Stereoselective Halogenation of Integral Unsaturated C‐C Bonds in Chemically and Mechanically Robust Zr and Hf MOFs

Abstract Metal–organic frameworks (MOFs) containing ZrIV‐based secondary building units (SBUs), as in the UiO‐66 series, are receiving widespread research interest due to their enhanced chemical and mechanical stabilities. We report the synthesis and extensive characterisation, as both bulk microcrystalline and single crystal forms, of extended UiO‐66 (Zr and Hf) series MOFs containing integral unsaturated alkene, alkyne and butadiyne units, which serve as reactive sites for postsynthetic modification (PSM) by halogenation. The water stability of a Zr–stilbene MOF allows the dual insertion of both −OH and −Br groups in a single, aqueous bromohydrination step. Quantitative bromination of alkyne‐ and butadiyne‐containing MOFs is demonstrated to be stereoselective, as a consequence of the linker geometry when bound in the MOFs, while the inherent change in hybridisation and geometry of integral linker atoms is facilitated by the high mechanical stabilities of the MOFs, allowing bromination to be characterised in a single‐crystal to single‐crystal (SCSC) manner. The facile addition of bromine across the unsaturated C−C bonds in the MOFs in solution is extended to irreversible iodine sequestration in the vapour phase. A large‐pore interpenetrated Zr MOF demonstrates an I2 storage capacity of 279 % w/w, through a combination of chemisorption and physisorption, which is comparable to the highest reported capacities of benchmark iodine storage materials for radioactive I2 sequestration. We expect this facile PSM process to not only allow trapping of toxic vapours, but also modulate the mechanical properties of the MOFs.

S10 formula and all values derived from the unit cell contents. Almost 80% of the cell volume is not occupied by the framework and contains diffuse and disordered solvent molecules. This electron density was accounted for using SQUEEZE within PLATON [S14]
peb-H 2 (0.082 g, 0.22 mmol, 1 eq) and concentrated HCl (0.02 ml) were added and the suspension was sonicated for a further 10 minutes before being placed in the oven at 120 °C for 60 hours. The bulk material was collected from the bottle upon completion, centrifuged once with fresh DMF (30 ml) and twice with acetone (2 x 30 ml), before being dried under vacuum ([Zr 6 O 4 (OH) 4 (peb) 6 ] n , 0.099 g, 0.035 mmol, 95% -average yield over 2 reactions).
Activation: Powder samples were added to 50 ml PYREX reagent bottles and left to stand in CHCl 3 . The CHCl 3 was exchanged for fresh CHCl 3 a further 4 times over 4 days, before being collected by centrifugation and dried under vacuum.

Single Crystals (4)
Single crystal synthesis of (4) was attempted using either L-proline or benzoic acid modulation, as in our experience these two modulators have provided the greatest enhancement of single crystal growth of Zr/Hf MOFs constructed from linear dicarboxylic acid ligands. [S7, S8] However, upon visual inspection of the crystals obtained from both sets of synthetic conditions it was obvious that the crystals obtained in the presence of benzoic acid S11 were much more defined in shape and were also larger in size (~100 µm compared with ~50 µm) than those synthesised using L-proline modulation. This difference was reflected during single crystal X-ray diffraction, with only poor low resolution data obtained for the L-proline modulated crystals. Therefore, single crystals were synthesised in the presence of benzoic acid to aid single crystal X-ray data collections (similar results were observed for the Hf analogue (5)).
peb-H 2 (0.082 g, 0.22 mmol, 1 eq) and HCl (0.02 ml) were added and the suspension was sonicated for a further 10 minutes before being placed in the oven at 120 °C for 24 hours. The bottles were removed from the oven after this period, and allowed to cool to room temperature. The crystals were left to stand in their mother solution.
Single crystal diffraction data were collected and processed using CrystalClear-SM Expert 3.1 b27. [S10] The structure was solved using Superflip [S15] and refined using SHELXL [S12] within Olex2. [S13] All non-hydrogen atoms were refined with anisotropic atomic displacement parameters (adps) with an enhanced rigid-body restraint (RIGU) applied. Hydrogen atoms were placed in geometrically calculated positions and included as part of a riding model, except the OH hydrogen atoms, which were not included explicitly in the model but are included in the cell contents and all values derived from them. Approximately 63% of the cell volume is not occupied by the framework and contains diffuse and disordered solvent molecules. This electron density was accounted for using SQUEEZE within PLATON [S14] which calculated a solvent accessible volume of 40049 Å 3 containing 7636 electrons, equivalent to approximately 190 molecules of DMF. The structure was deposited with the Cambridge Structural Database with deposition number CCDC 1443196.
Activation: Powder samples were added to 50 ml PYREX reagent bottles and left to stand in CHCl 3 . The CHCl 3 was exchanged for fresh CHCl 3 a further 4 times over 4 days, before being collected by centrifugation and dried under vacuum.
peb-H 2 (0.082 g, 0.22 mmol, 1 eq) and HCl (0.02 ml) were added and the suspension was sonicated for a further 10 minutes before being placed in the oven at 120 °C for 24 hours. The bottles were removed from the oven after this period, and allowed to cool to room temperature. The crystals were left to stand in their mother solution.
Single crystal X-ray diffraction data were collected and processed using APEX2, [S16] the structure was solved using Superflip [S15] and refined using SHELXL [S12] within Olex2. [S13] All non-hydrogen atoms were refined with anisotropic atomic displacement parameters (adps) with an enhanced rigid-body restraint (RIGU) and similarity restraint applied to the adps for C1 and an ISOR restraint to C1 and O3. Hydrogen atoms were placed in geometrically calculated positions and included as part of a riding model, except the OH hydrogen atoms, which were not included explicitly in the model but are included in the cell contents and all values derived from them. Approximately 63% of the cell volume is not occupied by the framework and contains diffuse and disordered solvent molecules. This electron density was accounted for using SQUEEZE within PLATON [S14] which calculated a solvent accessible volume of 40068 Å 3 containing 19296 electrons. The structure was deposited with the
However, we are aware that many of the MOFs desirable for these modifications may not be as chemically or mechanically stable as zirconium MOFs, and as such we have decided to investigate the use of N-bromosuccinimide (NBS) in the presence of N,N'-diphenylthiourea (DPT) as a milder, alternative brominating agent. [S17] The use of NBS in the presence of H 2 O is known to result in bromohydrination and so this was also explored as a potential route for the functionalisation of (1) (Scheme S3). Scheme S3. Reaction scheme for postsynthetic modification of (1) with NBS, highlighting potential products and the abbreviations used.
General Procedure for Postsynthetic Modification of (1) with NBS (1) (0.020 g, 0.053 mmol alkene, 1 eq) was suspended in the reaction solvent (5 ml) by stirring, then N,N'-diphenylthiourea (DPT) (10 mol %) was added if required, before cooling to 0 °C. N-Bromosuccinimide (0.047 g, 0.265 mmol, 5 eq) was added, the flask was sealed and the mixture was purged by passing N 2 through the system. The reaction mixture was left to stir under an N 2 atmosphere overnight, gradually warming to room temperature. The modified MOF was collected by centrifugation with fresh acetonitrile (2 x 10 ml) and acetone (2 x 10 ml), before being dried under vacuum. S15 1 H NMR spectroscopy of acid digested samples (DMSO-d 6 /D 2 SO 4 ) of the resulting MOFs was used to investigate the effect of reaction conditions on product distribution (Table S1). Addition of 15 equivalents of NBS (Entry 2) was necessary to quantitatively brominate (1).
Addition of N,N'-diphenylthiourea (DPT), which is known to promote bromohydrination, resulted in small amounts of the bromohydrinated product (1-Br-OH) when no water was added to the reaction (Entry 3), presumably resulting from excipient water. Increased amounts of NBS (Entry 4) gave slightly better conversion but still a low yield of (1-Br-OH).
Addition of water (Entry 5) gave significantly more bromohydrinated product, while the presence of water and DPT resulted in full conversion and up to 67% bromohydrination (Entries 6-8). Similar yields of (1-Br 2 ) and (1-Br-OH) could be achieved with as little as two equivalents of NBS in the reaction (Entries 9-10), and we found the reaction could also be scaled up with a similar product distribution (Entry 11). In our attempts to selectively bromohydrinate (1), the judicious adjustment of reaction conditions to facilitate the isolation of (1-Br-OH) did facilitate quantitative conversion of (1), but we were unable to exclude the formation of (1-Br 2 ). S16 Figure S1. Partial 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of acid digested samples of (1) (bottom) and the material after reaction with NBS/DPT (Table S1, Entry 11) (top). The stacked spectra clearly show that (1) has been completely converted into both (1-Br 2 ) and (1-Br-OH). Figure S1 shows a typical 1 H NMR spectrum of a bromohydrination attempt (Entry 11 from Table S1). The aromatic protons are overlaid for the two products, (1-Br 2 ) and (1-Br-OH), in the top 1 H NMR spectrum. The presence of the singlet at around 6.2 ppm is characteristic of bromoalkane protons, while the presence of two doublets at around 5.2-5.3 ppm reflects the loss of symmetry in the enantiomeric (R,S/S,R) bromohydrinated products (1-Br-OH). No succinimidyl by-products are observed in the 1 H NMR spectra of the digested MOFspresumably they are removed by the washing procedure.
Although useful for providing information regarding the conversion, 1 H NMR spectroscopic data were unable to provide any information on the retention of crystallinity of the MOF, although a solid phase reaction is suggested by the stereoselectivity that is observed.
However, this does not definitively prove that the reaction is occurring in a postsynthetic manner. Therefore, a typical bromohydrination reaction was scaled up (Entry 11) to allow the product to be analysed by powder X-ray diffraction (PXRD) ( Figure S2). S17 Figure S2. Comparison of the PXRD pattern of (1) with that of its postsynthetically derived products obtained using NBS.
Comparison of the PXRD patterns of (1) and its postsynthetically bromohydrinated/brominated products highlights the high mechanical stability of the MOF, as it is able to withstand a change in geometry of integral ligand atoms, as the central carbon atoms hybridisation changes from sp 2 to sp 3 . The chemical stability of (1) is also apparent, as the MOF remains crystalline in the presence of water, which is well known to result in the collapse of many late transition metal MOFs.
Whilst NBS has allowed us to bromohydrinate (1), we have chosen Br 2 as a favoured brominating agent for Zr-MOFs bearing integral unsaturated functionalities. . S18

S5. Bulk Postsynthetic Bromination of (3)
Upon consideration of the attempts to postsynthetically brominate (1) using either NBS or neat bromine, it is apparent that the greatest success has been obtained using bromine as the brominating agent. Therefore, neat bromine was selected for subsequent brominations of (3).

Bulk Bromination of (3)
Bulk powder of (3) (0.100 g, 0.50 mmol triple bond, 1 eq) was added to a 50 ml reagent bottle and left to stand in 15 ml CHCl 3 overnight. The CHCl 3 was removed and fresh CHCl 3 added, followed by the addition of bromine (385 µl, 7.52 mmol, 15 eq). The bottle was sealed before being stored in the dark for a period of 48 hours. The reaction product ( found. Bromination of (3) to (3-Br 4 ) was followed by liquid NMR spectroscopy using DMSOd 6 /D 2 SO 4 mixtures for digestion of the MOF ( Figures S3 and S4). The presence of D 2 SO 4 alters the pH of the NMR solution which in turn affects the chemical shifts of the signals from sample to sample, meaning that they cannot be compared directly between samples, but it is however still possible to gain an understanding of the species that are present in solution. Peaks marked with an asterix represent protons corresponding to benzoic acid, which was used as a modulator during synthesis of (3).

S19
The 1 H NMR spectrum of (3) ( Figure S3) contains small quantities of benzoic acid, which is also carried through to the brominated product (3-Br 4 ) suggesting that it is located at defect sites or becomes trapped in the pores of (3) during synthesis. There is limited information available from the 1 H NMR spectroscopic data due to the presence of only aromatic protons with the same splitting patterns in both the parent and brominated MOFs, although the increased splitting of the peaks that is observed is indicative of quantitative bromination to a single species, which we assume to be trans,trans-bdb-Br 4 -H 2 due to the steric restrictions imposed on the linker within the MOF. 13 C NMR spectroscopy ( Figure S4) was used to provide greater insight into the product distribution obtained. Figure S4. Stacked 13 C NMR spectra (DMSO-d 6 /D 2 SO 4 ) comparing acid digested samples of (2), (2-

Br 2 ) and (3-Br 4 ).
Previously, 13 C NMR spectroscopy provided definitive evidence of the stereoselective bromination of (2) to (2-Br 2 ) due to the occurrence of only six peaks for both species, confirming that only one of the two possible isomers had been obtained. [S8] Comparison with liquid phase bromination of the ligand contained within (2), as well as single crystal X-ray S20 crystallography, determined that trans-edb-Br 2 -H 2 was formed exclusively. Upon bromination, the dramatic upfield shift of the alkyne carbon from δ = 92.0 ppm to δ = 118.2 ppm provided conclusive evidence for the formation of the trans-dibromoalkene product, owing to the geometrical constraints of the MOF.
Likewise, in the 13 C NMR spectrum of (3-Br 4 ) ( Figure S4) we observe seven signals which correspond to the brominated product, alongside two small peaks around δ = 130 ppm which represent aromatic C-H signals of residual benzoic acid. The very close correlation of the chemical shifts for the trans-dibromoalkene carbon atoms between (2-Br 2 ) and (3-Br 4 ) (δ = 118.2 ppm vs δ = 117.6 ppm) suggests that the formation of the single species is the trans,trans-bdb-Br 4 -H 2 product -that is that the bromination of (3) occurs stereoselectively.
Direct comparison with the 13 C NMR spectrum of (3) was not possible, due to its very poor solubility.
In addition to the analytical techniques described in the manuscript, thermogravimetric analysis (TGA) was used to monitor the effect of bromination on the thermal stability of (3).
The TGA profile, collected under air, of (3) evidences its high thermal stability, with a major mass loss observed at around 460 °C, after losing ~5-10% mass which can be assigned to the exclusion of solvent molecules ( Figure S5, overleaf). Upon comparison of the brominated material (3-Br 4 ) with the parent material, it is evident that a two-step profile is observed (excluding solvent loss), with the first step representing debromination. Upon comparison of the profiles of (3) and (3-Br 4 ) it is clear that the brominated material is stable at higher temperatures, with the second pronounced mass loss now observed above 500 °C.
Assuming that the material prior to debromination represents the fully evacuated MOF, then the percentage mass loss during the first step can be used to provide an initial understanding of the bromine content of the MOFs. The mass loss during the first step of (3-Br 4 ), between ~200-450 °C, corresponds to a mass loss of ~26.4%, whilst the bromine content can be calculated to be 44.3%. This observation is inconsistent with the very close correlation found from elemental bromine analysis (44.3% calculated; 43.1% found) and the spectroscopic evidence which suggest quantitative bromination. Therefore we propose that in the case of (3-Br 4 ) it is not possible to separate the two mass losses for debromination and framework collapse using thermal analysis, hence there is a temperature range in which both processes are occurring simultaneously. This hypothesis is reinforced by the total mass loss observed between 200-550 °C of ~82.9%, which matches exactly the theoretical mass loss expected upon decomposition of (3-Br 4 ) to zirconium dioxide (Table S2). The mass loss in this temperature range for (3) is also very close to that expected for complete conversion to ZrO 2 .

S6. Solution Phase Bromination of bdb-Me 2
Solution phase bromination of bdb-Me 2 , the dimethyl ester of the linker bound within (3), was performed to provide a comparison of the product distribution, in terms of stereoselectivity, obtained when the brominations are carried out on the solid phase MOFs.
The bromination was carried out on the diester of the MOF linker as it exhibits much greater solubility in chloroform than the dicarboxylic acid itself.

Synthesis of bdb-Br 4 -Me 2
The diester bdb-Me 2 (0.060 g, 0.19 mmol, 1 eq) was added to a 50 ml reagent bottle and dissolved in 10 ml CHCl 3 by stirring. The mass spectrometry results indicate that the tetrabromodialkene species has been synthesised, with no evidence of any partially brominated products. This knowledge helped with the assignment of the 1 H NMR spectrum ( Figure S6, overleaf), which indicates the presence of more than one compound. The presence of three sets of resonances combined with the disappearance of the peaks relating to bdb-Me 2 provides evidence that the bromination has occurred quantitatively and that the product contains a mixture of three geometric isomers. 13 C NMR spectroscopic data was collected for the product, with a large number of resonances again suggesting the presence of multiple geometric isomers, although they could not be confidently assigned due to the similarity of their chemical shifts.
Comparison of the PXRD patterns of (4) and (5) with their brominated products (4-Br 4 ) and (5-Br 4 ) reveals that crystallinity is retained during bromination, with noticeable differences observed between the PXRD patterns of the starting materials and the halogenated products ( Figure S7, overleaf). Upon bromination of the alkynes to dibromoalkene units, there is an associated mechanical contraction, causing the peaks of the daughter materials to shift to slightly higher values of 2θ relative to their parents. The relative intensities and shapes of the peaks are also observed to change upon bromination, which together suggests that quantitative conversion has been achieved. Very similar patterns are observed for the brominated products (4-Br 4 ) and (5-Br 4 ), suggesting that they are structurally similar and that bromination occurs in a facile manner regardless of the metal used in the synthesis of the MOF. NMR spectroscopic analysis very difficult, and so small amounts of triethylamine were added to the solutions to solubilise the ligands. As a result of the low solubility it was not possible to follow the bromination of (4) and (5) using 13 C NMR spectroscopy, with only limited information available from 1 H NMR spectroscopic data ( Figures S8 and S9, overleaf).

The comparisons of both (4) with (4-Br 4 ) and (5) with (5-Br 4 )
show the formation of only a single species upon exposure to bromine, with changes in the chemical shifts observed. Both comparisons suggest the formation of a single brominated species of the same symmetry as the parent material, complementing the other analytical techniques utilised in the study. Figure S8. Stacked 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of acid digested samples of (4) vs (4-Br 4 ). Raman spectra of the MOFs are also indicative of quantitative bromination ( Figure S10, overleaf). In the parent MOFs, the signals at ~2220 cm -1 for (4) and (5) are characteristic of alkyne units. During bromination, these functional units are converted to dibromoalkene moieties, therefore it is expected that these peaks corresponding to C≡C bonds should no S27 longer be visible if the transformation has occurred quantitatively. In both brominated samples this is the case, with broadening of the alkene peak at ~1600 cm -1 observed, suggesting the appearance of a new alkene moiety, although the rapid damaging of the brominated samples by the Raman laser make this difficult to distinguish, particularly for (5-Br 4 ). It is clear however, that the signal associated with the alkyne unit is no longer present, indicating full conversion. Figure S10. Stacked Raman spectra highlighting the differences between a) (4) and its brominated product (4-Br 4 ) and b) (5) and its brominated product (5-Br 4 ). The Raman laser rapidly damaged the brominated materials, hence the poor quality of their spectra. Thermogravimetric analysis in air ( Figure S11, overleaf) was again employed to provide an understanding of the thermal stability of the interpenetrated MOFs and their brominated products. Both (4) and (5) demonstrate high thermal stabilities, with major mass losses observed at around 470 °C, as is expected for zirconium and hafnium MOFs.  (Table S3, overleaf). This further indicates that bromination has occurred quantitatively. Table S3. Summary of the expected (assuming complete conversion to ZrO 2 /HfO 2 ) and observed mass losses obtained from TGA measurements for the parent MOFs (4) and (5) as well as their brominated products (4-Br 4 ) and (5-Br 4 ).  Figure S12) to gain an understanding of the porosities of (4) and (5) as well as their brominated products, whilst also examining the effect of interpenetration.

Considering our efforts towards bulk phase brominations of the new zirconium MOF (3) with
UiO topology, and both the isoreticular interpenetrated zirconium (4) and hafnium (5) MOFs, we decided to investigate the feasibility of their transformation to (3-Br 4 ), (4-Br 4 ) and (5-Br 4 ), in turn, in a single crystal manner. The bromination of these MOFs, although it appears to be facile, does require a high degree of chemical and mechanical stability. This is due to the prolonged contact with a highly reactive bromine solution combined with the large amounts of mechanical strain that is induced during the contraction of the framework as the hybridisation of the integral carbon atoms changes from sp to sp 2 .

Single Crystal Bromination -General Procedure
Small amounts of single crystals of the MOF still in their mother liquor were added by pipette to a 10 ml vial containing fresh DMF (3 ml). The DMF was exchanged for fresh DMF twice before being exchanged for CHCl 3 (3 ml). The CHCl 3 was exchanged for fresh CHCl 3 a further two times. Bromine (100 µl for (3); 50 µl for (4) and (5)) was added, the vial was sealed and left to stand in the dark for 96 hours. The CHCl 3 was replaced multiple times with fresh CHCl 3 . Small portions of the crystals were added to a vial containing DMF to resolvate.
The DMF was exchanged for fresh DMF and the crystals left to stand.
Unfortunately, it did not prove possible to collect anything other than unit cell parameters for (3-Br 4 ), likely as a result of the significant disorder and frustration imposed by the linker.
The decrease in unit cell edge did, however, provide evidence of bromination and a concurrent mechanical contraction.
Single crystal diffraction data were collected and processed using CrystalClear-SM Expert 3.1 b27. [S10] The structure was solved using Superflip [S15] and refined using SHELXL [S12] within Olex2. [S13] Only the Zr atoms were refined with anisotropic atomic displacement parameters (adps), all other atoms were refined with isotropic adps and similarity restraints applied to the carbon atoms. Disorder was modelled as two 0.5 occupied sites for C7, both Br sites and C8 and C9 of the central aromatic ring (with C6 and C10 common to both orientations, see Figure S15). Distance and planarity restraints were applied. The Br atoms have large displacement ellipsoids and the residual electron density suggests that the S34 positions are smeared over a wider arc than has been modelled, particularly for Br2 which faces into the solvent accessible cavity. Hydrogen atoms were placed in geometrically calculated positions and included as part of a riding model, except the hydrogen atoms on C9, C10 and the OH hydrogen atoms, which were not included explicitly in the model but are included in the cell contents and all values derived from them (see Figure S15 for atom labelling). Approximately 44% of the cell volume is not occupied by the framework and contains diffuse and disordered solvent molecules. This electron density was accounted for using SQUEEZE within PLATON [S14]

S35
Single crystal diffraction data were collected using CrystalClear-SM Expert 3.1 b27 [S10] and processed using CrysAlis PRO 1.171.38.41. [S11] The structure was solved using Superflip [S15] and refined using SHELXL [S12] within Olex2. [S13] Only the Zr atoms were refined with anisotropic atomic displacement parameters (adps), all other atoms were refined with isotropic adps and similarity restraints applied. Disorder was modelled as two 0.5 occupied sites for C7 and both Br sites. Distance and planarity restraints were applied. The Br atoms have large displacement ellipsoids and the residual electron density suggests that the positions are smeared over a wider arc than has been modelled, particularly for Br2 which faces into the solvent accessible cavity. Hydrogen atoms were placed in geometrically calculated positions and included as part of a riding model, except the OH hydrogen atoms, which were not included explicitly in the model but are included in the cell contents and all values derived from them. Approximately 50% of the cell volume is not occupied by the framework and contains diffuse and disordered solvent molecules. This electron density was accounted for using SQUEEZE within PLATON [S14] which calculated a solvent accessible

S10. Iodination of (1), (2), (3) and (4)
Considering the reactivity of the MOFs towards bromine, we decided to investigate their potential as I 2 storage materials, which could find important applications in the nuclear industry to avoid the release of radioactive iodine. [S18] We designed two sets of experiments, firstly measuring the chemisorption of I 2 vapours by the MOFs by 1 H NMR spectroscopy, and secondly total iodine uptake through gravimetric experiments.

Iodine Chemisorption Experiments
In a typical reaction, (1), (2), (3) or (4) was added to an open-ended 1 dram vial. Separately, I 2 (15 eq per multiple bond) was added to a 30 ml screw top vial (Table S4). The small vial containing the MOF was then placed in the centre of the large vial, with the lid of the outer vial tightened to create a closed system. The experiments were allowed to stand at room temperature for the required number of days (1, 3, 7, 14 and 21 days, except for experiments containing (4) when data was also collected at 28 days). To stop the reaction, the small vial was removed and the contents added to a 15 ml centrifuge tube. The product was collected by centrifugation multiple times with CHCl 3 (10 ml) until washings were colourless, before drying under vacuum. The percentage chemisorbed I 2 was calculated from 1 H NMR spectroscopic integral ratios obtained from DMSO-d 6 /D 2 SO 4 digests of the products, as previously used to determine the extent of bromination. The bromination of (1) in chloroform is facile, with quantitative conversion achieved in the presence of as little as 5 equivalents of bromine. This is in stark contrast to the results obtained here, where iodine addition across the integral alkene bond through vapour diffusion does not occur even when exposed to 15 equivalents of iodine. This lack of reactivity is clear from the 1 H NMR spectroscopic data ( Figure S16), with no observed change in the deterministic alkene protons observed. Figure S16. Stacked partial 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of (1) compared with samples of (1) exposed to I 2 vapours for a set numbers of days. The spectra were referenced by setting the alkene proton singlet to 7.39 ppm, to counteract slight signal shifts due to differences in pH resulting from acid digestions.

S38
In contrast, 1 H NMR spectroscopic data reveal ( Figure S17) the successful addition of iodine from the vapour phase across the alkyne bonds of the edb 2ligands contained within (2), building on preliminary studies which investigated iodine chemisorption at 7 days. [S8] The comprehensive study conducted here however reveals that the amount of I 2 chemisorbed continues to increase until 14 days to a maximum value of ~84% conversion of the ligand to trans-edb-I 2 -H 2 , corresponding to 57% w/w irreversible chemisorption of iodine. Figure S17. Stacked partial 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of (2) compared with samples of (2) exposed to I 2 vapours for a set numbers of days. Red circles denote resonances assigned to trans,trans-edb-I 4 -H 2 . The spectra were referenced by setting the aromatic proton doublet to 7.63 ppm, to counteract slight signal shifts due to differences in pH resulting from acid digestions S39 Although complicated by the presence of residual benzoic acid modulator, the 1 H NMR spectra ( Figure S18) reveal that (3) is also reactive towards iodine vapours with new products obtained. A maximum value of 63% conversion of the alkyne bonds of bdb 2to iodinated products was observed, although only minimally rising from the 59% that was obtained after 14 days. After 1 day, there is evidence of formation of partially iodinated bdb-I 2 -H 2 alongside the fully iodinated trans,trans-bdb-I 4 -H 2 , with the majority of the iodination continuing to favour the formation of the fully iodinated ligand as the reaction time increases. The partially iodinated species is identifiable due to it lower symmetry, with some peaks coincident with the starting material and some with the fully iodinated linker. A maximum value of 88.5% w/w irreversible chemisorption of iodine was achieved. Figure S18. Stacked partial 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of (3) compared with samples of (3) exposed to I 2 vapours for a set numbers of days. Resonances marked with an asterix correspond to protons of residual benzoic acid from the synthesis of (3), red semicircles denote resonances assigned to protons of the partially iodinated bdb-I 2 -H 2 species, while red circles mark signals for the protons of trans,trans-bdb-I 4 -H 2 . The spectra were referenced by setting the aromatic proton doublet to 7.65 ppm, to counteract slight signal shifts due to differences in pH resulting from acid digestions.

S40
The reactivity of (4) towards I 2 vapours ( Figure S19) was initially slower than (2) or (3), however the rate of iodination increased and appeared to remain constant over a longer period of time with no decrease in the rate of iodination observed until 21 days. Similar to the iodination of (3), partially iodinated peb-I 2 -H 2 alongside the fully iodinated trans,trans-peb-I 4 -H 2 ligands was observed. Overall, 75% conversion of the alkyne units of peb 2to iodinated species was detected, corresponding to 79% w/w irreversible chemisorption of iodine. Figure S19. Stacked partial 1 H NMR spectra (DMSO-d 6 /D 2 SO 4 ) of (4) compared with samples of (4) exposed to I 2 vapours for a set numbers of days. Red semicircles denote resonances of protons of the partially iodinated peb-I 2 -H 2 , species while red circles mark signals of the protons of trans,trans-peb-I 4 -H 2 . The spectra were referenced by setting the aromatic proton doublet to 7.63 ppm, to counteract slight signal shifts due to differences in pH resulting from acid digestions.

S41
A plot of iodination conversion versus time reveals clear differences in the reactivity of the four MOFs towards iodine vapours ( Figure S20) which will influence their potential for use as iodine storage materials. The absence of iodinated products in the 1 H NMR spectra of digests of (1) is surprising, considering its high reactivity towards elemental bromine as well as NBS as a bromine source. The lack of reactivity of the alkene is in stark contrast to the alkyne units in (2), (3) and (4), which show total conversions in the 65-85% range after 21 days. Figure S20. Comparison of the I 2 chemisorption capacity of each of the four MOFs investigated.
During centrifugation steps, it became apparent that there were large quantities of iodine being removed by solvent rinses, suggesting iodine physisorption within the MOF pores.
Total iodine uptake measurements were then undertaken with data recorded over the same time range as the chemisorption experiments (although more data points were collected), allowing the total quantity of adsorbed iodine to be calculated.

Iodine Total Uptake Experiments
In a typical reaction, (1), (2), (3) or (4) (20 mg) was added to a pre-weighed 1 dram openended vial (Table S5). Separately, I 2 (15 eq) was added to a 30 ml screw top vial and the small vial containing the MOF was then placed in the centre of the large vial, with the lid of the outer vial tightened to create a closed system. The experiments were allowed to stand at room temperature with the small vial removed, weighed and placed back inside the large vial after 1, 3, 7, 14 and 21 days (data was also collected for the reaction containing (4) at 28 days). The gravimetric I 2 uptake of the MOFs was calculated from the observed increase in mass. Table S5. Summary of reaction quantities for total I 2 uptake experiments.
The simple experimental set-up was very efficient for determining the total iodine uptake capacity of the MOFs. Within a very short time (i.e. on the order of minutes) the MOF powders contained within the small vials were observed to darken, suggesting the diffusion of iodine occurred almost instantaneously. The uptake of iodine was monitored up to 21 days (28 days for the reaction containing (4)), and visualised by a plot of uptake versus time ( Figure S21, overleaf).
A noticeable observation from the gravimetric I 2 uptake curves is the capability of (1) to physisorb I 2 , yet no reaction occurs across the alkene bonds. There is a striking similarity between the weight % total I 2 uptake capacity of both (2) and ( Figure S21. Comparison of the gravimetric iodine uptake capabilities of the MOFs when exposed to iodine vapour. Again, similar to the chemisorption experiments it is clear that the uptake of I 2 by (4) is the slowest of the four MOFs, although it continues to adsorb I 2 for 28 days, which is the longest for the MOFs investigated. The ability of (4) to adsorb iodine for long periods of time results in an uptake performance that is by far the most superior, with a maximum uptake of 279% w/w recorded. The vast iodine uptake capacity of (4) is a result of both its ability to physisorb iodine within its pores, as well as irreversibly trapping it by chemical reaction across the alkyne units to convert them to diiodoalkene units. The chemisorptive capacity of (4) was actually lower than (3) when considered by weight %, therefore the superior uptake must be the result of a high tendency of I 2 to physisorb within the pores of (4). A plausible reason for the increased I 2 physisorption may be the result of the interpenetrated nature of (4), leading to an increased density of highly electron rich Zr 6 clusters. These results unambiguously highlight that zirconium MOFs are promising candidates as iodine capture/storage materials.

S11. Crystal Structures of Iodinated Linkers
The stereoselectivities of the iodination reactions were confirmed by the isolation of single crystals of iodinated linkers from digested samples (DMSO-d 6 /D 2 SO 4 ) of (2), (3) and (4) that had been exposed to iodine vapours. In all cases, crystals of DMSO solvates separated from the solutions.
Hydrogen atoms were placed at calculated positions and included in a riding model other than the methyl and OH hydrogen atoms which were refined as rigid-rotors. A region of poorly defined solvent was accounted for using SQUEEZE within PLATON [S14] which calculated a Single crystal X-ray diffraction data were collected and processed using APEX2 [S16] and the structure was solved using SHELXT and refined using SHELXL [S12] within Olex2. [S13] All non-hydrogen atoms were refined with anisotropic atomic displacement parameters (adps) with an enhanced rigid-body restraint (RIGU) applied. Hydrogen atoms were placed at calculated positions and included in a riding model other than the methyl and OH hydrogen

S47
Pyrollidine is able to successfully debrominate (1-Br 2 ), resulting in quantitative conversion back to the parent MOF (1). This is evident from the characteristic downfield shift of the alkene protons upon removal of the bromine atoms from the bromoalkane unit, regenerating an alkene. This transformation is an example of a tandem PSM event proving the high mechanical stability of these MOFs as the integral carbon atoms are observed to cycle from sp 2 to sp 3 and lastly back to sp 2 .