Capture and immobilisation of iodine (I₂) utilising polymer-based ZIF-8 nanocomposite membranes

Abstract

Polymer nanocomposites made up of nanoporous metal–organic frameworks (MOFs) are fast becoming a staple of next generation hybrid composites, and are currently being intensely developed for gas capture and separation. This work reports the first attempt to capture and retain iodine (I₂) using polymer-MOF (ZIF-8) nanocomposites. Membranes of ZIF-8-based nanocomposites (comprising either a glassy Matrimid or a rubbery polyurethane (PU) matrix) were prepared via a colloidal-mixing approach and their viability for I₂ capture and retention effects was determined through absorption experiments, nanoindentation mechanical measurements, and thermogravimetric (TGA) analysis. The absorption experiments demonstrated that I₂ capture and retention is possible in all of the nanocomposite membranes, although the PU/ZIF-8 30 wt% nanocomposite exhibited higher affinity for I₂ absorption (>32 wt%). It is reasoned that the molecular affinity and attraction between I₂, (2-methylimidazolate organic ligands of) ZIF-8 nanoparticles, and polymer matrices (Matrimid and PU) will catalyse the formation of weak secondary bonds, resulting in the ‘capture’ and ‘retention’ of I₂ within molecular segments of the polymers and inside the pores of ZIF-8. The enhancement of the Young's modulus (E) of the PU/ZIF-8 30 wt% nanocomposite (E increased by ∼6%) is postulated to be due to I₂ rigidification, while TGA analysis proved that I₂ retention within both free volume of the polymer and ZIF-8 sodalite cages remained intact up till the points of structural degradation, at ∼200 °C for the PU-based nanocomposites, and at ∼300 °C for the Matrimid-based nanocomposites. We propose that the affinity of the organic ligands in ZIF-8 and the formation of free volume in the nanocomposites from the presence of ZIF-8 attracted I₂, and the formation of secondary bonds between these constituents (H-bonds) strengthened not only the nanocomposite, but also kept I₂ from being released despite the larger pore size and gate-opening dynamics of ZIF-8. It was therefore concluded that a combination of nanoparticles of porous MOFs and a rubbery polymer is promising for further development to enable I₂ capture and retention applications.

---

1. Introduction

Nuclear technology is prominent in two sectors that are intrinsic parts of our lives: power generation¹ and medicine.² Nuclear power plants (NPP) are regarded as a source of clean, renewable, and cost-effective energy,³ while nuclear medicine forms the core of many diagnostic techniques and treatments for illnesses plaguing mankind.⁴ Nuclear-related processes produce many by-products such as isotopes of iodine (¹²⁹I and ¹³¹I), ¹³⁵Cs and ⁹⁹Tc, some of which emit alpha (α), beta (β), gamma (γ), and neutron (n), all of which are hazardous to humans and the environment. Due to their extensive half-lives (t_1/2), some ranging to millions and billions of years, a permanent and effective disposal approach for these by-products, aptly called nuclear waste products, is necessary.^5,6

Some common radionuclides that are produced as by-products of power generation are isotopes of iodine: ¹²⁹I, with a t_1/2 of 1.5 × 10⁷ years, which is detrimental to the environment, and ¹³¹I, with a t_1/2 of 8.02 days, which is detrimental to human metabolic processes.⁷ As such, many research works have focused on the successful capture and confinement of iodine.^8,9 Common industrial practice in capturing and storing iodine is dominated by activated carbon^10,11 and zeolite¹² filters and scrubbers; however, recent research works are oriented towards metal–organic frameworks (MOFs)^13,14 or variants of composites,^9,15,16 to varying degrees of success.

MOFs are an emerging class of hybrid materials^17–19 formed by coordinating multifunctional organic linkers to metallic ions or clusters, giving rise to a plethora of 3-D porous networks containing periodic nanosized pores or extended metal–organic channels.^20–22 The presence of metallic cores and organic linkers introduces tunable physico-chemical properties that are absent from purely metallic or purely organic materials.²³ Many MOFs are extremely porous (typically >1000 m² g⁻¹)²⁴ to accommodate guest encapsulation or host–guest confinement interactions,²⁵ while certain MOF structures feature coordinatively unsaturated metal sites²⁶ that are highly accessible for molecular adsorption. From the mechanical standpoint, the open frameworks of MOFs are relatively more flexible (lower stiffness)^27,28 and also softer compared to their constituents or inorganic structural analogues.²⁹ A notable example of MOFs that embody these properties is the subclass called Zeolitic Imidazolate Frameworks (ZIFs),³⁰ which are structural analogues of inorganic zeolites, established catalysts in the petrochemical industry. Ag-impregnated zeolites (Ag-MORs)³¹ have been used to capture iodine, and filters to that effect are available commercially. ZIFs exhibit a tunable accessible surface area³⁰ and a respectably high thermal stability (up to 500 °C),³² making them ideal candidates for many potential applications.^33–35 Recent reports reveal that the terahertz lattice dynamics and soft modes of ZIFs³⁶ may also play a role in guest sorption, retention and delivery.

ZIF-8 [Zn(mIm)₂; mIm = 2-methylimidazolate], which is a cubic ZIF material adopting a sodalite architecture (surface area up to 1950 m² g⁻¹),³² has been shown to be effective in the capture and retention of iodine in both the powdered form and extruded pellets.⁷ Despite its reported effectiveness, the deliverability of ZIF-8 remains an open question. Depositing loose powders of ZIF-8 into aqueous environments is clearly not a practical approach for industrial practices.³⁷ Therefore, recent work has focused on the formation of MOFs, and by extension, ZIF-based composites³⁸ and membrane systems^39,40 to improve robustness and enhance deliverability. There are plenty of examples of MOFs and ZIF-based composites,^41,42 encompassing polymer-based and ceramic-based matrices. However, MOF-polymer composites are especially favoured, due to the affinity between the organic ligands of MOFs and the polymer molecules/chains, which results in the formation of strong and continuous structures that retained all of the advantages of MOFs without compromising the structural features associated with polymers. The flexibility and ease of processing of polymers also make them easily deployable and impregnated with MOF nanoparticles at controlled loadings.^43,44 It should also be pointed out that recently reported work suggested that the deployment of composites,^15,16 when exploiting the advantages of both the matrix and fillers, is more effective in capturing and storing iodine compared to either constituents acting on their own.

Previous attempts have been made to capture and immobilise iodine within ZIF-8/HKUST-1, Ag, and glass powders (EG2992 and EG2998), forming a glass-composite material (GCM), which turned out to be a success, as reported by Sava et al.⁴⁵ Despite the effectiveness of this reported approach, its energy intensiveness and the requirement to employ a mixture of multiple constituents within a single system might prove to be over demanding for practical engineering solutions.

The current study aims to combine the ease of material processability and adaptability of polymers and the large surface area and regular pores offered by MOFs to create a composite membrane system, which is structurally resilient, easily deliverable, and mechanically robust to capture and retain iodine. This work aims to develop, assess and characterise the viability of ZIF-8/polymer nanocomposites as an iodine capture and retention model system. Herein, one of the objectives is to also minimise the energy expenditure associated with the capture and retention process, and thus to rely upon the molecular interactions and affinity between iodine, ZIF-8 nanoparticles, and polymer matrices (either glassy or rubbery) to expedite the uptake of iodine into the thin composite membranes. The introduction of this so-called passive polymer-based system represents an equally effective alternative that is cost effective and highly deliverable, the concept of which is potentially viable for development to enable industrial applications in the long run.

2. Materials and methods

2.1 Synthesis of ZIF-8 nanoparticles and processing of ZIF-8/Matrimid and PU/ZIF-8 nanocomposite membranes

The methods involved in synthesising ZIF-8 nanoparticles (Fig. 1) are reported elsewhere⁴⁶ and are replicated in this work. Polyurethane (PU) beads were purchased from Sigma Aldrich, while Matrimid®5218 flakes were supplied by Huntsman Materials (the polymeric structure for the polymers and ZIF-8 are presented in the ESI† in Fig. S2). Both polymers were used as is without any further alterations. A measured amount of PU, based on eqn (1), was dissolved in tetrahydrofuran (THF), while the same was done for Matrimid flakes in chloroform (CHCl₃), so as to achieve the targeted ZIF-8 nanoparticle loading of 30 wt%.where m_ZIF-8 is the weight (in grams) of ZIF-8 mixed into the colloidal solution, and m_PU is the weight (in grams) of the PU beads mixed into the composite solution. m_PU can be replaced with m_Matrimid in the case of Matrimid flakes.

Fig. 1 Scanning electron micrograph showing ZIF-8 nanoparticles prior to being embedded in polymer-MOF nanocomposite. The mean diameter of the nanoparticles is ca. 150 nm.

Both were magnetically stirred for 24 hours, forming a clear and transparent viscous solution for PU, and a clear yellowish viscous solution for Matrimid. Two vials of ZIF-8 nanoparticles were prepared (∼2 g each), dispersed in CHCl₃ and THF, and were added to the Matrimid and PU solutions at a measured rate, respectively. The combined PU/ZIF-8 and ZIF-8/Matrimid solutions were again magnetically stirred for 24 hours, forming opaque and viscous solutions. The PU/ZIF-8 solution was casted onto a PTFE substrate, while the ZIF-8/Matrimid solution was casted onto a glass substrate using an automated doctor blade machine at a constant speed of 15 mm⁻¹, with the thickness of the doctor blade set to 150 μm. After casting, the membranes were left to slowly cure in a glove bag saturated with THF vapour (PU/ZIF-8) and CHCl₃ vapour (ZIF-8/Matrimid) for 6 hours, after which they were removed and placed in a vacuum oven and dried for 24 hours at 85 °C to remove any residual THF/CHCl₃ within the PU, Matrimid, or ZIF-8 nanoparticles. After drying, the membranes were removed from the vacuum oven and carefully dislodged from the substrate, had their thicknesses measured, and stored for further tests. The methodology described above resembles that of our recent report to fabricate thin mixed-matrix membranes of MOF/polymer nanocomposites.⁴⁴

2.2 Iodine (I₂) capture

A solution of 10 mM iodine (henceforth called I₂) was prepared with cyclohexane (C₆H₁₂) as its base; both were purchased from Sigma Aldrich and used as is without further modifications or purification. The previously synthesised PU/ZIF-8 and ZIF-8/Matrimid nanocomposites were dried in a vacuum oven for 24 hours at 85 °C prior to the I₂ uptake experiments to remove any moisture or solvents remaining on the surface of or within the membranes. These nanocomposites were then carefully weighed, with the weight recorded for every sample. The nanocomposites were then placed in a 50 mL glass vial, and 1 mL of the aforementioned I₂ solution was sampled and placed into the vials and hermetically sealed. The vials were placed away from direct exposure to sunlight in order to prevent UV degradation or from influencing I₂ capture. Images of the vials were taken at designated intervals of 0, 1, 2, 4, 6, 24, 48, 72 and 96 hours. At the end of the uptake experiments, the membranes were removed from the solution, washed, and weighed.

The samples exhibiting the highest level of capture and retention of iodine were used for desorption experiments, where the sample was submerged in ethanol (EtOH) and allowed to desorb for a set amount of time. Images of the desorption experiments were taken to monitor the level of iodine released from the samples. The I₂-absorbing polymers and nanocomposites were subjected to multiple characterisation techniques to confirm their absorptive and retentive capabilities.

2.3 Scanning electron microscopy (SEM)

Matrimid, PU, and their corresponding ZIF-8 nanocomposites, as is, were submerged in liquid nitrogen (LN₂) and then fractured at the cross sections to produce a clear view of their through-thickness internal morphology. The fractured samples were mounted on an SEM stub and secured using carbon tapes at an angle of 90°. The samples were then sputter-coated with gold for 30 seconds, and were then loaded into the field emission gun SEM (TESCAN LYRA3 FEG-SEM/FIB apparatus) for imaging. Imaging was conducted under high vacuum (HV mode), at 5–10 kV, employing a magnification of 10 000–30 000×.

2.4 Thermogravimetric analysis (TGA)

The influence of I₂ uptake on the thermal and structural modifications of the pristine nanocomposite membranes was studied using the TGA method. The device used in this work is a TGA-Q50 (TA instruments), which comes with its own induction-heating chamber and platinum (Pt) sample holder. A portion of the nanocomposites was carefully sectioned from the main sample (∼1.5 mg), and placed onto the Pt sample holder. Prior to the experiments, the sample holder was calibrated to zero weight in order to prevent any foreign particles or objects from influencing the readings. The samples were then subjected to the heat treatment outlined and programmed into the software: a heating rate of 10 °C min⁻¹, an initial temperature of 50 °C and a final temperature of 800 °C. These parameters and temperatures were selected in order to guarantee that the samples (which include Matrimid, PU, and ZIF-8) undergo the total range of thermal phase changes and decomposition. The weight percent (wt%) variation of the samples was recorded at every temperature interval, and these data were then used to construct thermal decomposition plots of the polymer matrices and their corresponding nanocomposites, with and without I₂ guest molecules.

2.5 Nanoindentation mechanical characterisation

The blank and I₂-loaded nanocomposites were secured onto an aluminium stub using a mounting wax, and their quasi-static nanomechanical properties, specifically the corresponding Young's modulus (E) and nanohardness (H), were determined using the nanoindentation method via an MTS NanoIndenter XP (Agilent Technologies, USA), equipped with a Berkovich three-sided pyramid diamond tip. A collection of 20 indents (at a depth of 2000 nm) were made per sample at random locations on the membrane surface, as per the method described by Mahdi et al.⁴⁴

3. Results and discussion

3.1 Iodine uptake, immobilisation and release rates

Fig. 2 shows the cross-sectional images of the membranes. It is clear from the SEM images that the dispersion of the ZIF-8 nanoparticles throughout the polymer matrices is uniform. It should also be pointed out that the ZIF-8 nanoparticles embedded in Matrimid are encapsulated by Matrimid, exhibiting spherically coated surfaces,⁴⁴ while the ZIF-8 nanoparticles embedded onto PU seem to retain their original rhombic dodecahedron crystal configuration,⁴⁷ further increasing the accessibility to the ZIF-8 within the nanocomposite, which is important vis-à-vis I₂ capture and retention. The presence of ZIF-8 was also confirmed via X-ray diffraction performed on the membranes (see XRD in ESI,† Fig. S1).

Fig. 2 Cross-sectional SEM images of the polymer and polymer-MOF nanocomposites, with (a) Matrimid, (b) ZIF-8/Matrimid 30 wt%, (c) PU, and (d) PU/ZIF-8 30 wt%.

Fig. 3 shows the vials containing the Matrimid and PU-based ZIF-8 membranes, filled with I₂, at designated time intervals (0–96 h), while Fig. 4 shows the quantitative uptake capacity of the membranes, as measured by the wt% of absorbed I₂.

Fig. 3 Determination of I₂ absorption characteristics of the polymer-MOF nanocomposites. (a) The time it took for the composites to absorb I₂ up to 96 hours of exposure, where (1) Matrimid, (2) Matrimid/ZIF-8 30 wt%, (3) PU, and (4) PU/ZIF-8 30 wt%. (b) Membranes before absorption of I₂ (top) and after absorption of I₂ (bottom). (c) Optical micrographs taken at the cross section of the PU/ZIF-8 30 wt% membranes, where the dotted arrows indicate the I₂ flow gradient entering the composite. The darker region corresponds to the top membrane surface while the lighter region is the bottom surface (with respect to (a)). It can be seen that the mean membrane thickness is 100 μm.

Fig. 4 Weight differential of I₂ uptake in the polymer and 30 wt% nanocomposite samples. It was found that the weight uptake of I₂ in Matrimid and its composite is relatively small compared to that in PU and its corresponding nanocomposites.

The physical changes in the PU and PU-based nanocomposite are rather pronounced, with PU exhibiting a yellowish tinge, while the PU/ZIF-8 30 wt% membrane show a dark purple tinge (Fig. 3b). The colour changes are attributed to the absorbed I₂. Data in Fig. 4 shows that the PU/ZIF-8 30 wt% nanocomposite exhibiting a substantial weight gain of 32.4 wt%, while PU shows a much smaller increase of 5.9 wt%. It can be seen that higher concentrations of absorbed I₂ will result in the nanocomposites exhibiting a similar colour to that of I₂, while lower absorbed concentrations will result in a light brownish tinge. Due to the relatively small amount of absorbed I₂ in the Matrimid-based membranes (1.8 wt% and 2.7 wt% for Matrimid and Matrimid/ZIF-8 30 wt%, respectively), their corresponding colours post-absorption remained the same (Fig. 3b). It was evident that the uptake and retention of I₂ resulted in the change of colours of the polymers and their corresponding nanocomposites.

To confirm I₂ capture within the nanocomposite as opposed to mere surface coverage of I₂, we visually assessed the cross-section of the I₂-loaded composite (Fig. 3c). Optical images taken of the cross-section of the nanocomposite capturing the highest amount of I₂, i.e. PU/ZIF-8 30 wt%, exhibited features suggestive of I₂ transport across the nanocomposite interior. The colour gradient in Fig. 3c indicates the direction of the flow of I₂, proving that I₂ was indeed absorbed by the nanocomposite as opposed to only being present on the surface layer of the composite. Notably, the top and bottom surfaces of the nanocomposite could be easily distinguished. It can be seen that the top surface that was exposed to I₂ exhibited a darker shade of purple, as opposed to the bottom surface of the membrane sample. This differential colouration phenomenon elucidates that a concentration gradient is present in the composite, thus indicating that I₂ has infiltrated the nanocomposite but remained entrapped within as opposed to being filtered out of the membrane.

We examined the I₂ release characteristics by monitoring the change in colour of the solution, where the I₂-loaded PU/ZIF-8 30 wt% nanocomposite was immersed and hermetically sealed (Fig. 5). The polymer matrices, as discussed previously, decrease the accessibility to the high surface area of ZIF-8 nanoparticles, which extended the I₂ release time (beyond 6.5 h), relative to the case of (loose) MOF powders, where a substantial release of captured I₂ was reported after only two hours.⁴⁸ Furthermore, the hydrophobicity of the polymer matrices prevented access of polar solvents to the nanopores of the ZIF-8 nanoparticles, allowing the nanocomposites to retain I₂ within their microstructures. This was proven by immersing the I₂-loaded PU/ZIF-8 30 wt% composite into water, and after a period of 10 days (see Fig. S3 in the ESI†), there were no colour changes detected in the surrounding water, signifying that I₂ remained immobilised. Importantly, this result shows that despite the affinity of I₂ towards polar solvents (e.g. water), the desorption of I₂ can be suppressed by encapsulating it within a polymer/MOF nanocomposite where the matrix is an intrinsically hydrophobic phase. We reasoned that the complex microstructure of the nanocomposite (Fig. 2) further prevented open access to the adsorption sites located inside the sodalite cages of ZIF-8,⁷ such that the absorbed I₂ can remain immobilised within the membrane.

Fig. 5 Iodine release was investigated by monitoring the change in the colour of ethanolic solution in I₂-loaded PU/ZIF-8 30 wt% at different time points. Since the polymer matrix in the nanocomposite decreased the accessibility to the high surface area of ZIF-8 pores, I₂ release in ethanol was observed to slow down and remain virtually unchanged beyond 6.5 h (see ESI† Fig. S3 for the sample after 10 days of immersion time).

3.2 Nanoindentation properties

The nanoindentation technique was used to characterise the nanomechanical properties of the polymers and their corresponding nanocomposites pre- and post-I₂ absorption. The results are summarised in Fig. 6.

Fig. 6 Nanoindentation results of the Matrimid and PU-based nanocomposite membranes. (a & b) Young's modulus of Matrimid-based and PU-based membranes, respectively. (c & d) Nanohardness of Matrimid-based and PU-based membranes, respectively. Note that the glassy Matrimid matrix (whose T_g > room temperature, RT) is significantly stiffer and harder than the rubbery matrix of PU (T_g < RT). Specifically for the blank samples, we note that the decrease in the Young's modulus of Matrimid/ZIF-8 30 wt% compared with that of pristine Matrimid is associated with the incorporation of the ZIF-8 fillers (E_ZIF-8 ∼3 GPa),²⁸ which are more compliant than the Matrimid phase; this effect has been explained in detail in recent reports.^44,50 The nanoindentation data shown here were averaged from a surface penetration depth ranging from 1 to 2 μm.

The application of the nanoindentation technique is a practical choice: bulk-testing approaches such as uniaxial tensile tests of the membranes will result in the bulk properties (in this case, the matrix of the nanocomposites) being determined. The influence of the nanoparticles (ZIF-8) will subsequently be overlooked. The nanoindentation technique makes measured indents onto minute areas of the surface of the samples at controlled depths (up to ∼2 μm),^44,49 enabling it to investigate the delicate surface regions where the ZIF-8 nanoparticles, the polymer matrices, and incidentally, the absorbed I₂ are actually interacting. For the polymer matrices, this translates into probing the surface where the polymeric chains and I₂ are interacting. This method will allow us to measure relatively small mechanical property differences, more precisely capturing the change in the Young's modulus (elastic stiffness) and hardness (resistance against plastic deformation) of the ZIF-8 mixed-matrix membranes due to their absorption of I₂.

Fig. 6(a) and (c) show that the nanomechanical properties of the Matrimid-based membranes are only slightly affected by the absorption of I₂. This is reflected by the fact that the Young's Modulus (E) differs by ∼2% for Matrimid and no change was detected for its nanocomposite sample. While for the hardness (H) values, the error bars for both samples are within the range of each other, implying that the amount of I₂ absorbed by the Matrimid-based membranes are relatively small (all changes fall under the ∼5% range) vis-à-vis the nanomechanical properties. The glassy nature of the Matrimid matrix at room temperature (Matrimid glass transition temperature, T_g ∼345 °C)⁵¹ is important and may account for this outcome. It is envisaged that the molecular packing and rigidity underpinning the polymeric chains of Matrimid form an interconnected network that is sufficiently dense, thereby preventing the infiltration and uptake of I₂ guest molecules. We found that the introduction of ZIF-8 nanoparticles (fillers) into Matrimid, despite their expected disruption to the molecular packing of Matrimid,^30,31 did little to boost the overall I₂ absorptive capabilities, thus its marginal impact on the nanomechanical properties of Matrimid-based membranes.

However, a different trend was observed in the PU-based membranes (Fig. 6b and d). It is seen that the Young's modulus (E) and nanohardness (H) values changed by a discernible amount for neat PU (−25.5 MPa, corresponding to a ∼40% drop for E, and −3.5 MPa for H, corresponding to a drop of ∼27%) and PU/ZIF-8 30 wt% (28 MPa for E, corresponding to an increase of ∼6%, and 3 MPa for H, corresponding to a decrease of ∼11%) post-I₂ absorption. It should be pointed out that unlike in neat Matrimid, the nanomechanical properties in the neat PU sample displayed a contradictory behaviour where E and H values declined with absorption of I₂. This softening effect is speculated to be due to the pervasiveness of I₂ in the hard and soft segments of PU (see ESI† Fig. S2), which increased the interatomic distance between the polymeric chains, thus weakening the inter- and intra-molecular bonding interactions affecting their structural integrity.³⁴ We also propose that the looser molecular packing of the rubbery PU polymeric chain (T_g < RT) and the nominal molecular attachment of the ZIF-8 to the PU matrix accounted for the increased absorption of I₂ into the PU/ZIF-8 30 wt% nanocomposite, due to increased accessibility to ZIF-8 nanoparticles. The infiltration and I₂ uptake within the ZIF-8 nanoparticles will be assisted by the size of the porous cages of ZIF-8 (ref. 52) (11.4 Å against the size of the iodine molecule of 5.6 Å), if the (loosely packed) rubbery polymer chains will provide an open pathway to reach the ZIF-8 pores, thereby boosting the total I₂ uptake (see Fig. 4). Our results show that, for the PU/ZIF-8 30 wt% nanocomposite, upon I₂ absorption there is no major variation detected in E and H, both of which appear to lie within the respective statistical errors.

3.3 Thermogravimetric analyses (TGA) of the I₂-absorbing nanocomposite membranes

Fig. 7 shows the TGA profiles of the absorbing nanocomposite membranes and their blank counterparts. It is seen from Fig. 7(a)–(b) that the absorption by the Matrimid-based nanocomposites is almost negligible, as the profile of the absorbing sample is almost identical to the blank sample, with almost minute variations throughout the thermal decomposition profile. This is supported by the weight uptake differential data in Fig. 4, where the wt% changes post-I₂ absorption for the Matrimid-based nanocomposites remained under 3%.

Fig. 7 TGA plots of (a) Matrimid, (b) ZIF-8/Matrimid 30 wt%, (c) PU, and (d) PU/ZIF-8 30 wt%, with their respective blank counterparts. For an overall combined plot of TGA, including the ZIF-8 nanoparticles, refer to Fig. S5 in the ESI.†

On the contrary, as per Fig. 4, there is a marked rise in weight change (+32.4%) in the PU/ZIF-8 30 wt% composite, which is also reflected in the thermal decomposition profile of PU and PU-based nanocomposites. For PU, at ∼210 °C, the I₂-loaded composite began to decompose. This is speculated to be due to the initiation of thermal degradation of organics, and with PU molecular structure degradation, the I₂ bonded to the soft segment of the PU polymeric chain is being released. It should also be pointed out that the flash point of I₂ is at 184.3 °C; however, I₂ only began to evaporate from PU at 210 °C, meaning that the host–guest interactions shield the I₂ entrapped in PU from evaporating at its characteristically lower temperature stated above. We proposed that only when the chain integrity of PU was compromised that I₂ was exposed to the surrounding thermal conditions, thus triggering its evaporation process.

The thermal decomposition of the I₂-loaded sample was more rapid compared to its blank counterpart. The decomposition of the hard segment of PU appeared to begin at 310 °C for the I₂-loaded sample, which is earlier than the blank sample, where it began to decompose at 405 °C. The reason for this is that the I₂-loaded sample has formed an intrinsic part of the microstructure of PU, and once I₂ began to be liberated from the hard segments (ESI† Fig. S2) due to structural degradation from the soft segments, the structural integrity of the hard segment begins degrading also, in conjunction with release of I₂. A similar trend was observed in the PU/ZIF-8 30 wt% nanocomposite, where thermal decomposition of the loaded composite began a lot sooner than the blank nanocomposite; however, it is unique in that I₂ is released from the decomposed soft segment of PU and decomposing ZIF-8 (∼350 °C, see ESI† Fig. S5). The results in Fig. 3 suggest that I₂ is absorbed into ZIF-8 cages as well as the soft and hard segments of PU. Upon thermal decomposition of PU, ZIF-8 nanoparticles are exposed, and the organic ligands forming bonds with I₂ and the hard segment of PU will began to decompose, releasing I₂ and breaking the chemical and physical interactions formed with PU.

The phenomenon surrounding the release of I₂ from ZIF-8 cages is obvious in Fig. 7d, where from 300–550 °C, the I₂-loaded sample decomposes at a rate that is slower than its blank counterpart. This can be attributed to the entrapped I₂ in the ZIF-8 cages being released as the ZIF-8 cages collapse because of thermal decomposition.⁵³ At 550 °C, when I₂ was completely removed, a sharp drop was observed in the wt% of the nanocomposite. Subsequently at 600 °C, its thermal decomposition profile matched that of its blank counterpart, showing that all the entrapped I₂ was indeed removed from both PU and ZIF-8. What is important to establish here is that, by comparing Fig. 7c and d, there is strong evidence supporting the notion that thermal stability of the I₂-absorbing PU/ZIF-8 nanocomposite has significantly improved as a consequence of guest immobilisation (section 3.4).

3.4 Proposed mechanism responsible for iodine absorption and immobilisation in MOF-polymer nanocomposite membranes

Fig. 8 illustrates the proposed absorption mechanism that explains I₂ uptake and retention by the polymer-MOF nanocomposites. Based on the observed I₂ uptake by the polymers and their corresponding ZIF-8 nanocomposites, we propose that the uptake of iodine will be made more efficient with the utilisation of a flexible polymeric molecular backbone. The high surface area of ZIF-8 nanoparticles (∼1650 m² g⁻¹)⁵⁴ is expected to offer significant active sites to afford adsorption of I₂,⁵⁵ mostly via combined actions of chemisorption⁵⁶ and physisorption;⁷ however, retention of the adsorbate is significantly strengthened by the polymer matrices. Both polymers (Matrimid and PU) exhibited almost negligible amounts of I₂ uptake (Fig. 4), and it is presumed that the high uptake shown by the composites (especially PU/ZIF-8 30 wt%) is due to I₂ entering the pores of ZIF-8 nanoparticles instead of the free volume within the polymers formed by the presence of the nanoparticles. However, the functional groups and aromatic moieties in both Matrimid and PU are potential sites for the weak molecular interactions with I₂ molecules; therefore, the possibility of I₂ storage within inter-polymer-MOF pockets (free volume) cannot be completely ruled out. The presence of these unoccupied pockets could also potentially serve as active molecular flexible pathways that allow for I₂ to infiltrate the composites and guide the molecules towards the more active and high-surface area ZIF-8 nanoparticles cages.

Fig. 8 The proposed I₂ absorption mechanism by the polymer-MOF nanocomposite. The introduction of ZIF-8 nanoparticles within the polymer matrix offers porosity attributed to the intrinsic sodalite cages of ZIF-8 (filler phase) and free volumes arising from the disruption of the molecular packing of the nanocomposite by the nanoparticle fillers. These voids could serve as active sites, attracting I₂ molecules and forming bonds with the mIm ligands of ZIF-8, and the ‘hard’ and ‘soft’ segments of the PU (see ESI† Fig. S2) or the active coordination sites present in Matrimid.⁴⁴

Fig. 8 details the plausible molecular level interactions between I₂, ZIF-8, and the polymers (PU in the diagram) in the ZIF-8 nanocomposites. The light yellow spheres indicate the void formed by the presence of ZIF-8 nanoparticles, while the black circles represent the inter-polymer-ZIF-8 pockets. These free volumes could be of large and small dimensions, which is incidentally not necessary for strong interactions with I₂ when compared to the continuous array of ordered active voids of the ZIF-8 nanoparticles. Smaller volume voids with higher surface area may result in stronger nanoscale confinement with incoming foreign guest species, such as I₂. This effect could be investigated in greater detail by means of, for example, positron annihilation lifetime spectroscopy (PALS).^54,57

Furthermore, comparing the chemical structures of the polymer matrices could provide additional insights into their potential role in the capture and retention of I₂ within ZIF-8 polymer nanocomposites. PU is made up of amide linkages that are suitable for non-covalent interactions, while its flexible ether moiety could potentially provide a site for strong interaction and catalyse molecular dynamics within the matrix that improves small guest molecular mobility. However, in the case of Matrimid, its rigid aromatic backbone could potentially hinder the formation of inter-polymer-ZIF-8 pockets via its strong interchain aromatic interactions.

It is clear from our experimental data that the PU samples are more susceptible to I₂ uptake compared to the Matrimid samples. We reasoned that the intrinsic porosity of PU and its ability to interact with guest species via weak interactions could explain these differences. The molecular flexibility of the PU's primary chain allows for entropic dynamics of its rubbery structure to capture I₂ more rapidly. On the contrary, tightly packed structural configurations, which are especially prevalent in Matrimid (glassy phase), could result in a lower accessible surface area, rendering them less effective for I₂ immobilisation. The introduction of ZIF-8 nanoparticles into both polymer matrices will create strong “anchoring sites” and increase the free volume (due to disruption of molecular packing) that will enhance the resulting nanocomposite capacity to immobilise I₂. On this basis, the model we proposed in Fig. 8 represents the most probable capture and storage mechanism for the Matrimid/ZIF-8 30 wt% and PU/ZIF-8 30 wt% nanocomposites developed in this work.

4. Conclusions

This work represents the first reported attempt to capture and retain iodine within membranes made from ZIF-8/polymer nanocomposites. A few conclusions can be drawn from the body of this work, as listed below.

• Iodine was indeed captured and retained within the pristine polymer membranes (albeit at small wt%), but the immobilisation of I₂ in the nanocomposite is prominent particularly in the case of PU/ZIF-8 30 wt% membranes.

• We identified that the rubbery polymer matrix (e.g. PU) results in better absorptive and retention capabilities, which could be associated with the more accessible ZIF-8 pores to afford physisorption and chemisorption processes and further accompanied by physical confinement inside free volumes of the polymers.

• We projected that the mechanism that allowed iodine to be captured and retained within the polymers and nanocomposites can be attributed to the molecular interactions and affinity between the mIm (deprotonated ligand of HmIm) of ZIF-8, iodine molecules, and some of the polymeric chains of the matrices. Therefore, this passive process requires almost no injection of external energy, but relies completely on molecular and chemical affinities existing between the multiple organic–inorganic molecular constituents in the nanocomposite.

• The formation of ZIF-8/polymer nanocomposites improved the deliverability of the samples and allowed the retention of I₂ at temperatures far beyond its flashpoint, thus boosting the thermal stability as evidenced from the TGA experiments.

• The initial design proposal in the context of iodine capture and retention would suggest that a combination of highly porous MOFs containing strong sorption sites (e.g. ZIFs,³⁰ MILs,^58,59 CuBTC,⁶⁰ CPOs^61,62etc.) and rubbery polymers, such as PU, polydimethylsiloxane (PDMS) and polyisobutylene (PIB), will be favourable for maximising iodine capture rates in the emerging class of MOF-based mixed-matrix membrane (MMM)^39,63,64 nanocomposites.

• There is a wide scope for future developments in the aforementioned areas, not only to better elucidate the underpinning physico-chemical mechanisms but also to design and tune suitable combinations of membrane nanocomposites to yield bespoke guest immobilisation capacities.

Acknowledgements

E. M. Mahdi would like to thank Yayasan Khazanah (YK) for the DPhil scholarship that funded his research work at Oxford University. We acknowledge the Royal Society Research Grants (RG140296) for equipment funding. The authors thank Mr. S. Ying and Prof. A. M. Korsunsky for access to the Scanning Electron Microscope facilities (MBLEM Laboratory at Oxford, EU FP7 Project iSTRESS (604646)), and Prof. Steve Roberts and Dr. David Armstrong for usage of nanoindentation facilities at Oxford Materials. We acknowledge the provision of advanced material characterisation facilities, including TGA, by the Research Complex at Harwell (RCaH), Rutherford Appleton Laboratory (RAL), Oxfordshire. We thank Dr. Marek Jura and Dr. Gavin Stenning at R53 Materials Characterisation Laboratory in ISIS RAL, for access to the Rigaku X-ray diffractometer.

References

1. A. Verbruggen, E. Laes and S. Lemmens, Renewable Sustainable Energy Rev., 2014, 32, 16–28.
2. V. S. Ramos, V. R. Crispim and L. E. Brandao, Appl. Radiat. Isot., 2013, 82, 111–118.
3. V. H. M. Visschers, C. Keller and M. Siegrist, Energy Policy, 2011, 39, 3621–3629.
4. L. G. Carneiro, E. A. de Lucena, C. d. S. Sampaio, A. L. Dantas, W. O. Sousa, M. S. Santos and B. M. Dantas, Appl. Radiat. Isot., 2015, 100, 70–74.
5. R. P. Rechard, J. H. Lee, E. L. Hardin and C. R. Bryan, Reliab. Eng. Syst. Safe., 2014, 122, 145–164.
6. A. Salama, M. F. El Amin and S. Sun, Prog. Nucl. Energy, 2015, 85, 747–755.
7. D. F. Sava, M. A. Rodriguez, K. W. Chapman, P. J. Chupas, J. A. Greathouse, P. S. Crozier and T. M. Nenoff, J. Am. Chem. Soc., 2011, 133, 12398–12401.
8. D. F. Sava, K. W. Chapman, M. A. Rodriguez, J. A. Greathouse, P. S. Crozier, H. Zhao, P. J. Chupas and T. M. Nenoff, Chem. Mater., 2013, 25, 2591–2596.
9. S. Ma, S. M. Islam, Y. Shim, Q. Gu, P. Wang, H. Li, G. Sun, X. Yang and M. G. Kanatzidis, Chem. Mater., 2014, 26, 7114–7123.
10. O. B. Yang, J. C. Kim, J. S. Lee and Y. G. Kim, Ind. Eng. Chem. Res., 1993, 32, 1692–1697.
11. H. Sun, P. La, Z. Zhu, W. Liang, B. Yang and A. Li, J. Mater. Sci., 2015, 50, 7326–7332.
12. K. W. Chapman, P. J. Chupas and T. M. Nenoff, J. Am. Chem. Soc., 2010, 132, 8897–8899.
13. B. Assfour, T. Assaad and A. Odeh, Chem. Phys. Lett., 2014, 610–611, 45–49.
14. C. Falaise, C. Volkringer, J. Facqueur, T. Bousquet, L. Gasnot and T. Loiseau, Chem. Commun., 2013, 49, 10320–10322.
15. G. Massasso, J. Long, J. Haines, S. Devautour-Vinot, G. Maurin, A. Grandjean, B. Onida, B. Donnadieu, J. Larionova, C. Guerin and Y. Guari, Inorg. Chem., 2014, 53, 4269–4271.
16. G. Massasso, M. Rodriguez-Castillo, J. Long, A. Grandjean, B. Onida, Y. Guari, C. Guerin and J. Larionova, J. Mater. Chem. A, 2015, 3, 179–188.
17. H. C. Zhou, J. R. Long and O. M. Yaghi, Chem. Rev., 2012, 112, 673–674.
18. J. C. Tan and B. Civalleri, CrystEngComm, 2015, 17, 197–198.
19. H. Furukawa, K. E. Cordova, M. O'Keeffe and O. M. Yaghi, Science, 2013, 341, 1230444.
20. S. Kitagawa, R. Kitaura and S. Noro, Angew. Chem., Int. Ed., 2004, 43, 2334–2375.
21. P. Silva, S. M. Vilela, J. P. Tome and F. A. Almeida Paz, Chem. Soc. Rev., 2015, 44, 6774–6803.
22. S. Galli, N. Masciocchi, V. Colombo, A. Maspero, G. Palmisano, F. J. Lopez-Garzon, M. Domingo-Garcia, I. Fernandez-Morales, E. Barea and J. A. R. Navarro, Chem. Mater., 2010, 22, 1664–1672.
23. A. K. Cheetham, C. N. Rao and R. K. Feller, Chem. Commun., 2006, 4780–4795.
24. G. Férey, Chem. Soc. Rev., 2008, 37, 191–214.
25. M. D. Allendorf, M. E. Foster, F. Leonard, V. Stavila, P. L. Feng, F. P. Doty, K. Leong, E. Y. Ma, S. R. Johnston and A. A. Talin, J. Phys. Chem. Lett., 2015, 6, 1182–1195.
26. P. D. Dietzel, R. E. Johnsen, R. Blom and H. Fjellvag, Chem. – Eur. J., 2008, 14, 2389–2397.
27. J. C. Tan and A. K. Cheetham, Chem. Soc. Rev., 2011, 40, 1059–1080.
28. J. C. Tan, B. Civalleri, C. C. Lin, L. Valenzano, R. Galvelis, P. F. Chen, T. D. Bennett, C. Mellot-Draznieks, C. M. Zicovich-Wilson and A. K. Cheetham, Phys. Rev. Lett., 2012, 108, 095502.
29. J. C. Tan, T. D. Bennett and A. K. Cheetham, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 9938–9943.
30. A. Phan, C. J. Doonan, F. J. Uribe-Romo, C. B. Knobler, M. O'Keeffe and O. M. Yaghi, Acc. Chem. Res., 2010, 43, 58–67.
31. T. M. Nenoff, M. A. Rodriguez, N. R. Soelberg and K. W. Chapman, Microporous Mesoporous Mater., 2014, 200, 297–303.
32. K. S. Park, Z. Ni, A. P. Cote, J. Y. Choi, R. Huang, F. J. Uribe-Romo, H. K. Chae, M. O'Keeffe and O. M. Yaghi, Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 10186–10191.
33. S.-L. Li and Q. Xu, Energy Environ. Sci., 2013, 6, 1656.
34. S. N. Wijenayake, N. P. Panapitiya, S. H. Versteeg, C. N. Nguyen, S. Goel, K. J. Balkus, I. H. Musselman and J. P. Ferraris, Ind. Eng. Chem. Res., 2013, 52, 6991–7001.
35. M. R. Ryder and J. C. Tan, Mater. Sci. Technol., 2014, 30, 1598–1612.
36. M. R. Ryder, B. Civalleri, T. D. Bennett, S. Henke, S. Rudić, G. Cinque, F. Fernandez-Alonso and J. C. Tan, Phys. Rev. Lett., 2014, 113, 215502.
37. J. Caro, Curr. Opin. Chem. Eng., 2011, 1, 77–83.
38. N. Stock and S. Biswas, Chem. Rev., 2012, 112, 933–969.
39. B. Zornoza, C. Tellez, J. Coronas, J. Gascon and F. Kapteijn, Microporous Mesoporous Mater., 2013, 166, 67–78.
40. L. Diestel, N. Wang, B. Schwiedland, F. Steinbach, U. Giese and J. Caro, J. Membr. Sci., 2015, 492, 181–186.
41. H. B. Tanh Jeazet, C. Staudt and C. Janiak, Dalton Trans., 2012, 41, 14003–14027.
42. T. Yang, G. M. Shi and T.-S. Chung, Adv. Energy Mater., 2012, 2, 1358–1367.
43. Q. Song, S. K. Nataraj, M. V. Roussenova, J. C. Tan, D. J. Hughes, W. Li, P. Bourgoin, M. A. Alam, A. K. Cheetham, S. A. Al-Muhtaseb and E. Sivaniah, Energy Environ. Sci., 2012, 5, 8359.
44. E. M. Mahdi and J. C. Tan, J. Membr. Sci., 2016, 498, 276–290.
45. D. F. Sava, T. J. Garino and T. M. Nenoff, Ind. Eng. Chem. Res., 2012, 51, 614–620.
46. J. Cravillon, S. Münzer, S.-J. Lohmeier, A. Feldhoff, K. Huber and M. Wiebcke, Chem. Mater., 2009, 21, 1410–1412.
47. Y. C. Pan, D. Heryadi, F. Zhou, L. Zhao, G. Lestari, H. B. Su and Z. P. Lai, CrystEngComm, 2011, 13, 6937–6940.
48. S. Parshamoni, S. Sanda, H. S. Jena and S. Konar, Chem. – Asian J., 2015, 10, 653–660.
49. J. C. Tan, C. A. Merrill, J. B. Orton and A. K. Cheetham, Acta Mater., 2009, 57, 3481–3496.
50. N. W. Khun, E. M. Mahdi, S. Q. Ying, T. Sui, A. M. Korsunsky and J. C. Tan, APL Mater., 2014, 2, 124101.
51. A. C. Comer, D. S. Kalika, B. W. Rowe, B. D. Freeman and D. R. Paul, Polymer, 2009, 50, 891–897.
52. S. A. Moggach, T. D. Bennett and A. K. Cheetham, Angew. Chem., Int. Ed., 2009, 48, 7087–7089.
53. J. T. Hughes, D. F. Sava, T. M. Nenoff and A. Navrotsky, J. Am. Chem. Soc., 2013, 135, 16256–16259.
54. Q. Song, S. K. Nataraj, M. V. Roussenova, J. C. Tan, D. J. Hughes, W. Li, P. Bourgoin, M. A. Alam, A. K. Cheetham, S. A. Al-Muhtaseb and E. Sivaniah, Energy Environ. Sci., 2012, 5, 8359–8369.
55. T. D. Bennett, P. J. Saines, D. A. Keen, J. C. Tan and A. K. Cheetham, Chem. – Eur. J., 2013, 19, 7049–7055.
56. J. T. Hughes, D. F. Sava, T. M. Nenoff and A. Navrotsky, J. Am. Chem. Soc., 2013, 135, 16256–16259.
57. D. Cangialosi, H. Schut, A. van Veen and S. J. Picken, Macromolecules, 2003, 36, 142–147.
58. S.-H. Huo and X.-P. Yan, J. Mater. Chem., 2012, 22, 7449.
59. J. O. Hsieh, K. J. Balkus, J. P. Ferraris and I. H. Musselman, Microporous Mesoporous Mater., 2014, 196, 165–174.
60. A. K. Chaudhari, I. Han and J. C. Tan, Adv. Mater., 2015, 27, 4438–4446.
61. J. Kahr, R. E. Morris and P. A. Wright, CrystEngComm, 2013, 15, 9779.
62. P. D. Dietzel, P. A. Georgiev, J. Eckert, R. Blom, T. Strassle and T. Unruh, Chem. Commun., 2010, 46, 4962–4964.
63. H. B. T. Jeazet, C. Staudt and C. Janiak, Dalton Trans., 2012, 41, 14003–14027.
64. G. X. Dong, H. Y. Li and V. K. Chen, J. Mater. Chem. A, 2013, 1, 4610–4630.

---

Footnote

† Electronic supplementary information (ESI) available: XRD plots of the blank nanocomposites, raw data pertaining to nanoindentation, and TGA compilation of all the nanocomposites, including ZIF-8 nanoparticles. See DOI: 10.1039/c5me00008d

---