Insoluble Acyclic Cucurbit[n]uril-Type Receptors Capture Iodine from the Vapor Phase

Abstract

Nuclear energy makes large contributions toward meeting global energy needs, but societal concerns remain high given the impacts of the intended release of radioactive materials including ¹²⁹I and ¹³¹I. In this paper we explore the use of a homologous series of acyclic CB[n] type hosts (H1 – H4) as adsorbents of iodine from the vapor phase. We find that H2 – H4, but not H1 – perform well in this application with uptake capacities of 2.2 g g⁻¹, 1.5 g g⁻¹, and 1.9 g g⁻¹, respectively. The chemisorptive uptake process involves partial oxidation of catechol walled H2 to quinone walled host and capture of I₃ ⁻ and I₅ ⁻. Solid H2 can be regenerated by treatment with Na₂S₂O₄ and reused at least five times. The x-ray crystal structure of H2 is also reported.

Graphical Abstract

Catechol walled acyclic CB[n]-type receptors function as solid state sequestrants for iodine from the vapor phase with uptake capacities reaching 2.2 g g⁻¹ for H2. The uptake process involves both chemisorption (oxidation of H2; reduction of I₂) and physisorption processes (uptake of I₃⁻ and I₅⁻). H2 can be regenerated by treatment with aq. Na₂S₂O₄ and reused five times.

Introduction

The energy demand of modern society has increased rapidly in recent decades. The use of petroleum products to meet these needs has contributed significantly to environmental pollution and is implicated as a significant factor in climate change.^[1] Alternative power generation technologies include solar, wind, hydroelectric, geothermal, and nuclear.^[2] Among these alternative technologies, nuclear energy production is well established but societal concerns remain paramount given the well known disasters at Chernobyl, Three Mile Island, and Fukushima.^[3] Hazardous radioactive nuclides and species generated by nuclear power plants include ¹²⁹I, ¹³¹I, ³H, ¹⁴CO₂, and ⁸⁵Kr. Among these, ¹²⁹I and ¹³¹I have gained widespread attention due to their potential environmental impact. ¹³¹I has a short half-life (8.02 days), is taken up biologically, and is directly involved in human metabolic pathways.^[4] In contrast, ¹²⁹I has a half-life of 1.6 × 10⁷ years and therefore presents a persistent environmental hazard.^[5] Molecular iodine (I₂) readily sublimes above 45 °C, undergoes chemical reactions as either an oxidizing or a reducing agent, has high biocompatibility, and easily diffuses through the atmosphere and bodies of water. Therefore, the development of effective sequestrants to capture and store iodine is crucial.

An established technology to remediate radioactive iodine involves its chemical transformation into AgI using silver containing adsorbents.^[1a,6] Silver-based zeolites have been investigated for the uptake of iodine.^[7] However, the adsorbent capacity of these inorganic adsorbents are limited by their surface area and are costly which limits their practical application as adsorbents for radioactive iodine. Metal-organic frameworks (MOFs) – with their higher surface areas – have been investigated as adsorbents for I₂ and showed a higher adsorbent capacity than silver-based zeolites.^[1b,8] However, these MOFs exhibit low thermal and moisture stabilities which limit their use in nuclear power plants. Recently, a variety of macrocycles have been investigated as host adsorbents for I₂.^[9] For example, the Nau group reported that CB[6] binds to I₂ in water with a binding constant of 1.4 × 10⁶ M⁻¹ driven in part by C=O•••I₂ halogen bonding interaction.^[10] Subsequently, Cao and co-workers synthesized a porous polymer incorporating CB[6] and demonstrated the non-covalent uptake of I₂ vapor.^[11] Huang and co-workers pioneered the use of non-porous pillararenes for the reversible capture of iodine from the vapor phase and aqueous solution.^[1a,12] Other researchers have extended this line of inquiry by demonstrating I₂ capture using extended pillararenes and resorcinarenes,^[13] cyclooctaphenylenes and bipyridine cages,^[14] terphenarenes,^[15] bisindole macrocycles,^[16] cyclobenzoin hydrazones,^[17] Trögers-base macrocycles^[18] cycloviologens,^[19] and cyclotrihydroquinone macrocycles.^[20]

Over the past decade, our group has been intensively studying the synthesis and molecular recognition properties of acyclic CB[n]-type receptors and showed that they retain the essential molecular recognition properties of macrocyclic CB[n].^[21] Water-soluble acyclic CB[n]-type receptors function in a variety of important biomedical applications including as solubilizing excipients for the in vitro formulation and in vivo delivery of insoluble drugs,^[22] and as in vivo sequestrants for drugs of abuse (e.g. fentanyl, methamphetamine, PCP),^[23] neuromuscular blocking agents (rocuronium, vecuronium, and cisatracurium),^[24] and anesthetics (ketamine and etomidate).^[25] Recently, we reported the synthesis of a series of water-insoluble catechol-walled acyclic CBs for the solid-state sequestration of organic micropollutants from water.^[26] In this paper, we report our investigation of the use of solid H1 – H4 (Figure 1) as sequestrants for _I2 from the vapor phase.

Figure 1.

Molecular structures of acyclic CB[n]-type receptors studied in this paper.

Results and Discussion

This results and discussion section is subdivided as follows. First, we study the ability of the homologous series of acyclic CB[n]-type hosts H1 – H4 to capture iodine from the vapor phase. Next, we report the crystal structure obtained for H2. Next, we show that the most efficient sequestrant (H2) can be regenerated and reused over five cycles without significantly decreasing the iodine uptake. Finally, we probe the details of the uptake process by synthetic and spectroscopic means.

Selection of H1 –H4 as Sequestrants for I₂

Previous reports of I₂ capture by solid macrocyclic hosts^[1a,11] rely on non-covalent interactions between host and I₂. Previously, we found that catechol-walled host H4 functioned best as sequestrant for organic micropollutants from water^[26] by non-covalent interactions. The OH-groups of H4 serve to reduce the inherent water solubility of H4 (3.4 mM) but play an otherwise passive role. Below we demonstrate that the OH-groups of the best _I2 sequestrant (H2) play an active role by undergoing partial oxidation to the ortho-quinone walled acyclic CB[n]. We further demonstrate that the ortho-quinone walled acyclic CB[n] can be chemically reduced to H2 and reused.

Capture of Iodine from the Vapor Phase Using H1 – H4

Previous workers have performed I₂ vapor uptake experiments at 75 °C by gravimetric methods using a closed system comprising a capped outer jar along with three uncapped inner vials (iodine vial, adsorbent vial, reference vial) as shown in Figure S3.^[14a,15,27] To allow direct comparisons with the previously reported adsorbents, we employed an identical experimental setup. Previously, we have found that glycoluril tetramer derived acyclic CB[n]-type hosts display more potent molecular recognition properties due to their more fully formed ureidyl carbonyl portals which enhance ion-dipole interactions.^[28] Accordingly, we initially screened the uptake of _I2 vapor as the length of the glycoluril oligomer was increased from monomer (H1) to tetramer (H4). Solid H1 – H4 (10 mg) were activated before use by heating at 90 °C under high vacuum overnight. Experimentally, we monitor the I₂ uptake by measuring the change in weight of the adsorbent and the reference vials as a function of time and use eq. S1 to calculate the I₂ uptake (q_t, g g⁻¹) at each time point. Figure 2a shows the plot of iodine uptake as a function of the glycoluril oligomer length (H1 – H4) over time. Figure 2b shows the dramatic change in color of solid samples of H2 (beige to black) that occur during the I₂ uptake experiments at five time points. Hosts H2 and H4 display highest levels of I₂ uptake, reaching levels of 2.2 and 1.9 grams of I₂ per gram of host, respectively. The 2.2 g g⁻¹ uptake capacity of iodine displayed by H2 is equivalent to 12 atoms of iodine per molecule of H2. Glycoluril trimer derived host H3 also captures I₂ reaching a level of 1.5 g g⁻¹ at 62 hours. Conversely, glycoluril monomer derived host H1 performs quite poorly and only adsorbs 0.05 grams _I2 per gram H1. As an important control experiment, we measured the uptake of water gravimetrically under identical conditions at the 62 h timepoint. The water uptake is negligible for H2 (0.015 g/g), but more substantial for H3 (0.14 g/g) and H4 (0.285 g/g); the iodine uptake values reported above have taken the uptake of water into account. The excellent performance of H4 and H3 to a lesser extent was not surprising given the known binding affinity trends of acyclic CB[n] hosts^[28] as the glycoluril oligomer is lengthened. The outstanding performance of H2, however, was quite surprising given that glycoluril dimer derived hosts were previously shown to be quite complementary to planar aromatic cations.^[29] For comparison, the _I2 vapor uptake of H2 (2.2 g g⁻¹) is ten times higher than the non-porous ethylated pillar[6]arene reported by Huang and co-workers.^[1a] Table 1 presents the results of iodine uptake from the vapor phase for a variety of supramolecular systems reported in the literature to allow a more complete comparison with the results reported herein. Han and co-workers^[27b] reported that the imine-linked covalent organic framework (COF) prepared from tris(4-formylphenyl)amine and 1,3,5-tris(4-aminophenyl)benzene achieves a superior I₂ uptake of 7.94 g g⁻¹ in 96 hours. This high _I2 uptake value was attributed to the very high Brunauer-Emmett-Teller (BET) surface area of this COF (2348 m² g⁻¹). To determine whether the _I2 uptake trends seen for H1 – H4 were related to differences in surface area, we performed BET measurements for H1 – H4. Figure 2c shows the BET isotherm recorded for H2 which allowed us to calculate the surface area of H2 as 5.8 m² g⁻¹. Related BET measurements for H1 (47.0 m² g⁻¹), H3 (33.9 m² g⁻¹), and H4 (25.2 m² g⁻¹) were also performed (Supporting Information, Figures S16). Given the similar uptake capacities of H2 and H4 and their disparate BET surface areas suggests that iodine uptake along this series of compounds is not primarily dependent on surface area and lead us to delve further into the I₂ uptake process by H2.

Figure 2.

a) Plot of iodine vapor uptake by H1 – H4 over a period of 62 hours. Key: H1 (●), H2 (○), H3 (□), H4 (■). b) Photographs illustrating the color change of H2 at different time intervals (0, 2.5, 9, 20, and 50 hours) in the presence of iodine vapor. c) N2 adsorption isotherm of H2. Key: Adsorption (●), Desorption (○).

Table 1.

Iodine uptake capacities (g/g) for selected MOFs and macrocyclic host systems.

Host Material[b]	I2 Uptake (g/g)	Research group
Thiophene MOF (DUT 67 (Zr))	0.84	Wen Zhang[8b]
Thiophene MOF (DUT 68 (Zr))	1.08	Wen Zhang[8b]
Ethylated pillar[6]arene	0.2	Feihe Huang[1a]
Cationic macrocycle MOC-2	2.25	Li-Hui Cao[19]
cyclotrixylohydroquinoylene	0.76	Leyong Wang[20]
Tröger’s base [3]arenes	4.02	Leyong Wang[18b]
Tröger’s base cuboid	3.43	Juli Jiang[18a]
Bisindole [3]arenes	5.12	Cheng Yang[1b]
Aluminum Molecular Rings	0.5	Jian Zhang[30]
Bipyridine Cage	3.23	Xiaodong Chi[14a]
Tower[6]arene	2.08	Ying-Wei Yang[13b]
CB[6] polymer	0.25	Rong Cao[11]
Terphen[n]arenes	0.67	Chunju Li[15]

X-Ray Crystal Structure of H2

We were fortunate to obtain single crystals of H2 upon recrystallization from aq. TFA/MeOH, and the structure was solved by x-ray crystallography (CCDC 2359141). The crystalline form of H2 incorporates solvating H₂O and MeOH molecules. Figure 3 shows a cross-eyed stereoview of the packing of H2 in the crystal. A solvating MeOH molecule partially fills the cavity of H2 and forms a H-bond to the ureidyl C=O portal (O-H•••O= angle: 156.028°; OH•••O=C distance: 2.104 Å; O•••O=C distance: 2.892 Å) of H2. The individual units of H2 form tapes^[31] due to π-π interactions between the convex faces of the o-xylylene sidewalls roughly along the xz-diagonal as described previously for related compounds.^[32] The mean planes of the aromatic sidewalls are parallel (interplanar angle: 0°) and separated by 3.55 Å which is only slightly longer than the preferred π-stacking distance (≈ 3.4 Å).^[33] The individual tapes of H2 align with their long axes parallel to form sheets. The tapes are held together by bridging water molecules that form two H-bonds from the ureidyl C=O groups of one tape to the intracavity MeOH of the opposing tape of H2. The metrics for these two H-bonds are: 1) O•••O distance, 2.774 Å; OH•••O distance, 1.710 Å; O-H•••O angle 172.623° and 2) O•••O distance, 2.712 Å, OH•••O distance, 1.680 Å; O-H•••O angle, 163.204°. Finally, these sheets stack along the y-axis whereby the knobby equatorial CH3-groups on the convex face of the glycoluril dimer backbone of H2 pack into holes on the opposing sheet.

Figure 3.

Cross-eyed stereoview of the packing of H2 in the crystal. Color code: C, grey; H, white; N, blue; O, red; H-bonds, red-yellow striped.

Regeneration and Reuse of H2

Based on the noteworthy uptake of I₂ vapor by H2, we decided to study its regeneration and reuse over five consecutive cycles (Figure 4a). Since H2 is completely insoluble in common organic solvents, we decided to wash the solid generated by I₂ uptake with organic solvents. Surprisingly, the solubility characteristics of the generated solid was different than H2 which suggested that H2 underwent a covalent chemical change during iodine uptake. Given that I₂ is an oxidizing agent, we suspected that H2 was oxidized to the corresponding quinone DQ (Figure 4b) during the I₂ uptake process. Accordingly, we suspended the solid in 5% aqueous sodium dithionite (1.5 mL) and stirred overnight at room temperature and then obtained the residual solid by centrifugation. ¹H NMR spectroscopy (Figure S6, TFA solvent) established that the regenerated solid was H2 (93% purity). Alternatively, we also tried washing the solid generated after iodine uptake with 5% aq. Na₂S₂O₃ but observed less efficient regeneration of H2 as determined by ¹H NMR spectroscopy. We suspended the regenerated H2 in water (1.5 mL) and incubated it for 20 minutes at 25 °C, followed by centrifugation and removal of the supernatant to remove residual Na₂S₂O₄ and iodinated species. This washing process was repeated four times in total, and H2 was dried under high vacuum. The regenerated H2 was submitted for the next I₂ uptake cycle. Figure 4a shows the results of the regeneration and reuse of H2 over five cycles. The data shows that the first cycle has a lower I₂ uptake capacity (1.4 g g⁻¹) than the subsequent cycle which we attribute to the larger amount of H2 used in cycle 1 (50 mg) which results in less efficient diffusion of I₂ into the bottom of the 1.8 mL sample vial. The use of a larger amount of H2 in cycle 1 was necessitated by mechanical losses of material that occur during the regeneration process. Even though it shows that there is a loss in the iodine uptake over the five consecutive cycles, we believe the iodine uptake of H2 to be noteworthy, given that it still adsorbs 1.3 g g⁻¹ in the fifth cycle. Based on the reductive regeneration process, the iodine uptake of H2 from the vapor phase likely involves both chemisorption and physisorption processes. Related regeneration processes using 5% aq. Na₂S₂O₄ were performed for the solids generated by the use of H3 and H4 for I₂ uptake. ¹H NMR spectroscopy of the regenerated solids showed the presence of pure H3 and H4 (Supporting Information, Figure S7–S8).

Figure 4.

a) Iodine vapor uptake in five consecutive cycles after the regeneration of H2. b) Structure of bisquinone DQ and Me₂Cat.

Independent Synthesis of Diquinone DQ

In order to provide further support for formation of ortho-quinone walled acyclic CB[n] during I₂ uptake, we decided to independently synthesize DQ. For this purpose, we allowed H2 to react with sodium periodate in acetonitrile for 4 h. Pure DQ was obtained in a meager 8% yield after washing with water to remove inorganic salts. The IR spectrum of DQ showed the C=O stretch of the ortho-quinone 1659 cm⁻¹. The ¹H NMR spectrum of DQ recorded in DMSO-d₆ displays the three CH₃ singlets (12:6:6 ratio) and four doublets for the two diastereotopic CH₂⁻ groups (4:4:2:2 ratio) that are expected given the depicted C_2v⁻ symmetric structure of DQ (Supporting Information, Figure S1). The ¹³C NMR spectrum of DQ recorded in TFA displays 11 resonances also in accord with C_2v⁻ symmetry. Reduction of DQ with Na₂S₂O₄ delivered H2 in 94% yield which shows that ortho quinone walled acyclic CB[n] are viable intermediates in the uptake process presented herein.

Probing the Nature of the Solid Generated after I₂ uptake by H2 and after Regeneration with Na₂S₂O₄

First, we performed powder X-ray diffraction (PXRD) analysis of H2 and the solid generated after I₂ uptake by H2 to probe the crystallinity before and after the uptake process. Figure 5a shows the PXRD patterns obtained for pure H2 and the solid generated after I₂ uptake. The PXRD pattern of pure H2 shows a series of well-defined reflections that indicates the crystallinity of the H2 sample used in the I₂ uptake experiments. In contrast, the PXRD pattern measured for the solid after I₂ uptake is broad and featureless which indicates that the sample has become amorphous during the I₂ uptake process.^[14a,27b] The PXRD patterns recorded for the solid obtained after I₂ uptake using H3 or H4 are also broad and featureless (Supporting Information, Figure S14–S15). In contrast, the PXRD pattern of the solid obtained after uptake of I₂ by H1 shows low-intensity diffraction peaks related to the crystalline structure of H1 (Supporting Information, Figure S12) as expected based on the very poor I₂ uptake capacity of H1 (Figure 2a). Next, we decided to probe the nature of the adsorbed iodine species using Raman spectroscopy.^[8b,27] Figure 5b shows the Raman spectrum recorded for the solid obtained after H2 was used to capture iodine. The Raman spectrum displays peaks at 110 cm⁻¹ and 170 cm⁻¹ which are assigned to I₃⁻ and I₅⁻, respectively.^[34] This result suggests that the uptake of I₂ by H2 causes some oxidation processes and formation of HI as byproduct which further reacts with I₂ to generate I₃⁻ and I₅⁻. The location of I₃⁻ and I₅⁻ in the solid material is unclear, but given the preference of CB[n]-type hosts for cations we suspect they reside outside the cavity probably H-bonded to the residual OH groups. We also measured the IR spectrum of the solid generated after iodine uptake (Supporting Information, Figure S18) and observed the presence of a band at 1653 cm⁻¹ which we attribute to the presence of an ortho-quinone moiety (vide infra) and band at 3358 cm⁻¹ which we attribute to the OH groups of residual H2. In addition, the positive ion mode electrospray mass spectrum of the solid (Supporting Information, Figure S17) showed the presence of ions at 689.30, 687.30, and 685.27 which we attribute to [H2 + H]⁺, monoquinone [H2 – 2H + H]⁺, and [DQ + H]⁺. Finally, we recorded the ¹H NMR spectrum of the solid generated after iodine uptake (Supporting Information, Figure S11) and observed at least two sets of new resonances for the bridging CH₂-groups and the ArCH₃ groups which reflect the partial oxidation of the aromatic walls. After establishing that both chemisorption (e.g. oxidation of H2) and physisorption (e.g. capturing of I₃⁻ and I₅⁻) we wanted to further probe the regeneration process. Washing of the solid generated after iodine uptake with Na₂S₂O₄ regenerates H2 as described above (Supporting Information, Figure S6). After washing the solid with water, we recorded the Raman spectrum of both the supernatant and the solid and observed signals for I₃⁻ at 110 cm⁻¹ and I₅⁻ at 170 cm⁻¹ in both phases. The partial retention of I₃⁻ and I₅⁻ in the regenerated solid helps explain the slow decrease in uptake capacity over the five cycles seen in Figure 4a.

Figure 5.

a) Powder X-ray diffraction pattern of H2 before (red) and after (blue) iodine uptake. b) Raman spectra of H2 after iodine uptake.

Importance of the glycoluril oligomer backbone

We performed additional control experiments to probe the influence of the glycoluril oligomer backbone on the iodine uptake and regeneration process. Iodine uptake experiments performed using wall Me₂Cat show that I₂ uptake is substantial (3.37 g g⁻¹ iodine in 50 hours). However, Me₂Cat cannot be regenerated by washing with 5% aq. Na₂S₂O₄ (Figure S21) likely due to decomposition reactions of the corresponding ortho-quinone. Next, we monitored the thermal stability of solid Me₂Cat at 75 °C over 24 hours (¹H NMR monitoring, Figure S20) and conclude that Me₂Cat is stable under these conditions. Therefore, the poor regenerability of Me₂Cat is not due to thermal degradation of Me₂Cat. Accordingly, we conclude that a key function of the glycoluril backbone of H2 – H4 is to suppress decomposition reactions of the ortho quinone walled acyclic CB[n] intermediates and enable the regeneration process.

Conclusion

In summary, we have studied the uptake of iodine from the vapor phase to solid acyclic CB[n]-type receptors H1 – H4. The x-ray crystal structure of H2 was obtained as its methanol solvate; the individual molecules arrange in the form of tapes due to π-π interactions between the convex faces of H2. We find that H2 – H4 all perform well as sequestrants for iodine with uptake capacities of 2.2, 1.5, and 1.9 g g⁻¹ after 62 hours. The results of Raman spectroscopy show that iodine is taken up in the form of I₃⁻ and I₅⁻ whereas mass spectrometry and ¹H NMR spectroscopy show that H2 is partially transformed to ortho quinone walled acyclic CB[n] (e.g. DQ) which establishes that both chemisorption and physisorption are operative in the iodine uptake process. The spent H2 adsorbent can be regenerated by treatment with Na₂S₂O₄ which reduces the quinone walls back to H2. The regenerated H2 can be reused over 5 cycles, albeit with a slightly reduced iodine uptake capacity. The work further establishes that water insoluble acyclic CB[n] have great potential as solid state adsorbents of toxic materials like radioactive iodine.

Supporting Information Summary

The data that support the findings of this study are available in the supplementary material of this article. Deposition Number 2359141 (for H2) contains the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service. The authors have cited additional references in the Supporting Information.^[26,35]

Data Availability Statement.

The data that support the findings of this study are available from the corresponding author upon reasonable request.