Anisotropic Swelling of a Single‐Crystalline Hydrogen‐Bonded Organic Framework Induced by Iodine Vapor Uptake

ABSTRACT

In this paper, we report that iodine vapor adsorption induces anisotropic swelling of a porous hydrogen‐bonded framework (HOF) composed of hexacarboxylic acid with hexaazatrinaphthylene (CPHATN). Single crystals of HOF CPHATN‐1a exhibited swelling along the stacking direction upon exposure to iodine after an induction period, achieving a maximum elongation ratio of 60% and iodine uptake up to 30% at 22°C by weight. Single‐crystal X‐ray diffraction analysis, microscopic observation, micro‐CT, Raman spectroscopy, as well as molecular dynamics calculations, revealed that iodine adsorption onto the HOF proceeds in the following two steps: (1) adsorption into void channels and (2) intercalation into the layers. The latter process causes anisotropic elongation of the crystals and perturbs crystalline periodicity. Intercalation of iodine into the interlayer was facilitated by attractive halogen bonds between the σ‐holes of iodine and the pyrazine nitrogen atoms of CPHATN. This is the first proposal of the swelling mechanism of organic porous crystals by iodine based on single‐crystal structures, and the results provide new insights into the effects of inclusion of heavy‐element molecules on porous frameworks, contributing the development of porous materials.

A single crystal of a layered hydrogen‐bonded organic framework (HOF) undergoes iodine uptake‐induced anisotropic swelling via adsorption into void channels and intercalation into the layers.

1. Introduction

Porous materials capable of selectively or massively adsorbing iodine have been actively researched for applications such as removing and trapping environmental pollutants, including radioactive ¹²⁹I species [1, 2], tuning conductivity through charge doping [3, 4, 5], and achieving regular spatial arrangements of exotic polyiodide ions [6, 7, 8]. Moreover, by inserting highly volatile iodine molecules with large absorption coefficients into porous crystals, fundamental properties of porous materials can be readily evaluated using spectroscopic and diffraction experiments, contributing to the development of new porous materials. To date, iodine sorption behaviors of various porous materials, such as metal‐organic frameworks (MOFs) [9] and covalent‐organic frameworks (COFs) [10], have been revealed. In particular, porous molecular crystals (PMCs) and porous organic crystals (POCs), which assemble molecules through intermolecular interactions, yield single crystals suitable for single‐crystal X‐ray diffraction (SCXRD) analysis. This allows precise discussion on iodine‐framework interactions and structure‐property relationships based on the crystal structure incorporating iodine, providing valuable insights from a structural perspective for the development of porous materials [11, 12, 13, 14]. Among these, we focused on hydrogen‐bonded organic frameworks (HOFs), porous crystals formed by the aggregation of molecules via hydrogen bonding [15, 16, 17, 18, 19, 20, 21]. This is because design principles that leverage the high directionality of hydrogen bonding enable the systematic construction of desired frameworks, such as isoreticular frameworks [22, 23, 24, 25]. Some HOFs were reported to show excellent iodine uptake [26, 27, 28, 29, 30, 31]. For example, Claverie, Dory, and coworkers synthesized a series of HOFs based on amide‐substituted tribenzodehydro[12]annulene derivatives and demonstrated that one of the HOFs, HOF_T‐Hex, exhibited significant iodine capture capacity over 6 g/g [26]. Ke and coworkers reported that single crystals of cross‐linked HOFs adsorbed iodine with expansion of the crystals [27, 28]. We also discovered that iodine molecules form threefold helical strands within the one‐dimensional channel of a HOF by adsorbing iodine onto activated empty HOF single crystals and analyzing their crystal structures [29]. The orthogonal arrangement of iodine molecules in the helical strand indicates the presence of typical halogen bonds between the sigma holes of the iodine molecules and the surrounding electron‐rich belts [32, 33].

This study reports the first observation of iodine uptake‐induced, anisotropic swelling behavior in activated HOF CPHATN‐1a [34], composed of hexaazatrinaphthylene‐based hexacarboxylic acid CPHATN (Figure 1a, b). Swelling is a well‐known phenomenon for networked materials: the folding of polymer sheets to form architectures [35, 36] and the actuation of composites combining polymers and organic crystals were achieved by swelling [37]. In the case of inorganic clay materials, 100‐fold swelling and shrinkage of layered clay crystals composed of H_0.8[Ti_1.2Fe_0.8]O₄·H₂O induced by polar amine was reported by Sasaki and coworkers [38]. Solvatechromic behaviors of porphyrin cations intercalated in a clay were induced by swelling of the clay with organic solvents [39]. However, HOF crystals that swell anisotropically upon iodine adsorption remain entirely unexplored, except for cross‐linked frameworks reported by Ke and coworkers [27, 28].

FIGURE 1.

(a) Molecular structure of CPHATN. (b) Crystal structure of HOF CPHATN‐1a. (c) Schematic representation for the iodine absorption‐induced anisotropic swelling process of a single crystal of HOF CPHATN‐1a. The process includes (1) adsorption of iodine into channel voids and (2) intercalation of iodine into the interlayer spaces.

Exposing single crystals of HOF CPHATN‐1a to iodine vapor resulted in anisotropic swelling along the stacking direction after an induction period, achieving an elongation of the crystals up to +60%. After iodine removal, the crystals shrink back to their original size, while cracks formed along the layers in the crystals. Through thorough analyses, including SCXRD analysis, microscopic observation, micro‐CT, and theoretical calculations, it was suggested that iodine adsorption onto HOF proceeds in the following two stages: (1) adsorption into channel‐like voids while maintaining the HOF crystal structure, followed by (2) intercalation of iodine molecules into the interlayer spaces (Figure 1c). Among them, the second process may cause anisotropic elongation of the crystal. Intercalation of iodine into the interlayer was facilitated by halogen interactions between the σ‐hole of the iodine and the pyrazine nitrogen atoms of CPHATN, as well as dispersion forces working between the π‐conjugated planes of the molecule and iodine. This is the first proposal of the swelling mechanism of organic porous crystals by iodine based on single‐crystal structures, and the results provide new insights into the effects of inclusion of heavy‐element species on porous frameworks, contributing the development of porous molecular materials.

2. Results and Discussion

2.1. Preparation of Activated HOF CPHATN‐1a

Bulk crystals of the 1,2,4‐trichlorobenzene (TCB) solvated HOF CPHATN‐1(TCB) were prepared by slow evaporation of a solution of CPHATN dissolved in N‐methylpyrrolidone and TCB at 100°C, according to our previous report [34]. The activation of the HOF was conducted by heating at 190°C under reduced pressure conditions, giving single crystals of activated HOF CPHATN‐1a, in which H‐bonded hexagonal network sheets are slip‐stacked in an AB manner to form the layered HOF.

2.2. Iodine Sorption Behaviors

Bulk crystals of CPHATN‐1a were placed under the saturated iodine vapor at room temperature (22°C) to observe iodine adsorption behavior of the HOF (Figure S1). The vapor pressure of iodine at this temperature is ca.0.04 kPa (0.3 mmHg) [40]. Morphological changes of representative crystals are shown in Figure 2 and Figure S2. The HOF crystals changed color from yellow to black via red, indicating uptake of iodine in the void channels of the HOF. Moreover, the rectangular‐shaped crystals elongated anisotropically along the longer direction upon further exposure to iodine vapor. When the iodine vapor was removed and the crystals were exposed to the atmosphere, the elongated crystals shrank. Leaving the crystals at room temperature resulted in them remaining colored and slightly elongated (Figure S3). When the crystals were heated at 100°C to remove the included iodine, the resulting crystals exhibited the original crystal sizes, but they had cracks. Their detailed structures will be discussed later. Indexing of the single crystal revealed that the widest crystal face corresponds to the (011) plane, and the observed elongation occurs along the crystallographic a axis, corresponding to the direction of the void channels running. The cracks generated seem to be parallel to the {1 ‐2 ‐2} plane, corresponding to the plane formed by the H‐bonded 2D hexagonal sheet (Figure 3). The morphological changes of single crystals under iodine vapor were further monitored under brighter ambient light (Figure S4). It is observed that Iodine also penetrated through micro‐cracks and defects generated during the activation, leading to the progression of striped coloring. Elongation also progressed as the black stripes grew.

FIGURE 2.

(a) Morphological changes of single crystalline HOF CPHATN‐1a under iodine vapor at room temperature. (b) Anisotropic swelling of the selected five single crystals exposed to iodine vapor for 2, 10, and 24 h, and subsequently left under air for 1 h at room temperature, and left for another 48 h at 100°C. Scale bar: 100 µm.

FIGURE 3.

Relationship between crystalline morphology and molecular arrangements for CPHATN‐1a.

To quantify the crystal elongation upon iodine adsorption, the morphology of randomly selected 28 single crystals of CPHATN‐1a from two batches was monitored for 47 h under saturated iodine vapor at room temperature (Figure S5). In all cases, the elongation of crystals was observed along the long axis direction (Figure 4a): The average value of the elongation ratio is 47%, and the observed maximum ratio was 62%, although the elongation behaviors are sensitive to the conditions, and therefore, are dependent on experimental batches. It is noteworthy that the time course of the elongation shows a sigmoidal profile (Figure 4b). The elongation started after induction periods of ca. 20 h and reached the plateau after ca. 47 h. To quantify the kinetics of the elongation, the Avrami equation was applied to the morphological change in the elongation period as the time‐dependent volume fraction [41, 42]. The resultant Avrami exponent is n = 1.88, which is close to 2, indicating the sporadic nucleation (n _N = 1) and one‐dimensional transformation (Figure S6). Induction period, on the other hand, could not be fitted by the equation. When the crystals swollen with iodine are exposed to the atmosphere, the expansion ratio rapidly decreased to 20% within 10 min and then reached a plateau (Figure 4c).

FIGURE 4.

(a) Crystal lengths of 28 single crystals of HOF CPHATN‐1 before and after exposure to iodine vapor for 47 h at room temperature (22°C) and their elongation ratio. (b,c) Time‐dependent change of elongation ratio of the single crystals upon (b) adsorption and (c) desorption processes at 22°C. (d, e) Time‐dependent changes of mass upon (d) iodine adsorption and (e) desorption processes at 22°C. In the desorption process, the sample was consequently heated at 100°C to remove iodine completely. (f) Photographs of crystalline powder of CPHATN‐1 at states i–v during iodine sorption experiments.

Time‐dependent changes of weight under iodine vapor were investigated at 22°C using crystalline powder of CPHATN‐1a (Figure 4d). The weight increased gradually over time and saturated at approximately 130%. The molar ratio of CPHATN to I₂ after the saturation was 1:1.3. When adsorption was performed at 30°C, the weight increased to 140% due to the rise in saturated vapor pressure (Figure S7), and the resultant molar ratio of CPHATN to I₂ was 1:1.8. The desorption process was also recorded at 22°C. Weight reduction occurred gradually and reached 120% (Figure 4e). After 48 h, the sample was heated at 100°C, resulting in the complete removal of the iodine. The color changes of the samples corresponded to the changes in the mass (Figure 4f). It should be noted that the weight changes show no stepwise profile corresponding to the elongation, and that the host/guest ratio estimated from the weight changes is closely similar to that crystallographically estimated for the single crystals that accommodate iodine but has not started elongation (details to follow later), indicating that the elongation may not need further adsorption of iodine, in other words, iodine molecules accommodated in the channels move between layers to elongate the crystals. Another possibility cannot be ruled out. Namely, elongation has not yet started in bulk crystals; the crystal elongation within bulk solids may be less likely to occur than in isolated single crystals.

2.3. Crystallographic Analysis

To reveal the crystal structure of iodine‐included HOF CPHATN‐1(I₂), a single crystal of CPHATN‐1a was sealed in a glass capillary in the presence of a piece of iodine and subjected to SCXRD analysis after ca. 10, 30, and 60 h to give crystal structures CPHATN‐1(I₂‐1), CPHATN‐1(I₂‐2), and CPHATN‐1(I₂‐3), respectively (Figure S8). Note that the iodine uptake in this condition proceeded slowly, and therefore, expansion of the crystal had not yet started. After 60 h of exposure to iodine, the framework was basically the same as the empty HOF CPHATN‐1a (Table S1, Figure 5a and Figure S9) [43], while each cell parameter slightly altered with the exposure time (Figure 5b). The a axis is slightly shrunken by 0.8%, while the b and c axes are elongated by 3.3% and 0.7%, respectively. The γ became larger by 2.7%, while the α became smaller by 1.2%. Consequently, the cell volume became larger from 3127.50(13) ų to 3212.55(12) ų by 2.7%. The superimposed structure of these unit cells clearly shows that the rotational conformation of the peripheral phenylene groups beside the included iodine gradually changes from 79.2° to 89.4° as the exposure time increases (Figure 5c and Figure S10, Table S2). This phenyl group is the site of inherently large static or dynamic displacement. The observed changes in the other phenylene groups are much smaller (i.e., less than 1.9°).

FIGURE 5.

(a) Crystal structure of CPHATN‐1(I₂‐1). Anisotropic displacement ellipsoids were shown with 50% probability. (b) The changes in the cell parameters of CPHATN‐1a upon iodine adsorption. (c) Distortion of the unit cell and conformational changes of carboxy phenyl groups.

The iodine molecules are accommodated in the void channel, although they are highly disordered within the cavity, and therefore, it becomes challenging to accurately locate the iodine species inside. In structural refinement, residual electron peaks were assigned to iodine atoms, and their chemical occupancy was optimized. The molar ratios of CPHATN/I₂ estimated from chemical occupancies optimized in the crystallographic analysis are 1/0.6, 1/1.25, and 1/1.4 for CPHATN‐1(I₂‐1), CPHATN‐1(I₂‐2), and CPHATN‐1(I₂‐3), respectively. When the exposure time to iodine was further prolonged, the crystallinity of the single crystals became lower, resulting in failure of the SCXRD analysis (Figure S11). This suggests that excessive iodine insertion into the crystal disrupts its framework structure.

Next, iodine uptake of bulk crystals of CPHATN‐1a was monitored by powder X‐ray diffraction (PXRD) as shown in Figure 6 and Figure S12. Crystalline powder of CPHATN‐1a was sealed in a glass capillary with a piece of iodine and subjected to PXRD measurement using synchrotron X‐ray radiation with a wavelength of 1.00 Å. Again, iodine uptake may occur with different time courses from those in other experimental systems due to the different conditions. Of the two distinctly observed diffraction peaks at 2θ of 2.7° and 2.9° corresponding to the (001) and (010) peaks, the former temporarily weakens in intensity over time before strengthening again. During this process, the peak position remains unchanged. Meanwhile, the intensity of the (010) peak decreased and began to shift slightly toward the low‐angle direction starting at 66 h. The peak at 2.5° marked with an asterisk is not attributable to CPHATN‐1a itself; it is considered to be contamination from another crystalline phase generated during the activation. While it becomes prominent during the iodine adsorption process, its intensity subsequently decreases to a negligible level. The (01−1) peak at 4.4° exhibits a significant increase in diffraction intensity due to iodine adsorption, while its peak position remains unchanged. Since the (0 1−1) plane is located above aligned vacancies, it is thought that iodine incorporation into these vacancies enhances its diffraction intensity. Of the three weak diffraction peaks observed for CPHATN‐1a at 5.3–6.0°, two peaks showed an increase in intensity and shift into the smaller angle region by 0.1° as the duration of iodine exposure increased.

FIGURE 6.

Time‐dependent PXRD pattern changes of CPHATN‐1a upon I₂ exposure and simulated pattern of the obtained crystal structures. The peak marked with an asterisk is not attributable to CPHATN‐1a, indicating the existence of another crystalline phase generated during the activation.

From the above, it was found that the periodicity of the crystals changes slightly upon iodine adsorption. This is consistent with the changes observed in SCXRD analyses: simulated PXRD patterns of CPHATN‐1(I₂‐1), CPHATN‐1(I₂‐2), and CPHATN‐1(I₂‐3) are in good agreement with the trends of the PXRD patterns as shown at the bottom of Figure 6. However, any changes in diffraction patterns that suggest the significant elongation of the crystals were not observed. These results can be interpreted as follows. When a certain amount of iodine enters the empty spaces of the HOF, the HOF can keep single crystallinity except for slight changes in the cell parameters. Once the crystal starts to elongate, an amorphous region forms where periodicity is lost and thus becomes undetectable by diffraction, while the remaining crystalline parts shows the diffraction.

2.4. Microscopic CT Analysis

To investigate the morphology of the anisotropically swollen crystals in detail, micro‐computed tomography (micro‐CT) images using synchrotron X‐ray radiation were acquired for a crystal of CPHATN‐1a before iodine adsorption, an iodine‐included crystal after elongation had saturated, and a crystal from which iodine had been removed at 100°C overnight. The resultant phase images viewed from the orthogonal three different directions are shown in Figure 7. Note that each is not the same crystal, but rather an image of a separate crystal in that state.

FIGURE 7.

Micro‐CT images from the orthogonal three directions for (a,d) activated HOF, (b,e) fully elongated HOF under I₂ vapor, and (c,f) de‐iodinated HOF by heating. (a–c) Images of the entire crystals. (d–f) Magnified images of the crystals.

The crystals before iodine adsorption exhibit a very smooth appearance both on the surface and internally, despite having some defects. Fully elongated crystals exhibit large cracks and irregularities on their surfaces. Even in enlarged images, numerous fine cracks densely occur. Crystals after iodine removal show long cracks, but enlarged images reveal no multiple cracks like those in fully elongated crystals. Except for a few long cracks, the surface remains very smooth. Furthermore, examining the surface image and two internal cross‐sectional images of the fully elongated crystal shown in Figure S13 reveals that the interior is considerably smoother compared to the surface, which exhibits numerous cracks. This indicates that iodine is distributed fairly uniformly within the crystal (at least within the resolution limits of the present micro‐CT analysis). Fluorescence X‐ray analysis was also performed on the iodine‐doped crystal (Figure S14); analysis with sufficient resolution to observe an inhomogeneous distribution of iodine could not be achieved.

The above results indicate that iodine adsorption consists of the following two processes. (1) In the initial stage of iodine adsorption, iodine is absorbed into the void channels of the HOF. Consequently, while the crystal becomes colored, its crystal lattice and morphology remain unchanged. (2) In the subsequent process, iodine molecules intercalate between stacked molecules of the framework. Attractive interactions between iodine and CPHATN molecules cause the intercalation, and this pushes the crystal apart in the stacking direction and disrupts the crystal periodicity.

2.5. Crystal Structure of I₂ Eliminated HOF

It is remarkable that we successfully performed structural analysis on one of the single crystals that had undergone iodine swelling and had been left at room temperature under ambient conditions for 7 days, although the data quality was insufficient (i.e., R ₁ = 0.13) (Table S3). The obtained crystal structure, referred to as CPHATN‐2(I₂), is shown in Figure 8, along with that of CPHATN‐1(I₂‐3) for comparison. CPHATN‐2(I₂) still contains the encapsulated iodine and exhibits different cell parameters from those of CPHATN‐1(I₂‐3). The molecules in CPHATN‐2(I₂) essentially maintain the 2D hexagonal network through intermolecular H‐bonding between the peripheral carboxyl groups, although one of the carboxylic acid dimers is significantly deformed. Both HOFs have layered structures of the 2D hexagonal sheets with two types of stacking manners. The first layer (colored in orange) and the second layer (colored in cyan) are stacked so that all peripheral carboxyphenyl groups contact each other between the first and second layers. The second and third layers (colored in cyan and dark blue, respectively), on the other hand, have less overlap between them. There are gaps with a width of 5.8 Å between the carboxyphenyl groups in CPHATN‐2(I₂): the width is wider by ≈3 Å than that of CPHATN‐1(I₂‐3). Iodine molecules have penetrated into the gaps of CPHATN‐2(I₂).

FIGURE 8.

Crystal structures of iodine included HOFs (a,c) CPHATN‐2(I2) and (b,d) CPHATN‐1(I2‐3). Stacking manners of (a,b) layers and (c,d) HATN cores. Color code: orange (the first layer), cyan (the second layer), dark blue (the third layer), purple (iodine).

Focusing on a stacking manner of the HATN cores, in CPHATN‐1(I₂‐3), the cores are slip‐stacked in an antiparallel manner with a stacking distance of 3.07 Å between the first and second layers, while the cores are stacked with a stacking distance of ca. 3.13 Å so that one side of each overlaps between the second and third layers. In the case of CPHATN‐2(I₂), the cores in the first and second layers are stacked similarly to those in CPHATN‐1(I₂‐3) with a stacking distance of 3.05 Å. On the other hand, the cores in the second and third layers are stacked antiparallel so that their central phenazine skeletons overlap with a stacking distance of ca. 3.17 Å, reducing the steric hindrance around the nitrogen atom. Consequently, the iodine molecule forms halogen bonds with four nitrogen atoms of the HATN cores: N···I distance and I–I···N angles are 2.936 Å and 111.03°, 3.451 Å and 170.62°, 3.386 Å and 130.59°, and 2.951 Å and 130.51°. The molar ratio of CPHATN and iodine was estimated by optimization of the chemical occupancy to be 1/1.7. In the framework of CPHATN‐1, nitrogen atoms of the HATN core are difficult for guest species to approach from the outside of the HOF due to the steric hindrance.

In HOF crystals swollen with iodine, iodine likely infiltrated the interlayer spaces. Subsequently, during the deiodination stage at room temperature, the crystals were gently annealed, yielding a single crystal suitable for structural analysis. Although the simulated PXRD pattern of CPHATN‐2(I₂) differs from that observed in the iodine exposure experiments (Figure 6), the crystal structure is intrinsically linked to the intercalation behavior of iodine into the layered HOF. It goes without saying that this structure provides profound insights into that behavior. We also succeeded in the structural analysis of the crystal after the iodine had been completely removed, referred to as CPHATN‐3 (Figure S15 and Table S3), although many of the crystals lost their single crystallinity. The crystal structure is closely similar to that of CPHATN‐1a. Moreover, a PXRD pattern of a bulk sample after removing iodine by heating shows characteristic diffraction peaks ascribable to CPHATN‐1a, although the pattern is broader than before iodine adsorption (Figure S16). It should be noted that an additional broad shoulder or peak not ascribable to that of CPHATN‐1a was also observed at ca. 4.8° and 9.5°. These results indicate that the iodine adsorption causes the framework of CPHATN‐1a to become more disordered and to partially adopt a different molecular arrangement.

2.6. Microscopic Raman spectroscopy

Microscopic Raman spectroscopy was performed on two types of single crystals of well‐swollen HOFs and a slightly shrunk HOF obtained by leaving for 1 h under air, as shown in Figure 9. Crystal (1) is the iodine‐swollen HOF sealed in glass under an iodine atmosphere. Crystal (2) is a swollen HOF immediately after being removed from an iodine vapor environment into atmospheric conditions. Iodine solid showed a characteristic band at 185 cm^{− 1}. Crystals (1) and (2), swollen with iodine, both exhibited a broad band ranging from 150 to 200 cm^{− 1} with a peak at 166 cm^{− 1}. This band includes N···I stretching and I···I stretching modes, because it is reported that the pyridine‐iodine system exhibits a Raman band at 163 cm^{− 1}, which is interpreted as a stretching vibration mode between the nitrogen (N) atom and the iodine (I) atom in a charge‐transfer bond (i.e., halogen bond) [44]. This is consistent with the crystal structure of CPHATN‐2(I₂). A HOF crystal left to stand for one hour and allowed to shrink was Raman inactive in the 100–500 wavenumber region. This indicates that iodine had already leached out at the depth of penetration of the excitation light.

FIGURE 9.

Normalized microscopic Raman spectroscopy on single crystals of iodine‐induced swollen HOFs [crystal (1): dark blue, crystal (2): orange], 1 h‐left HOF (red line), and I2 solid (green).

2.7. Calculation of Stacking Interaction Energy

Intermolecular interactions of the stacked CPHATN molecules in activated HOF CPHATN‐1a, iodine included HOF CPHATN‐1(I₂‐3), and iodine annealed HOF CPHATN‐2(I₂) were calculated. As described above, these crystals with the P‐1 space group exhibit two types of stacking manners, as observed between dimers A and B (Figure 8c,d). Therefore, the stacking interactions in each dimer are calculated at B97D/6‐311+G(d,p) level of theory using the core parts and the whole molecules (Figures S17 and S18). The calculated interaction energies are listed in Table 1. CPHATN‐1a and CPHATN‐1(I₂‐3) exhibit slightly larger interaction energy in the core part of dimer A [E _int(A‐core) −23.8 and −23.9 kJ/mol, respectively] compared with that of dimer B (E _int(B‐core): −21.8 and −22.0 kJ/mol, respectively). Interaction energies of dimers of the molecules, on the other hand, show almost no differences between dimers A and B. The Interaction energies for dimer A in CPHATN‐2(I₂) [E _int(A‐core): −24.0 kJ/mol and E _int(A‐all: −46.1 kJ/mol)] are comparable to those in the former two, although the stacking geometry is different. On the other hand, the energy of dimer B [E _int(B‐core): −16.9 kJ/mol, E _int(B‐all): −37.7 kJ/mol)] is obviously lower than those of the former two. The destabilization of dimer B in CPHATN‐2(I₂) was caused by iodine entering the gap to form halogen bonds with the nitrogen atoms of CPHATN, altering the dimer's geometry. However, the loss is thought to be compensated for by the interaction with iodine.

TABLE 1.

Intermolecular interaction energies (E _int/kJ mol⁻¹) between stacked molecules calculated at B97D/6‐311+G(d,p) level of theory.

	CPHATN‐1a	CPHATN‐1(I2‐3)	CPHATN‐2(I2)
E int(A_core) a	−23.8	−23.9	−24.0
E int(B_core) b	−21.8	−22.0	−16.9
E int(A_all) c	−47.6	−47.7	−46.1
E int(B_all) d	−48.0	−47.6	−37.7

^a E _int calculated for the cores in dimer A.

^b E _int calculated for the cores in dimer B.

^c E _int calculated for the molecules in dimer A.

^d E _int calculated for the molecules in dimer B.

2.8. Molecular Dynamics Simulation

To investigate the dynamic structural changes of the HOF incorporating iodine, molecular dynamics (MD) calculations were performed on systems where 1, 2, and 3 equivalents of iodine molecules were introduced within the initial crystal structure created by repeating the CPHATN‐1a unit cell 12 times in the a‐axis direction and 6 times in the b‐ and c‐axis directions. Since the crystal structure of CPHATN‐2(I₂) suggests that the halogen bond between the nitrogen of the pyrazine ring and iodine may play an important role in the crystal elongation, we set a quadrupolar iodine model (model‐II) to mimic the electrostatic halogen bond, in addition to the conventional apolar iodine model (model‐I). In model‐II, for convenience, a charge of –0.10 was assigned to each of the iodine atoms, and dummy atoms carrying only the charge of +0.10 were introduced at both ends of the iodine molecule. No further tuning was attempted for the parameters of the halogen bond, given that quantitative assessment of the σ‐hole parameters is rather involved. Instead, we focused on qualitative aspects brought by the electron deficiency around iodine and the molar ratio of I₂ to CPHATN was varied beyond the experimentally observed. It should be noted that unlike experiments, there was no process of I₂ uptake in MD. Iodine molecules were located (“forced” to be) in HOF when the initial configuration of MD was prepared. MD was performed 5 times, and each run was done with 100‐ns equilibration and 100‐ns production. The time scale of MD is much smaller than that of the experiment, and phenomena within some hundred nanoseconds are captured in MD.

The cell parameters of the system are listed in Table 2. When 1 or 2 equivalents of iodine are added to the system, the respective axis lengths in both model‐I and model‐II systems remain virtually unchanged (≲1%), from the ones without added iodine. In the timescale of a few hundred nanoseconds employed in MD and with the amounts of iodine introduced in the present experiments, the crystalline lattice of CPHATN‐1 is hardly affected. This is in agreement with the experimental observations and holds true irrespective of whether the effects of halogen bonds are incorporated or not. Significant changes in the cell parameters are seen in the MD timescale only with 3 equivalents of model‐II iodine. Model‐I iodine does not form halogen bonds, and accordingly, it is inferred from MD that below µs, the HOF structure is distorted through interactions with σ‐holes only when excessive amounts of iodine are placed.

TABLE 2.

Cell parameters changes of the iodine‐added HOF estimated by MD simulation.

	a /Å	b/Å	c/Å
no I2 [ a ]	8.84	12.08	12.71
1 eq. I2 [ b ]	8.80(8.88)[ e ]	12.15(12.08)[ e ]	12.64(12.71)[ e ]
2 eq. I2 [ c ]	8.96(8.97)[ e ]	12.10(12.09)[ e ]	12.77(12.74)[ e ]
3 eq. I2 [ d ]	9.16(8.91)[ e ]	12.72(12.38)[ e ]	12.50(12.65)[ e ]

^a Without the addition of I₂.

^b With adding 1 eq. of I₂.

^c With adding 2 eq. of I₂.

^d With adding 3 eq. of I₂.

^e The values for model‐I and model‐II are listed in each entry, with the former in parentheses.

With the σ‐holes, electrostatic interactions can be operative between the iodine molecule and the framework. Indeed, plotting the distribution of the distance between the nitrogen atom and the center of mass of the iodine molecule reveals that a peak appears at ≈0.5 nm only in the model‐II systems (Figure 10). This peak indicates a halogen bond formation between the pyrazine nitrogen atom and an iodine molecule. In the case of model‐I, peaks are observed only at larger distances due to weaker interactions. With 1 or 2 eq of iodine, though, the effect of the σ‐hole on the distance distribution in Figure 10 is not significant. The peak at ≈0.5 nm is evident only with 3 equivalents of model‐II iodine.

FIGURE 10.

Distribution of the distance between the nitrogen atom and the center of mass of the iodine molecule with (a) one, (b) two, and (c) three equivalents of iodine molecules. (d) Magnified profiles of (c). Cyan: model‐I. Red: model‐II. g(r) means the radial distribution function, which is the ratio of the local density of iodine at a distance r to the bulk density throughout the simulated system.

3. Conclusions

In this paper, we reported an anisotropic swelling behavior of the layered hydrogen‐bonded organic framework (HOF) CPHATN‐1a composed of hexacarboxylic acid with a hexaazatrinaphthylene core, upon adsorption of iodine vapor. Single crystals of HOF CPHATN‐1a exhibited swelling along the stacking direction upon exposure to iodine after an induction period, achieving a maximum elongation ratio of 60% and iodine uptake up to 30% by weight. Based on single‐crystal X‐ray diffraction, micro‐CT, Raman spectroscopy, and theoretical calculations, we proposed that iodine adsorption proceeds in the following two steps: (1) adsorption into void channels while maintaining the HOF crystal structure and (2) intercalation of iodine molecules into the layers. The latter process causes anisotropic elongation of the crystals and perturbs crystalline periodicity. Intercalation of iodine into the interlayer was facilitated by an attractive halogen bond between the σ‐holes and the pyrazine nitrogen atoms of CPHATN. This is the first precise proposal of the iodine‐induced swelling mechanism of organic porous crystals. The present study provides new insights into the effects of the inclusion of heavy‐element molecules on porous frameworks, offering a profound understanding of the development of porous materials.

Funding

This work was supported by KAKENHI (JP23H04029, JP24K01468, JP25H01672, JP25H02042, JP22H05148, and JP23K27313) from JSPS and Fugaku Supercomputer Project (Nos. JPMXP1020230325 and JPMXP1020230327) and Data‐Driven Material Research Project (No. JPMXP1122714694) from MEXT Japan. I.H. thank Hoansha Foundation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: smll74287‐sup‐0001‐SuppMat.pdf.

Supporting File 2: smll74287‐sup‐0002‐MovieS1.mp4.

Supporting File 3: smll74287‐sup‐0003‐cif.zip.

Data Availability Statement

The data that supports the findings of this study are available in the supplementary material of this article.