicon wafer was reacted with (tridecafluoro-1,1,2,2-tetrahydrooctyl)-1-trichlorosilane (UCT) by vapor deposition with 120 °C. After the wafers were cured at 65 °C for 1 h, the inlet and outlet of the chips were punched using a 1-mm-diameter punch. Finally, the PDMS replicas were irreversibly bonded to a glass slide via oxygen plasma (Harrick Plasma). The straight and main fluid channels have a width of 200 μm and a height of 69 μm. There are 48 rooms in each microfluidic chip, with each room having a diameter of 450 μm and constriction width and length of 40 μm and 100 μm, respectively.

Preparation of static droplet arrays with tunable gradients

The method used in this study differs from Sun's⁴¹ method in that air was used as a continuous phase in this study. Sun's method is not suitable for samples that are soluble in oil. This will not happen if the air is injected from the opposite direction, and the air will not contaminate the sample. The dilution process was performed in the following four stages. First, a black dye aqueous solution was injected at a rate of 5 μl/min into the microchannels with a syringe linked to a syringe pump (PHD2000, Harvard Apparatus, USA). Second, the air was injected at a rate of 1 ml/min to push the solution in the microchannel and not in the rooms. Next, 1.8 μl of de-ionized water was injected as a dilution agent at a rate of 0.3 μl/min. Finally, the air was injected from the opposite direction at a rate of 1 ml/min. In this study, to obtain a regular and wider range of gradients, initially, the volume of de-ionized water (1.8 μl) was kept constant while varying its injection rate from 0.3, 0.7, 1.0 to 1.3 μl/min. Then, the injecting volume of de-ionized water was adjusted from 1.8, 2.7 to 3.0 μl, keeping the injection rate constant at 0.3 μl/min. Finally, a bidirectional concentration gradient curve was obtained by varying the injection sequence of the black dye aqueous solution and de-ionized water and setting the volume of the last injection solution from 1.8, 2.2, 2.5, 2.9, 3.2, 3.6 to 4.3 μl, while keeping the last injection rate constant at 0.3 μl/min.

Synthesis of HKUST-1 crystals by bulk reaction

10 μl of the precursor solutions with different concentrations of NaCl (i.e., 0.4, 0.2, 0.1, and 0M) and SDS (1, 5, 10, and 50 mM) were dropped onto a hydrophilic modified glass slide. It was naturally dried in a clean environment.

Synthesis of HKUST-1 crystals in microfluidic chips

The crystal precursor solution containing different concentrations of additives was injected into the chip at a rate of 5 μl/min, and the air was pushed in at a rate of 1 ml/min in the reverse direction so that the crystal precursor solution was retained only in each crystal chamber.

Then, the effect of SDS on the morphology of HKUST-1 formed on the droplet array with SDS concentration gradients produced by the microfluidic chip was studied. For this analysis, first, the HKUST-1 precursor solution was injected at a rate of 5 μl/min. Then, the air was pushed in the reverse direction at a rate of 1 ml/min, so that the crystal precursor solution was only retained in the crystal chamber. Finally, a 10 mM SDS/DMSO solution was introduced into the chip at a rate of 0.3 μl/min. The injection was stopped when the volume reached 1.8 μl (from Room 1 to Room 48).

All reactions were performed under normal pressure at room temperature.

Characterization

Separate the PDMS chip from the glass slide and take out the sample in the room for characterization. The crystal morphologies were characterized by JEOL JSM-7800F Prime field-emission scanning electron microscopy (SEM) at a voltage of 5 kV and using a secondary electron detector in a vacuum environment. X-ray diffraction (XRD) measurements of powders and patterns were performed using a Schmalz thin film (Cu-Kα1 = 1.5406 Å; Cu-Kα2 = 1.5444 Å) in θ/2θ geometry. The samples were measured between 5° and 50°, with a step-size of 0.04° and a scan-speed of 0.3°/min.

Concentration gradient experiment based on microfluidics

Based on the results of Sun et al.,⁴¹ two main parameters were set to adjust the concentration gradient and range in droplet arrays: the flow rate and the volume of the diluting liquid. First, we studied the effects of the diluting liquid flow rate on the range of gradients of the droplet array. Figure 2(a) presents the dilution results when the volume is kept constant; the concentration gradient distribution becomes narrower as the dilution rate increases. This implies that a lower flow rate of the diluting liquid can yield a wider range and steeper gradient. This phenomenon can be attributed to the increasing duration for dilution when the flow rate of the diluting liquid is reduced. At a dilution flow rate of 0.2 μl/min, the range of concentration gradient was 0.38 C₀ to 0.97 C₀ (where C₀ is the initial concentration before dilution). This flow rate is not very stable because the extremely low rate leads to the capillary force in the microchannel being more prominent. Therefore, to analyze the effect of dilution volume on the concentration gradient, a dilution rate of 0.3 μl/min was used; at this rate, the concentration gradient ranged from 0.42 C₀ to 0.89 C₀.

FIG. 2.

Concentration gradient varied by adjusting the dilution parameters. (a) Varying dilution rate at a constant dilution volume of 1.8 μl. (b) Varying dilution volume at a constant dilution rate of 0.3 μl/min.

Furthermore, the volume of the diluting liquid was varied to tune the gradient range. To obtain a wider range of gradients of the droplet array, the dilution rate was maintained at 0.3 μl/min, and the concentration gradient curve was obtained by varying the dilution volume [Fig. 2(b)]. The experimental results indicate that the concentration in each room decreases with increasing dilution volume, while the slope of the concentration gradient remains almost constant, suggesting that the concentration gradient range remains almost unchanged. By reversing the order of solution injection, that is, the aqueous solution and the ink solution, a bidirectional concentration gradient curve with different gradient ranges can be obtained by varying the dilution volume at a constant flow rate (0.3 μl/min). From the bidirectional concentration gradient curve, the concentration value of multiple components in a certain room can be determined (Fig. 3). This is very important when multiple-reagent reactions are processed; for example, when two kinds of solutions (e.g., A and B) are successively introduced into a chip, the concentration of A and B can be obtained at a certain position in the chip by reading this curve. Thus, high-efficiency chemical experiments can be conducted.

FIG. 3.

Bidirectional concentration gradient curve obtained by varying the dilution volume while keeping the dilution rate constant at 0.3 μl/min.

Effect of SDS on the morphology of HKUST-1 crystals formed by bulk reaction vs microfluidics without concentration gradients

The morphology of HKUST-1 crystals prepared by bulk reaction and a microfluidic device both gradually evolved from truncated cubic to cubic octahedron, truncated octahedron, and octahedron with the increase in SDS concentration. The evolution of the morphology of HKUST-1 crystals prepared by the microfluidic method was more obvious and controlled than that prepared in bulk. As Fig. 4 shows, HKUST-1 polycrystals were cubic crystals when no additives were present; when 1 mM SDS was added to this reacting system, crystals with a truncated cube structure of uniform size of 1 μm appeared. When the SDS concentration increased to 5 mM, the obtained crystals predominantly had cubic octahedral structures and the crystal size was relatively uniform, 1–2 μm. When the SDS concentration was 10 mM, the crystals produced mostly had truncated octahedral structures with slightly larger crystals, approximately 3 μm. When the SDS concentration increased to 50 mM, the morphology of the crystal was an octahedral structure with significantly reduced size, and some crystals with hollow structure were present.

FIG. 4.

Scanning electron micrography (SEM) images of HKUST-1 crystals prepared by bulk reaction [(a)–(d)] and the microfluidic method [(e)–(h)] with SDS addition at different concentrations at room temperature. The concentration of SDS was 1 mM [(a) and (e)], 5 mM [(b) and (f)], 10 mM [(c) and (g)], and 50 mM [(d) and (h)].

The regulation of the crystal morphology of HKUST-1 by SDS was consistent with the structure-directing effect of NaCl added in the reaction system (Figs. S1–S3 in the supplementary material). However, at the same SDS concentration, the size uniformity of crystals obtained by the bulk reaction was inferior to the size uniformity of crystals obtained by the microfluidic method. This is because the microfluidic chip used here provided a well-defined reacting position for crystallization. The rooms in this microfluidic device had a defined volume and were independent of each other, which facilitated the reaction with a quantitative volume and an isolated, confined environment. These conditions are important for crystallization to ensure an unaffected reaction environment comprising quantitative materials and uniform reaction conditions. This implies that the amount of additives was equal, which is beneficial for us to study the effect of the concentration of the additives on the morphology of crystals. Crystallization of HKUST-1 is a self-assembly process induced by evaporation, so the evaporation rate has a significant effect on crystallization behavior. In the rooms of microfluidic chips, the evaporation rate is uniform and not largely affected by the airflow rate, which has a great effect on the crystallization process in bulk reaction. The microfluidic chip used in this study consists of microchannels that are several hundred micrometers wide, largely reducing the diffusion distance. This improves the nucleation rate and promotes a uniform concentration distribution in each room, resulting in the uniform occurrence of nuclei in the solution. The formation of nuclei is the main reason for finally forming fairly uniform crystals in the microfluidic device.

The XRD patterns presented in Fig. 5 show that, for the crystals prepared by the bulk reaction, the signal intensity of the (222) crystal facets increased significantly concerning the signal intensity of the (400) crystal facets as the SDS concentration increased. As for the crystals fabricated in the microfluidic chip, the signal intensity of the (222) crystal facets was stronger, while that of the (400) crystal facet was weaker with increasing SDS concentration.

FIG. 5.

XRD patterns of HKUST-1 crystals prepared via different methods under the effect of SDS addition. (a) Crystals prepared via the bulk method; (b) crystals prepared via the microfluidic device.

Effect of SDS on the morphology of HKUST-1 crystals prepared by microfluidics with concentration gradients

Based on the results of the previous concentration gradient experiments, a static droplet array containing a bidirectional concentration gradient of SDS and the crystal precursor was prepared using the microfluidic chip containing 48 crystal rooms. Since the SDS concentration exceeded 10 mM, it did not considerably affect the morphology of HKUST-1 crystals. Therefore, to produce the concentration gradient, a DMSO solution was added along with 10 mM SDS. Figure 6(a) presents the crystal array produced under this condition and Fig. 6(b) shows the SDS concentration and precursor concentration in different droplets obtained from the concentration gradient experiments. The figures indicate that the crystals in the room increased from Room 1 to Room 48 with increasing concentration of the precursor solution [the black areas in Fig. 6(a) are extremely crystallized HKUST-1 obtained in each room].

FIG. 6.

Concentration gradient droplet array prepared by the microfluidic chip to investigate the effect of SDS concentration on the crystallinity of HKUST-1. (a) Optical micrograph of the prepared crystal array. (b) The composition concentration gradient curve of the 48 rooms.

The SEM images in Fig. 7 present the clear evolution of the morphology of HKUST-1 crystals under the effect of different concentrations of SDS: from octahedron, gradually to the truncated octahedron, and then into a cubic octahedron, and finally to cubic structures. Room 1 consisted the highest SDS concentration and the lowest precursor concentration, which was 0.27 C₀ (Cu²⁺_C0 = 1M, BTC³⁻_C0= 0.5M). The above results indicate that when the SDS concentration was 10 mM, Cu²⁺ concentration was 1M and BTC³⁻concentration was 0.5M. The crystals obtained in Room 1 had a uniform truncated octahedron structure because the concentration ratio of SDS to the precursor in this room was considerably larger than the concentration ratio in Room 48, where the SDS concentration was 10 mM; lower SDS concentration and higher precursor concentration yielded small uniform cubic crystals. In this experiment, from Room 1 to Room 48, the SDS concentration decreased, whereas the precursor concentration increased; therefore, the shape of the crystals formed changed from octahedron, truncated octahedron, cubic octahedron, truncated cube to cube.

FIG. 7.

SEM images of HKUST-1 crystals prepared with droplet arrays with SDS concentration gradient by using a microfluidic chip.

Effect of SDS on growth of HKSUT-1

By comparison with the crystals prepared by adding NaCl to the precursor solution (Fig. S2 in the supplementary material), those obtained by adding SDS are smaller in size, and the effect of SDS on the crystal morphology is much more efficient because the amount of SDS added is much less than that of NaCl to obtain a similar morphological change of HKUST-1 crystals. Based on these observations, we believe that SDS acts as an anionic surfactant that can ionize Na⁺ ions and dodecyl sulfonate (${\mathrm{C}}_{12}{\mathrm{H}}_{25}\mathrm{S}{\mathrm{O}}_3^{-}$) anions in polar solutions (DMSO). After adding SDS to the precursor solution of HKUST-1 crystals, Na⁺ and ${\mathrm{C}}_{12}{\mathrm{H}}_{25}\mathrm{S}{\mathrm{O}}_3^{-}$ ions are surrounded by BTC³⁻ ions and Cu²⁺ ions, respectively, due to the electrostatic interaction; this greatly hinders the formation of SBUs in the reacting system. The rate of formation of the {111} crystal facets with lower surface energy reduces; therefore, the truncated cube and the truncated octahedron structures in which the {111} plane and the {100} plane coexist in the crystal structure appear. As the SDS concentration further increases, Na⁺ ions and dodecyl sulfonate anions hinder the binding of BTC³⁻ ions and Cu²⁺ ions, and the area occupied by the {111} plane in the crystal structure becomes larger. The final crystallization is an octahedral structure composed of eight {111} faces. When the concentration of SDS is further increased because Na⁺ ions and dodecyl sulfonate anions around BTC³⁻ ions and Cu²⁺ ions reach saturation, generation of more resistance to crystal crystallization is no longer possible; thus, the shape of the crystal remains as octahedral. However, due to the electrostatic interaction, the SDS ionization ions are difficult to diffuse and aggregate ions due to the clustering of BTC³⁻ ions and Cu²⁺ ions, making the final crystal growth difficult. Therefore, when the SDS concentration is 50 mM, the crystal size is greatly reduced. Based on this, the effect mechanism of SDS on HKUST-1 crystal morphology is proposed (Fig. 8).

FIG. 8.

Proposed mechanism of SDS on the morphology of HKUST-1 crystals.

CONCLUSIONS

In this study, a microfluidic chip that can generate a static droplet array with a continuous concentration gradient in one step was designed. The concentration gradient of the droplet array can be tuned by varying the dilution rate and dilution volume. The effects of the addition of NaCl and SDS on the morphology of HKUST-1 crystals were investigated and compared. The morphology of crystals evolved from cube to truncated cube, cubic octahedron, truncated octahedron, and finally to octahedron. Compared with the effect of NaCl on the morphology of HKUST-1 crystals, SDS tunes the morphology of the crystals more efficiently. SDS can act as a capping agent during HKUST-1 synthesis, and hence, it is expected that longer times can lead to more well-defined crystals habit. For the preparation of HKUST-1 crystals, compared with the bulk reaction method, the microfluidic method's reaction was faster and the orientation of the crystals obtained by this method was more regular. Through the microfluidic chip approach, 48 droplets with a wide range of the concentration gradient of SDS and precursor solution could be produced, which facilitated clear observation of the evolution of HKUST-1 crystals under the effect of SDS.

SUPPLEMENTARY MATERIAL

See the supplementary material for the effect of NaCl on the growth of HKUST-1 crystals.

AUTHORS’ CONTRIBUTION

Q.W. and X.W. contributed equally to this work.

DATA AVAILABILITY

The data that support the findings of this study are available within the article and its supplementary material.