Exploiting a Lattice-Driven Gating in a Cu-Based Zeolitic Imidazolate Framework for Efficient High-Temperature Hydrogen Isotope Separation

doi:10.21203/rs.3.rs-4724831/v1

Download PDF

Article

Exploiting a Lattice-Driven Gating in a Cu-Based Zeolitic Imidazolate Framework for Efficient High-Temperature Hydrogen Isotope Separation

https://doi.org/10.21203/rs.3.rs-4724831/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study presents a novel approach for hydrogen isotope separation using Cu-ZIF-gis, a Cu-based zeolitic imidazolate framework with ca. 2.4 Å channels. Traditional Kinetic Quantum Sieving(KQS) method involves exploiting diffusion barriers, caused by the flexibility of local linker molecules or unit cell breathing phenomena. However, these methods are ineffective at high temperatures, as such opening transition typically completes below 77K. Cu-ZIF-gis leverages a new approach named "Lattice-driven gating", which uses lattice expansion to control the aperture at elevated temperatures up to 180K. Despite the lack of strong binding sites, it achieves effective isotope uptake via KQS above 120K and shows exceptional stability, remaining intact for nearly 3 years in air. With the help of Quasi-Elastic Neutron Scattering(QENS), we observe substantial differences in the molecular mobility of H₂ and D₂ above 150K. This method is compatible with existing LNG cryo-infrastructure, marking a significant advancement in sustainable isotope separation technologies.

Physical sciences/Chemistry/Energy

Physical sciences/Chemistry/Chemical engineering

Hydrogen Isotope Separation

Zeolitic Imidazolate Framework (ZIF)

Lattice-Driven Gating

Kinetic Diffusion Limitation

Deuterium, a stable isotope of hydrogen, plays a critical role in enhancing the durability and luminous efficiency of semiconductors and display devices, as well as serving as a fusion fuel for energy production.^1,2 By exchanging hydrogen-deuterium at the interface, for instance, semiconductor devices can improve their lifespan by 10–50 times.³ This significant improvement has garnered considerable interest from the electronics industry. Consequently, there is a growing demand for energy-efficient and environmentally sustainable hydrogen isotope separation technologies. However, separating hydrogen isotopes in industrial processes poses significant challenges due to their identical physicochemical properties, leading to energy-intensive operations and low selectivity at elevated temperatures.⁴ Traditional methods like cryogenic distillation dominate the field but are notorious for their high energy consumption and operational inefficiencies.⁵ Although effective at low temperatures, these techniques struggle to maintain performance as temperatures rise, leading to increased costs and environmental impacts. Recently, confined porous systems have emerged as efficient methods for hydrogen isotope separation. Crystalline nanoporous materials can separate isotopes through nuclear quantum effects, which amplify the isotope effect due to the large mass ratio between H₂ and D₂.⁶ Two primary nuclear quantum effects, Kinetic Quantum Sieving (KQS) and Chemical Affinity Quantum Sieving (CAQS), have been explored for D₂/H₂ separation. KQS-based separation relies on differences in de Broglie wavelength at cryogenic temperature, while CAQS operates based on differences in zero-point energy (ZPE), with D₂ exhibiting stronger binding energy due to its lower ZPE.^{7,8,9,10,11,12,13,14} Although CAQS-based isotope separation shows high selectivity at elevated temperatures and seems promising^{8,13,15,16,17,18}, it has a critical drawback: the instability of open metal sites. These sites are prone to oxidation and structural degradation, severely compromising the material's durability.¹⁹ Therefore, achieving hydrogen isotope separation at high temperatures using a structurally stable method like KQS would be advantageous for practical applications.

To achieve efficient KQS, diffusion barrier differences for selected isotopes must be created. For that, current methods have utilized local flexibility (some molecules/atoms at the aperture vibrate) or breathing phenomenon (the unit cell structure undergoes rapid changes) to create these barriers. Both of these phenomena for isotope separation typically start to be observed below 77 K, leading to cryogenic flexibility. Previous studies have shown that the selectivity for hydrogen isotopes (S_D2/H2) through KQS decreases with increasing temperature as these barrier differences typically vanish above 77 K. For example, in the breathing phenomenon, Kim et al. observed a decrease in S_D2/H2 in MIL-53(Al) from 13.6 at 40 K/10 mbar to 2.05 at 70 K/10 mbar.²⁰ Similarly, in the case of local flexibility, Teufel et al. and Muhammad et al. reported reductions in S_D2/H2 from 6.9 (40 K) to 2.8 (70 K) in MFU-4(Zn) and from 16.7 (25 K) to 2.9 (60 K) in CoFA, respectively, as temperatures rose.^9,21 This trend underscores the challenge of maintaining high S_D2/H2 at elevated temperatures using KQS. However, one promising approach to achieving high selectivity for KQS at elevated temperatures is through thermal gating in a relatively rigid structure. Temperature-controlled pore size modulation in metal-organic frameworks (MOFs) can dynamically control effective pore sizes to optimize gas separation. Increasing temperature induces thermal expansion within the framework, contributing to a thermal gating phenomenon^22,23,24 that facilitates selective gas adsorption based on temperature changes. The conventional gating effect has been utilized to explore various gas storage and separation applications, including CO₂,²⁵ CH₄,²⁶ H₂,²⁷ and N₂.²⁸ However, very few examples have been reported of using this temperature gating effect in hydrogen isotope separation. For instance, a covalent organic framework (COF) functionalized with pyridine showed thermal pore opening only at cryogenic temperature, and as a result, the maximum uptake was observed at 50 K and demonstrated the S_D2/H2 of 3.8 at 50 K/14 mbar.¹⁴ FMOF-Cu possesses three interconnected cavities with bottleneck apertures measuring 3.6 Å and 2.5 Å. The cavities become accessible in succession after the thermal gating aperture. All cavities are fully accessible at 77 K, and a maximum S_D2/H2 of 4 was reported at 77 K/50 mbar.²⁹ A recent report involving 2α, a metal-organic cage (MOC), also demonstrated that the maximum uptake was observed at 77 K after temperature-triggered gate opening, and S_D2/H2 of 2.8 was recorded at 77 K/200 mbar.³⁰ Their strategies aimed to make locally flexible molecules or atoms more rigid so that vibrations would only occur at high temperatures (above 100 K). However, despite these efforts, the local molecular vibration approach was insufficient to desorb hydrogen isotopes effectively above 100 K. Therefore, instead of relying on molecular vibrations, using thermal lattice expansion to control pore size could be more effective for high-temperature pore regulation.

Hence, we propose a new approach called "Lattice-driven Gating (LDG)" effect. This method involves using the expansion of the entire lattice to modulate the pore size at high temperatures specifically. In this study, we are focusing on highly flattened tetrahedral Cu-based zeolitic imidazolate frameworks (Cu-ZIFs) material that exhibits an LDG effect, which allows for a significant enhancement in hydrogen isotope uptake at elevated temperatures. The H₂ binding energetics of Cu-ZIF-gis is significant as the complete desorption requires heating up to ~ 200 K, the highest H₂ desorption temperature range for any MOFs without open metal sites. Thus, Cu-ZIF-gis is expected to show high S_D2/H2 through KQS invoked by optimal pore size created by the lattice-driven gating phenomenon at elevated temperatures under 2,000 mbar. Hence, unlike previous studies that required low operational temperatures with a few tens of mbar, our findings demonstrate the potential of Cu-ZIF-gis to integrate seamlessly into existing infrastructures, such as those used in LNG (at 111 K) processing.

Pore structure characterization and single-gas sorption isotherm of Cu-ZIF- gis

Cu-ZIF-gis was synthesized according to the previous literature method, a solvothermal reaction between Cu(NO₃)₂·3H₂O and 2-nitroimidazole (nImH) in N,N-dimethylformamide (DMF) (see experimental section).³¹ The powder X-ray Diffraction (XRD), FT-IR, and thermal gravimetric analysis (TGA) data of Cu-ZIF-gis, matched well with the previous report (Figs. 1a, b and Supplementary Figs. 1–2).³¹ Cu-ZIF-gis is a remarkably stable material in air, for almost three years, which is a beneficial property for practical application (Fig. 1b). The structure consists of 2-nitroimidazolate (nIm) linkers and highly-flattened tetrahedral copper(II) nodes, resulting in narrow cylindrical channels with a cross-sectional diameter of 2.32 ~ 2.44 Å. Considering the kinetic diameter of H₂ (2.89 Å) and incorporating nitroimidazole linkers, it may possess a great structural deformation potential realized by H₂ stimuli-responsive ZIF.³² For instance, although the Cu-ZIF-gis has an internal void (Fig. 1), nitrogen cannot enter at 77 K due to the larger probe (N₂) size of 3.64 Å.³¹ However, H₂ can diffuse inside at 77 K with a large sorption hysteresis (Supplementary Fig. 3). Similarly, at even higher temperatures (90–210 K), oxygen sorption exhibits a growing hysteresis with increasing temperature, reaching a maximum uptake before gradually decreasing uptake with eventually disappearing the hysteresis phenomenon (Fig. 1c and Supplementary Fig. 4). The hysteresis and the complex gas adsorption behavior might originate from the temperature-dependent, adsorption-induced structural response of the Cu-ZIF-gis (Supplementary Figs. 5–7 and Supplementary Table 1). Therefore, the Cu-ZIF-gis is anticipated to efficiently separate D₂/H₂ when the pore size is optimal for KQS at a certain temperature range.

The single-component equilibrium adsorption isotherms of Cu-ZIF-gis for H₂ and D₂ were measured at temperatures ranging from 40 K to 150 K, as shown in Figs. 2a, b. D₂ and H₂ isotherms at 40 K exhibit negligible uptake, as expected. This is because the size of the channel aperture is smaller than the kinetic diameter of the hydrogen isotope (2.89 Å), which prevents gas molecules from penetrating the channel, known as pore-blocking. Generally, increasing isotherm temperature results in a decrease in gas uptake due to adsorption thermodynamics. However, at 77 K, the H₂/D₂ isotherms exhibited a noticeable uptake increase with significant desorption hysteresis (similar to O₂ isotherms). Consequently, Cu-ZIF-gis showed a clear thermal gate opening, increasing the sorption amount for both H₂ and D₂ with temperature. This is attributed to the gradual lattice expansion, which results in a larger aperture and leads to an increase in uptake for both isotopes up to 100 K. Notably, there was a significant difference in the amount of H₂ and D₂ uptake at 87 K and 100 K. Even at a temperature as high as 120 K, hysteresis loops were still observable for H₂ and D₂. This indicates that equilibration hadn't been achieved yet under measurable conditions, suggesting a substantial diffusion limitation for H₂ and D₂ under a temperature-dependent and confined 1D channel structure. In other words, the structure's thermal lattice expansion becomes more pronounced with increasing sorption temperature, resulting in increased uptake up to 100 K due to a weakened diffusion barrier for both isotopes. Please note that although the lattice expansion effects enable gate opening, the effective aperture is still not wide enough to allow the hydrogen isotope to move freely in the temperature range of 77–120 K. The maximum uptake of H₂ and D₂ are 1.7 mmol g^− 1 and 2.3 mmol g^− 1 at 100 K, respectively. The uptake of D₂ is higher than H₂ in all the isotherms, which is attributed to the quantum effect in confined Cu-ZIF-gis.³³

Temperature-dependent XRD experiments are carried out to demonstrate in more detail how the Cu-ZIF-gis structure reacts to changes in temperature. The XRD patterns are recorded as the sample is ramped from 20 K to 300 K at a rate of 3 K min^− 1 under high vacuum. As the temperature increased, the position of the major peaks (2θ ~ 15, 16, and 19.1°) shifted to a slightly but clearly lower angle, indicating a thermal expansion of the crystal lattice. These results also indirectly support the temperature-dependent LDG effects that may cause channel width changes in Cu-ZIF-gis (Fig. 2c). Indeed, this represents a departure from previously reported MOFs and ZIFs with rigid structures, which display consistent XRD patterns without clear peak shifts with temperature^34,35,36.

Hydrogen Isotope Separation Performance

To further explore the desorption behavior and separation performance of hydrogen isotopes on Cu-ZIF-gis, advanced cryogenic thermal desorption spectroscopy (AC-TDS) measurements were employed. The TDS spectra of single-component H₂ and D₂ were measured at their respective liquefaction (20 K for H₂ and 23 K for D₂) temperature and room temperature (Supplementary Figs. 8–9). Similar to single-component H₂ and D₂ isotherm studies, no desorption signal was observed when gas was dosed at 20 K or 23 K, while the clear desorption curves starting from 80 K to 190 K were observed after loading at room temperature for both H₂ and D₂, with calculated amounts of 1.7 mmol g^− 1 and 2.2 mmol g^− 1, respectively. The desorption signal for pure isotope exposure at the liquefaction temperature was not observed because the structure of Cu-ZIF-gis was in a closed gate state, blocking hydrogen isotopes from entering the pore. When hydrogen isotopes are exposed to Cu-ZIF-gis at room temperature and then cooled to 18 K, the structure locks into a closed gate state after absorbing the isotopes. This means that any isotopes already inside the Cu-ZIF-gis from the room-temperature exposure are trapped inside. By that, desorption of trapped isotopes requires high thermal energy, extending up to 180 K for complete release. Notably, the desorption peak maxima of ca. 130 K is the highest desorption temperature among MOFs without strong binding sites (open metal sites) reported so far. As a result, the temperature-dependent LDG effect makes the pores accessible to isotopes at high temperatures.

Pure H₂ and D₂ TDS patterns of Cu-ZIF-gis were compared with various porous materials to gain insights into the desorption behavior of hydrogen isotope at high temperatures, as shown in Fig. 3a. Five representative samples are compared regarding the desorption curve, which is correlating sorption enthalpy and pore size. MOF-303, with an aperture size of 6 Å, exhibits a TDS spectrum that undergoes complete desorption below 60 K, indicating typical physical adsorption behavior (desorption energy ~ 5.61 kJ mol^− 1).³⁷ However, MOF-74(Ni) with open metal sites as strong adsorption sites shows the desorption of hydrogen isotopes even above 90 K (heat of adsorption ~ 13 kJ mol^− 1).^8,38 Partially fluorinated FMOF-Cu has three pores of different sizes, which are connected by a narrow bottleneck aperture.²⁹ This aperture allows H₂ access to the hidden third cavity through linker vibration. Due to this unique FMOF-Cu structure, hydrogen isotopes could adsorb till 120 K. As another similar example, the metal-organic cage (MOC) material reported by He et. al. consists of a flexible narrow window of 3.0 Å or less and organic macrocycles.³⁰ This material also exhibits a partial gate-opening effect, resulting in maximal desorption occurring at 100 K even in the absence of an open metal site. Still, a desorption signal starts already to be observed at a very low temperature of 30 K too, which is caused by an interparticle spacing. Finally, it has been observed that in the Cu-ZIF-gis, hydrogen isotopes are desorbed completely at a temperature of 180 K, one of the highest compared to other materials reported previously. This observation highlights the importance of diffusion limitations caused by long and narrow cylindrical 1D channels, which enable the release of hydrogen isotopes at elevated temperatures, even in the absence of strong binding sites.

Furthermore, desorption energy (E_d) was calculated using the Falconer and Madix equation, assuming that at peak temperature, the desorption peak maxima correspond to the desorption rate, and fractional surface coverage at desorption peak maxima is not a function of heating rate.³⁹ For E_d calculation, pure H₂ and D₂ TDS measured at the heating rate of 3, 4.5, and 6 K min^− 1 were used. Please note that the desorption energy in porous materials is the sum of binding energy and diffusion barrier energy. In most cases, the diffusion barrier energy can be negligible for systems with large pores. However, for porous materials with highly confined systems like Cu-ZIF-gis, the diffusion barrier energy difference caused by heating ramp rates may become significant and affect the overall desorption energy due to very strong diffusion limitation. The E_d for H₂ and D₂ were almost identical values of 30.3 kJ mol^− 1 and 29.7 kJ mol^− 1, respectively. Indeed, the surface binding energy (physisorption) of D₂ on Cu-ZIF-gis is typically higher than H₂ due to the quantum statistical mass effect,³³ but the diffusion energy barrier difference by heating ramp rates of H₂ is higher than D₂ due to the temperature-dependent gating effect in Cu-ZIF-gis during desorption (Supplementary Fig. 10).

The hydrogen isotope separation performance of Cu-ZIF-gis was investigated through TDS measurements for an equimolar H₂/D₂ mixture (Supplementary Fig. 8). TDS spectra and their selectivities (S_D2/H2) obtained at various exposure times (t_exp ~ 0.5, 60, and 120 min) at 1000 mbar (exposure pressure, P_exp) were shown in Supplementary Figs. 11, 12. The isotope uptake increased with exposure time but became mostly saturated after 60 min. These results imply almost equilibrium of isotope uptake at 60 min. Moreover, no significant change in SD2/H2 was observed at all temperatures, even after 120 min exposure. Figure 3b and Supplementary Fig. 13 show TDS spectra of 1:1 H₂/D₂ mixture under different loading pressures (P_exp ~ up to 2000 mbar) and exposure temperatures (T_exp = 77–120 K) after 60 min of exposure (t_exp). Note that the TDS mixture spectra have been re-plotted separately for each isotope (Fig. 3b, top D₂, and bottom H₂). TDS analysis showed that the uptake of isotope increased up to 100 K and revealed the existence of (at least) two major desorption peaks. As Cu-ZIF-gis has only one channel with no open metal site incorporated, the existence of two desorption peaks may imply two different desorbed phases caused by the thermal gating effect (Supplementary Fig. 14). Moreover, in the TDS spectrum, complete desorption was achieved at around 140 K at T_exp ~ 77 K, but the desorption extended up to 180 K at T_exp ~ 120 K. These results indicate that as the temperature increases, gas molecules can diffuse deeply into the internal channel, requiring higher temperatures to release internally trapped isotopes.

Figure 3c shows the D₂/H₂ selectivity (S_D2/H2) corresponding to the loading pressure (10-2000 mbar) and exposure temperature (77–120 K) at 60 min of exposure time. The S_D2/H2 at T_exp ~ 77 K and P_exp ~ 10 mbar was found to be 4.0. As the loading pressure increased, the S_D2/H2 decreased gradually and saturated to ca. 3.2 after 1000 mbar. At T_exp ~ 87 K, the maximum S_D2/H2 of 2.8 was achieved at P_exp ~ 10 mbar. With a further increase in the temperature, the S_D2/H2 decreases. The decrease in S_D2/H2 at higher temperatures can be ascribed to the progressive expansion of the lattice framework, resulting in getting larger channels and increased accessibility of both hydrogen isotopes. At 87, 100, and 120 K exposure temperatures, the S_D2/H2 remains constant or increases/decreases negligibly. The comparison of S_D2/H2 at 10 and 2000 mbar and various temperatures with respect to corresponding D₂ uptake is shown in Fig. 3d. It can be seen that the maximum S_D2/H2 of 4.0 at T_exp ~ 77 K and 2000 mbar was obtained despite the lowest D₂ uptake. Based on the sorption isotherms, the LDG effect starts to be observed from 77 K, forming narrow pores for hydrogen isotopes in Cu-ZIF-gis. Therefore, the molecular confinement caused by the gating effect at 77 K maximizes, resulting in a high S_D2/H2 despite low isotope uptake. However, with increasing exposure temperature until 100 K, the signal intensity for both D₂ and H₂ increased, resulting in higher total isotope uptake, but due to the actively occurring gating effect caused by lattice expansion, accessibility of the pores to hydrogen isotopes is enhanced. As a result, the uptake of hydrogen isotope increased, leading to a decrease in S_D2/H2. Nevertheless, S_D2/H2 of 1.7 was still obtained even at a high temperature of 120 K and 2000 mbar, indicating that hydrogen isotope separation is possible at high temperatures and high pressures. These observations suggest that utilizing Cu-ZIF-gis can make the adsorptive D₂/H₂ separation process more energy efficient and benefit industrial processes requiring high pressure, such as the PSA process.

As shown in Fig. 3e, the S_D2/H2 of Cu-ZIF-gis was compared with the previously reported S_D2/H2 of various MOFs exploiting the ‘KQS’ effect (We intentionally excluded the CAQS materials due to their stability issues). Generally, the separation efficiency through KQS in most MOFs has been limited below 77 K, Supplementary Table 2. This is because, at high temperatures (above 77 K), hydrogen isotopes cannot attach to MOFs in an adsorbed state due to low binding energy. In the case of Cu-ZIF-gis, hydrogen isotopes are desorbed completely at a temperature of 180 K; therefore, the separation of D₂ and H₂ can be attempted at 120 K, which is one of the highest operating temperatures reported. As a result, Cu-ZIF-gis could be a promising adsorbent for increasing the operating temperature even above the LNG liquefaction temperature of 111 K of the KQS effect (The operating temperature of 111 K or higher means that LNG cryo-infrastructure already exists and can be used for isotope separation immediately).

Microscopic observation of diffusion dynamics

Quasi-Elastic Neutron Scattering (QENS) analysis further elucidates the role of thermal lattice expansion in Cu-ZIF-gis for hydrogen isotope separation, providing deeper insights into H₂ and D₂ motion in Cu-ZIF-gis.^41,42,43 Our QENS analysis reveals the temperature-dependent mobility of adsorbed hydrogen isotopes within the Cu-ZIF-gis framework. The fitting process for QENS data was performed using the QCLIMAX package in the Integrated Computational Environment-Modeling & Analysis for Neutrons(ICE-MAN).⁴⁴ Despite the experimental data's asymmetry (originating from the instrument), the obtained QENS spectra are well-fitted with delta, Lorentzian functions, and constant background (Figs. 4a, b & Supplementary Figs. 15–30). The delta Function (red line) represents the elastic scattering component, indicating the presence of immobile or very slow-moving isotopes within the pores of Cu-ZIF-gis. The broad Lorentzian (blue line) component accounts for the quasi-elastic scattering, indicating isotope molecules' diffusive motion. The width of the Lorentzian peak provides information about the diffusion coefficients and the nature of the molecular motion. Constant background (yellow line) accounts for any residual background noise of the measured data. We have analyzed the Elastic Incoherent Structure factor (EISF), which is the contribution of elastic scattering to the total scattering intensity (elastic + quasielastic) and provides details about localized hydrogen isotope motion geometry.⁴⁵ (Eq. (1))

$$\:EISF\left(Q\right)\:=\:\frac{{I}_{elastic}\left(Q\right)}{{I}_{elastic}\left(Q\right)+{I}_{Quasielastic}\left(Q\right)}$$

The geometry of the confined molecular motion of hydrogen isotopes can be determined by fitting the Q-dependence of the EISF. Therefore, the diffusion in a sphere model derived by Volino and Dianoux has been employed to describe the localized motions of isotopes within the confined space.⁴⁶

$$\:EISF\left(Q\right)=A+\left(1-A\right){\left(\frac{3{j}_{1}\left(Qr\right)}{Qr}\right)}^{2}$$

where A is the fraction of immobile molecules, while (1-A) is the fraction of mobile molecules that undergo localized motions within a sphere of radius r, and j₁ is the spherical Bessel function of the first order. As shown in Figs. 4c, d, the fitting of the EISF for H₂ and D₂ at various temperatures were obtained using Eq. 2. The diffusion in a sphere model analysis shows a greater confinement effect for D₂ compared to H₂ within the Cu-ZIF-gis structure, suggesting differential mobility that facilitates effective isotope separation. This is particularly evident at low loading (0.7 mmol g^− 1), where the localization radius for H₂ (1.6–1.7 Å) was also larger than that for D₂ (1.7–1.9 Å) (Fig. 4e) across all exposure temperatures, indicating that H₂ undergoes diffusive motion within these confined pores, whereas D₂ appears to be more confined and immobilized (see mobile fraction of (1-A), Fig. 4f). This means D₂ is more likely to remain within the porous framework than H₂, which also aligns with previously reported results.⁴⁷ Hence, this subtle difference in single-component isotopes of r and A values in Figs. 4e, f implies that the temperature-dependent mobility of D₂ is more pronounced than H₂, potentially affecting separation selectivity under the mixture.

Table 1
Fraction of mobile molecules and localization radius at various temperatures in Cu-ZIF-*gis* determined by hydrogen isotope diffusion model Eq. (2).
Temperature [k]	Mobile Fraction [%]		Localization of radius, r [Å]
Temperature [k]	H₂	D₂	H₂	D₂
77	0.20 ± 0.03	0.13 ± 0.02	1.68 ± 0.14	1.61 ± 0.09
100	0.28 ± 0.06	0.21 ± 0.04	1.70 ± 0.15	1.67 ± 0.13
150	0.43 ± 0.10	0.39 ± 0.07	1.91 ± 0.18	1.73 ± 0.15
* Please note that the immobile fraction also contains the Cu-ZIF-gis itself

Our study concludes by demonstrating the novel mechanism of LDG and highlighting the exceptional performance and robustness of Cu-ZIF-gis, a Cu-based ZIF, for hydrogen isotope separation. We particularly focused on its efficiency at elevated temperatures and pressures. Our advanced cryogenic thermal desorption spectroscopy (TDS) measurements reveal that Cu-ZIF-gis maintains significant D₂/H₂ selectivity even above the LNG temperature of 111 K. This performance based on the KQS is significantly higher than the operational temperatures for traditional methods (or any other KQS materials), attributed to the material's LDG effect, which dynamically modulates pore sizes to optimize isotope separation. The temperature-triggered LDG in Cu-ZIF-gis enhances its ability to adsorb and separate hydrogen isotopes at elevated temperatures at 120 K, allowing for efficient separation of D₂ from H₂ under high-pressure conditions, thereby eliminating the need for low-temperature operations. Note that its operating temperature above 111 K makes it a viable solution for integration into existing LNG cryo-infrastructure, facilitating large-scale industrial applications. The material also exhibits excellent structural stability, retaining its framework integrity even after prolonged (for 34 months) exposure to air and multiple adsorption-desorption cycles, ensuring long-term performance and reliability in industrial settings. The findings are supported by complementary QENS experiments. These experiments reveal that there is a dynamic difference that allows for the effective separation of D₂ from H₂, even at high temperatures of up to 150 K. This highlights that Cu-ZIF-gis offers a groundbreaking approach to hydrogen isotope separation. It leverages both the LDG effect and the distinct mobility dynamics revealed by QENS to achieve high performance under practical conditions. The Cu-ZIF-gis's high efficiency at elevated temperatures and pressures, as well as robust durability, address the limitations of conventional methods, providing a more energy-efficient and environmentally sustainable alternative.

Material Synthesis

Cu-ZIF-gis was prepared on a 5-fold scale using the literature method.³¹ Cu(NO₃)₂·3H₂O (121 mg, 0.5 mmol, Sigma-Aldrich) and 2-nitroimidazole (113 mg, 1.0 mmol, TCI) were dissolved in N,N-dimethylformamide (DMF) (15 mL, Daejung Chemicals & Metals Co., Ltd.) in a 30 mL vial. The vial was tightly capped with Teflon tape and heated in an oven at 80°C for one day to give blue square crystals. The crystals were filtered and washed with neat DMF (10 mL × 3) and diethyl ether (10 mL × 3) and dried under reduced pressure for about ten hours. The yield was 40% based on one mole of the metal salt.³¹

General Materials Characterization

X-ray diffraction patterns (XRD) experiment was carried out on the Rigaku MiniFlex 600 using Cu Kα radiation in the 2θ range of 5 ° to 40 ° with a scanning rate of 10 deg min^− 1. Fourier transform infrared (FT-IR) was obtained under transmission mode in the range of 4000 to 500 cm^− 1 using Thermo Fisher Scientific Nicolet iS50. Thermogravimetric analysis (TGA) was conducted in a temperature range from 25 ℃ to 700 ℃ with a heating rate of 10 ℃ min^− 1 under N₂ atmosphere with a flow rate of 15 cc min^− 1 using TGA-N-1000.

Gas sorption Isotherms Measurement

A H₂, D₂ and O₂ sorption experiment was conducted by a fully automated Quantachrome Autosorb-iQ2 with a home-built temperature-controlled cryocooler. Temperature calibration was performed due to the difference in position between the temperature sensor and the sample. The exact temperature was determined through the saturation pressure of various gases (H₂, D₂, O₂). Additionally, volume calibration was performed using the exact temperature obtained through temperature calibration. The existing volume calibration method using helium has a problem with helium adsorption at cryogenic temperatures. Therefore, the adsorption amount of the empty cell at each temperature was measured using a measuring gas, and then the measured value of the sample cell containing the sample was calibrated. Prior to measurement, Cu-ZIF-gis (ca. 100 mg) was activated under an ultra-high vacuum at 433 K for 10 hours to remove the moisture and gas molecules. The H₂ and D₂ sorption isotherms were obtained at 40, 77, 87, 100, 120, and 150 K, and O₂ sorption isotherms were obtained at 90, 110, 135, 160, 185, and 210 K.

In-situ X-ray Diffraction Patterns (XRD) measurement

In-situ XRD measurements were conducted to investigate the structural transition in Cu-ZIF-gis as a function of temperature. Structural studies of the Cu-ZIF-gis were performed using Rigaku MiniFlex 600 with CuKα radiation. Prior to measurement, Cu-ZIF-gis was activated at 100°C for 3 hours to remove the moisture and impurities. In-situ XRD pattern was recorded within the 14 ° to 20 ° range with a scanning speed of 3 deg min^− 1 in a temperature range of 20 K to 300 K with a heating rate of 6 K min^− 1 under vacuum conditions.

Thermal Desorption Spectroscopy (TDS) Measurement

TDS experiment was performed using an advanced cryogenic thermal adsorption spectroscopy (AC-TDS) apparatus equipped with quadrupole mass spectroscopy (QMS) and turbo molecular pump (TMP) for ultrahigh vacuum (10^− 8 mbar). The sample was loaded in a Copper sample holder and activated under vacuum at 433 K for 5 hours in order to remove water and adsorbed gas impurities if any. The procedure of TDS measurement proceeded as shown in Supplementary Fig. 8. For pure H₂ and D₂ TDS measurement, the sample was exposed to pure H₂ or D₂ at room temperature and liquefaction temperature (20 K for H₂, 23 K for D₂). Afterward, unadsorbed gas was evacuated at 20/23 K and cooled down to 18 K, and finally, the sample was heated to 240 K with a ramping rate of 6 K min^− 1. The desorbing gas was continuously recorded by quadrupole mass spectroscopy (QMS). In the H₂/D₂ 1:1 mixture measurements, the sample was exposed to H₂/D₂ 1:1 mixture at various exposure temperatures (77, 87, 100, and 120 K) for 60 min, and unadsorbed gas was evacuated at exposure temperatures. Then, the sample was cooled down to 18 K. Subsequently, with an increase in the temperature with a ramping rate of 6 K min^− 1 to 240 K, the gas molecules started desorbing from the sample and were quantified by QMS. The area under the obtained desorption peak of H₂ and D₂ was proportional to the desorbed amount of hydrogen isotopes, which can be quantified after calibration of TDS. The S_D2/H2 was directly calculated using the quantified area under desorption peak by calibration constant after calibration of TDS using Pd₉₅Ce₅ alloy and TiH₂ powder.

Quasi-Elastic Neutron Scattering (QENS) Measurements

The QENS measurement was carried out by disk chopper time-of-flight (TOF) spectrometer, IN5 at the Institut Laue-Langevin (Grenoble, France). The QENS spectra were collected using an incident neutron with a wavelength of 5 Å in a momentum transfer range of 0.3 Å⁻¹ to 1.9 Å⁻¹ with a step size of 0.2 Å⁻¹. The sample was loaded into an aluminum sample can in the glove box to prevent air exposure. Before the QENS measurement, the sample holder loaded Cu-ZIF-gis was activated at 433 K under a vacuum condition to remove the moisture and impurities. Afterward, the sample can be transferred to the IN5. During QENS measurement, H₂ and D₂ gases were injected into the sample, dosing amounts of 0.7 and 2.0 mmol g^− 1 at each temperature (77 K − 150 K) through an in-situ gas dosing system. The standard vanadium and empty can measurements were conducted for data normalization and background correction.

Competing interests

The authors declare no competing interests.

Author contributions

MJ, JP and HO conceptualized the project. MJ, JP and RM developed the methodology. CJ and JK synthesized and characterized the sample. MJ, JP and TP performed sorption & TDS experiments and analyzed the data. JY conducted in-situ XRD measurement. MJ, JP, JO, JTP, and MR performed the QENS experiment, and MJ, JP, AJR-C, JTP, MR, and HO analyzed and visualized the data. HO was responsible for funding acquisition and provided supervision throughout the project. MJ, JP, RM, JK, MR and HO wrote the original draft. MJ, JP, RM, JTP, JK, MR and HO reviewed and edited the paper.

Acknowledgments

This research was funded by the National Research Foundation of Korea (NRF), Government of Korea (MSI) (2022R1A2C3005978, RS-2023-00281671, NRF-2020R1A2C1004717). The computing resources were made available through the VirtuES and the ICE-MAN projects, funded by the Laboratory Directed Research and Development program and Compute and Data Environment for Science (CADES) at ORNL. We acknowledge the Institut Laue-Langevin (Grenoble, France) for the allocation of beam time.

Avouris P et al (1996) STM-induced H atom desorption from Si (100): isotope effects and site selectivity. Chem Phys Lett 257:148–154
Cheng K, Lee J, Lyding JW (2000) Approach to enhance deuterium incorporation for improved hot carrier reliability in metal-oxide-semiconductor devices. Appl Phys Lett 77:2358–2360
Lyding JW, Hess K, Kizilyalli IC (1996) Reduction of hot electron degradation in metal oxide semiconductor transistors by deuterium processing. Appl Phys Lett 68:2526–2528
Greenwood NN, Earnshaw A (2012) Chemistry of the Elements. Elsevier
Laboratory ORN (2005) (Oak Ridge National Laboratory Oak Ridge, TN
Bondorf L et al (2022) Isotope-selective pore opening in a flexible metal-organic framework. Sci Adv 8:eabn7035
Savchenko I et al (2015) Hydrogen isotope separation in metal-organic frameworks: Kinetic or chemical affinity quantum-sieving? Microporous Mesoporous Mater 216:133–137
Kim JY et al (2017) Exploiting diffusion barrier and chemical affinity of metal–organic frameworks for efficient hydrogen isotope separation. J Am Chem Soc 139:15135–15141
Muhammad R et al (2021) Exploiting the specific isotope-selective adsorption of metal–organic framework for hydrogen isotope separation. J Am Chem Soc 143:8232–8236
Park J, Attia NF, Jung M, Lee K, Oh H (2020) Biobased derived nanoporous carbon for hydrogen isotope separation. Microporous Mesoporous Mater 304:109291
Xiong R et al (2021) Hydrogen isotopes separation in Ag (I) exchanged ZSM-5 zeolite through strong chemical affinity quantum sieving. Microporous Mesoporous Mater 313:110820
Kim JY, Oh H, Moon HR (2019) Hydrogen isotope separation in confined nanospaces: carbons, zeolites, metal-organic frameworks, and covalent organic frameworks. Adv Mater 31:1805293
Ha J, Jung M, Park J, Oh H, Moon HR (2022) Thermodynamic Separation of Hydrogen Isotopes Using Hofmann-Type Metal-Organic Frameworks with High-Density Open Metal Sites. ACS Appl Mater Interfaces 14:30946–30951
Oh H et al (2013) A cryogenically flexible covalent organic framework for efficient hydrogen isotope separation by quantum sieving. Angew Chem Int Ed 52:13219–13222
Weinrauch I et al (2017) Capture of heavy hydrogen isotopes in a metal-organic framework with active Cu (I) sites. Nat commun 8:14496
Zhang L et al (2022) Chemical Affinity of Ag-Exchanged Zeolites for Efficient Hydrogen Isotope Separation. Inorg Chem 61:9413–9420
Oh H, Savchenko I, Mavrandonakis A, Heine T, Hirscher M (2014) Highly effective hydrogen isotope separation in nanoporous metal-organic frameworks with open metal sites: direct measurement and theoretical analysis. ACS Nano 8:761–770
Xiong R et al (2020) Highly effective hydrogen isotope separation through dihydrogen bond on Cu (I)-exchanged zeolites well above liquid nitrogen temperature. Chem Eng J 391:123485
Katugampalage TR et al (2024) Bimetallic Fe: Co metal-organic framework (MOF) with unsaturated metal sites for efficient Fenton-like catalytic degradation of oxytetracycline (OTC) antibiotics. Chem Eng J 479:147592
Kim JY et al (2017) Selective hydrogen isotope separation via breathing transition in MIL-53 (Al). J Am Chem Soc 139:17743–17746
Teufel JS (2012) Experimental investigation of H₂/D₂ isotope separation by cryo-adsorption in metal-organic frameworks
Aljammal N, Jabbour C, Chaemchuen S, Juzsakova T, Verpoort F (2019) Flexibility in metal-organic frameworks: A basic understanding. Catalysts 9:512
Chen K et al (2022) Gating effect for gas adsorption in microporous materials—mechanisms and applications. Chem Soc Rev 51:1139–1166
Schneemann A et al (2014) Flexible metal-organic frameworks. Chem Soc Rev 43:6062–6096
Kotani R, Kondo A, Maeda K (2012) Gate adsorption of CO₂ on a flexible one-dimensional copper-based coordination polymer crystal. ChemComm 48:11316–11318
Zhao D, Yuan D, Krishna R, van Baten JM, Zhou H-C (2010) Thermosensitive gating effect and selective gas adsorption in a porous coordination nanocage. ChemComm 46:7352–7354
Ma S, Sun D, Wang XS, Zhou HC (2007) A mesh-adjustable molecular sieve for general use in gas separation. Angew Chem Int Ed 46:2458–2462
Kim H et al (2008) Temperature-triggered gate opening for gas adsorption in microporous manganese formate. ChemComm 39:4697–4699
Zhang L et al (2019) Exploiting dynamic opening of apertures in a partially fluorinated MOF for enhancing H2 desorption temperature and isotope separation. J Am Chem Soc 141:19850–19858
He D et al (2022) Hydrogen isotope separation using a metal–organic cage built from macrocycles. Angew Chem Int Ed 61:e202202450
Jung C et al (2023) Porous zeolitic imidazolate frameworks assembled with highly-flattened tetrahedral copper (ii) centres and 2-nitroimidazolates. ChemComm 59:4040–4043
Nam J et al (2023) Zeolitic Imidazolate Frameworks as Solid-State Nanomachines. Angew Chem Int Ed, e202404061
Chen B et al (2008) Surface interactions and quantum kinetic molecular sieving for H₂ and D₂ adsorption on a mixed metal-organic framework material. J Am Chem Soc 130:6411–6423
Javed A, Strauss I, Bunzen H, Caro J, Tiemann M (2020) Humidity-mediated anisotropic proton conductivity through the 1D channels of Co-MOF-74. Nanomaterials 10:1263
Li J-N et al (2024) Host–Host Interactions Enhanced the Structural Rigidity in a 6-Fold Interpenetrated Diamondoid Metal-Organic Framework Exhibiting C₂H₂/C₂H₄ Separation. Inorg Chem 63:8286–8293
Miyazaki I, Masuoka Y, Suzumura A, Moribe S, Umehara M (2022) Direct Sintering Behavior of Metal Organic Frameworks/Coordination Polymers. ACS omega 7:47906–47911
Kim H et al (2023) High D₂/H₂ selectivity performance in MOF-303 under ambient pressure for potential industrial applications. Sep Purif Technol 325:124660
Pham T et al (2015) Understanding the H2 sorption trends in the M-MOF-74 series (M = Mg, Ni, Co, Zn). J Phys Chem C 119:1078–1090
Falconer JL, Schwarz JA (1983) Temperature-programmed desorption and reaction: applications to supported catalysts. Catal Rev Sci Eng 25:141–227
Mondal SS et al (2019) Systematic Experimental Study on Quantum Sieving of Hydrogen Isotopes in Metal-Amide‐Imidazolate Frameworks with narrow 1‐D Channels. ChemPhysChem 20:1311–1315
NaraseGowda S, Brown CM, Tyagi M, Jenkins T, Dobbins TA (2016) Quasi-elastic neutron scattering studies of hydrogen dynamics for nanoconfined NaAlH₄. J Phys Chem C 120:14863–14873
Jung M et al (2021) Elucidation of Diffusivity of Hydrogen Isotopes in Flexible MOFs by Quasi-Elastic Neutron Scattering. Adv Mater 33:2007412
Pefoute E, Kemner E, Soetens J, Russina M, Desmedt A (2012) Diffusive motions of molecular hydrogen confined in THF clathrate hydrate. J Phys Chem C 116:16823–16829
Ramirez-Cuesta AJ, Cheng Y (2021) The ICEMAN, a heterogeneous platform for analysis of neutron scattering data. Neutron News 32:15–16
Bée M (1992) A physical insight into the elastic incoherent structure factor. Phys B: Condens Matter 182:323–336
Volino F, Dianoux A (1980) Neutron incoherent scattering law for diffusion in a potential of spherical symmetry: general formalism and application to diffusion inside a sphere. Mol Phys 41:271–279
Yang D et al (2024) Investigation of the Dynamic Behaviour of H₂ and D₂ in a Kinetic Quantum Sieving System. ACS Appl Mater Interfaces 16:12467–12478

There is NO Competing Interest.

fSI.docx
GA.png
Graphical abstract

Download PDF

Version 1

posted

You are reading this latest preprint version

Exploiting a Lattice-Driven Gating in a Cu-Based Zeolitic Imidazolate Framework for Efficient High-Temperature Hydrogen Isotope Separation

Status:

Version 1

Abstract

Figures

Introduction

Results

Hydrogen Isotope Separation Performance

Microscopic observation of diffusion dynamics

Conclusions

Methods

Material Synthesis

General Materials Characterization

Gas sorption Isotherms Measurement

In-situ X-ray Diffraction Patterns (XRD) measurement

Thermal Desorption Spectroscopy (TDS) Measurement

Quasi-Elastic Neutron Scattering (QENS) Measurements

Declarations

Competing interests

Author contributions

Acknowledgments

References

Additional Declarations

Supplementary Files

Status:

Version 1