Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Mar 18;147(12):10640–10646. doi: 10.1021/jacs.5c00904

Discrimination of Perfluorinated Arenes/Alkanes by Modulable Polyaromatic Capsules

Urara Kai 1, Ryuki Sumida 1, Yuya Tanaka 1, Michito Yoshizawa 1,*
PMCID: PMC11951150  PMID: 40099388

Abstract

graphic file with name ja5c00904_0008.jpg

Efficient and selective binding of perfluorocarbons (PFCs), comprising only fluorine and carbon atoms, unlike hydrocarbons, remains quite difficult owing to the repulsive nature of fluorine. Here we describe that a cavity modulation strategy enables metal-linked polyaromatic capsules to quantitatively bind PFCs with excellent selectivity. From a mixture of perfluoroarene and the corresponding perfluoroalkane (i.e., perfluoronaphthalene and perfluorodecalin), a Pt(II)-linked capsule exclusively binds the arenes in water at room temperature, via effective D-A-A-D π-stacking interactions. The size-selective binding toward perfluoroarenes (i.e., perfluoronaphthalene and perfluorobenzene) is improved by using the analogous N-doped capsule from 85% to quantitative selectivity. Furthermore, unlike the Pt(II)-capsule, an isostructural Pd(II)-linked capsule displays the unusual length/shape-selective binding of linear/cyclic perfluoroalkanes (i.e., perfluoroheptane and perfluorodecalin) under similar conditions. Recognition of substituted hydrogen atoms on perfluorobiphenyl can also be accomplished using the Pt(II)-capsule. Various PFCs are thus clearly distinguished for the first time by the modulable polyaromatic capsules under ambient aqueous conditions.

Introduction

Perfluorocarbons (PFCs), composed of only fluorine and carbon atoms, display unique characteristics, derived from the extremely high electronegativity and low polarizability of the multiple fluorine atoms.1 Unlike the corresponding hydrocarbons, for example, perfluoroarenes provide highly electron-deficient aromatic frameworks, capable of efficiently stacking with nonfluorinated arenes. Perfluoroalkanes possess poorly interactive and relatively rigid frameworks, apt to repel nonfluorinated hydrocarbons as well as water molecules. These unusual, perfluoro-based physicochemical features have attracted attention from the viewpoints of synthetic, materials, and process chemistries.1 In addition, the effective collection and elimination of PFC derivatives (e.g., PFOA and PFOS) are of crucial importance from the viewpoint of recent environmental chemistry.2 To reveal and control intermolecular interactions at the molecular level, plenty of synthetic host compounds have been widely developed for various nonfluorinated aromatic and aliphatic guests, including complex biomolecules.3,4 However, to our surprise, PFCs can be bound only by a couple of (supra)molecular hosts, i.e., by tubular aliphatic cavities,5 a dimeric bowl-based aliphatic/aromatic cavity,6 and a cage-shaped aromatic cavity (Figure 1a–c).7 These cavities unfortunately provide relatively weak interactions with PFCs of various sizes and shapes so that their binding selectivity remains indistinct so far. In addition, there have been no reports on porous solid materials (e.g., metal–organic and covalent organic frameworks) for nonfunctionalized PFCs with high efficiency and selectivity.8

Figure 1.

Figure 1

Open in a new tab

Synthetic host compounds for perfluorocarbons (PFCs), providing (a) aliphatic, (b) aliphatic/aromatic, and (c) aromatic cavities, and (d) a polyaromatic cavity (this work). (e) Polyaromatic capsules 1a–c with a virtually closed cavity (details: see Figure S1) and the crystal structure of 1a (R = R′ = H and the counterions are omitted for clarity).9c,9d

For the discrimination of both aromatic and alkyl PFCs without functional groups, we herein focused on a capsular polyaromatic cavity (Figure 1d), expected to facilitate host–guest multi-interactions and size/shape complementarities. With the aid of polyaromatic capsules (Figure 1e) in water at room temperature, here we report that (i) Pt(II)-linked capsule 1a exclusively binds perfluoronaphthalenes over the corresponding perfluoroalkane and also binds the perfluoroarenes more selectively than smaller perfluorobenzenes (85% selectivity). (ii) The crystallographic analysis of the obtained host–guest composite indicates the formation of a D-A-A-D stacking structure in the cavity. (iii) The nitrogen-doping of the polyaromatic cavity allows Pt(II)-capsule 1b to bind perfluoronaphthalenes with improved selectivity (∼100% selectivity). Notably, (iv) unlike Pt(II)-capsule 1a, isostructural Pd(II)-linked capsule 1c binds perfluoroalkanes with excellent length- and shape-selectivity, under similar conditions. (v) Recognition of substituted hydrogen atoms on perfluorobiphenyl can be also accomplished by capsule 1a with quantitative selectivity. To clearly estimate actual interactions with not functional groups but perfluoro moieties, we focused on nonfunctionalized PFCs in this study.

For the present study, we selected three types of polyaromatic capsules 1a–c (Figure 1e),9 among various polyaromatic hosts reported previously,10 because of the following reasons. The anthracene-based electron-rich shell of capsule 1a (M = Pt(II) and X = CH)9c,9d allows the well-defined, spherical cavity (∼1.3 nm in diameter and ∼580 Å3 in volume)11 to selectively bind electron-deficient perfluoroarenes through electrostatic interactions. The cavity flexibility can be finely tuned by using N-doped capsule 1b (M = Pt(II) and X = N),9e which is an analogue of 1a, where the inner CH parts of the phenylene spacers are replaced by N ones. The replacement of the Pt(II) hinges on 1a with Pd(II) ones renders capsule 1c (M = Pd(II) and X = CH)9a,9b a flexible host with the same cavity volume, on the basis of the kinetically labile coordination bonds, to bind relatively rigid perfluoroalkanes with high efficiency and exclusive selectivity. Accordingly, the cavity properties of the present (nearly) isostructural capsules establish the distinct discrimination of various PFCs, for the first time, even from their mixtures in water, unlike other host cavities shown in Figure 1a–c. It should be noted that the observed high selectivities of 1a–c cannot be demonstrated toward the corresponding nonfluorinated hydrocarbons under the same conditions.

Results and Discussion

Binding Selectivity toward Perfluoroarenes

To clarify the discrimination ability of Pt(II)-capsule 1a toward aromatic and alkyl PFCs, we first employed perfluoronaphthalene (FN) and perfluorodecalin (FD; 23:77 cis/trans mixture), due to the same carbon atom number and similar molecular size,11 and revealed the binding preference of 1a toward different perfluoroarenes. When a mixture of FN and FD (5.0 μmol each) was combined with 1a (0.25 μmol) in D2O (0.5 mL) at room temperature for 1 h (Figure 2a), the quantitative formation (100% selectivity) of host–guest composite 1a·(FN)2 based on 1a was confirmed by NMR, MS, and X-ray crystallographic analyses.12,13 After removal of excess free guests by filtration, in the 1H NMR spectrum, all of the host aromatic signals (Ha–i) were shifted slightly, owing to the host–guest interactions in the cavity (Figure 2b,c). Particularly, the upfield shift of the inner phenylene signal (Ha) of 1a was observed by Δδ = −0.57 ppm. The NMR spectrum was consistent with that of 1a·(FN)2 (Figure 2d), prepared from 1a and FN under the same conditions. The 19F NMR spectrum of the product showed six signals derived from bound (FN)2, in the range of −151.5 to −148.0 ppm, without free FN signals (Figure 2f). These signals are shifted and desymmetrized, as compared with those of free FN in CDCl3 (Figure 2e), suggesting tight host–guest interactions in the dissymmetric host cavity.9 The formation of 1:2 host–guest composite 1a·(FN)2 was supported by the ESI-TOF MS spectrum, where molecular ion peaks corresponding to [1a·(FN)2 – n·NO3–]n+ (n = 4, 3) were observed at m/z = 972.3 and 1040.3 (Figure 2g), besides ion peaks for [1a·FN – n·NO3–]n+ generated only under the MS conditions. Neither NMR signals nor MS peaks derived from 1a·FD were detected from the reaction mixture (Figure 2d,f,g). Notably, 1a·(FN)2 was obtained quantitatively even using a 2.0 equiv. mixture of FN and FD (each) under the same conditions (Figure S7b).12,14,15

Figure 2.

Figure 2

Open in a new tab

(a) Selective binding of FN from a mixture of FN and FD by capsule 1a in water. 1H NMR spectra (500 MHz, D2O, r.t.) of (b) 1a, (c) 1a·(FN)2, and (d) products after mixing FN and FD with 1a at r.t. for 1 h. 19F NMR spectra (376/470 MHz, r.t.) of (e) FN in CDCl3 and (f) products after mixing FN and FD with 1a in D2O at r.t. for 1 h. (g) ESI-TOF MS spectrum (H2O) of 1a·(FN)2 after the competitive binding experiment and the expanded and simulated signals of [1a·(FN)2 – 4·NO3–]4+.

The X-ray crystal structure of 1a·(FN)2 revealed that two molecules of FN are fully accommodated within the closed polyaromatic shell of capsule 1a (Figure 3a).12,13 Pale yellow single crystals of 1a·(FN)2 were obtained by slow concentration of the saturated H2O solution at room temperature for 1 month. In the cavity, two molecules of FN are stacked with an interplanar distance of 3.4 Å, at the one of the fluorinated benzene rings, which is sandwiched between the two anthracene panels of 1a with interplanar distances of 3.4–3.6 Å (Figure 3b).16 The observed quadruple stack most likely stems from donor–acceptor–acceptor–donor (D-A-A-D) π-stacking interactions that stabilize the otherwise repulsive perfluoroarene–perfluoroarene (A–A) interactions. The 1:2 host–guest structure is stable enough at room temperature for 1 week and at 80 °C for 3 h (Figure S2b).12,15 The first binding constant (Ka1) of 1a toward FN was relatively estimated to be 2.5 × 106 M–1 by competitive binding experiments using FN and 18-crown-6 or phenylalanine dipeptide (Figure S8a and Table S1a),12,17 owing to the insolubility of FN in water. As control experiments, previous tubular and cage-like hosts (i.e., β-cyclodextrin5 and Fujita’s cage,7 schematically drawn in Figure 1a,c) showed poor to weak interactions with FN even under similar aqueous conditions (binding efficiency: 0.03 and 40% based on the host; Figures S9 and S10, respectively).12

Figure 3.

Figure 3

Open in a new tab

(a) X-ray crystal structure of 1a·(FN)2 (the side chains are replaced by hydrogen atoms and counterions/solvent molecules are omitted for clarity) and (b) the highlighted host–guest and guest–guest π-stacking interactions in the cavity.

The polyaromatic cavity of Pt(II)-capsule 1a distinguished the size of perfluoroarenes. Treatment of a mixture of FN and perfluorobenzene (FB; 20 equiv. each) with 1a in D2O at room temperature gave rise to a mixture of 1a·(FN)2 and hetero composite 1a·(FN·FB) in a 7:3 ratio and quantitative fashion based on 1a (Figure 4a, left).12 In the 1H NMR spectrum, two inner phenylene signals (Ha and Ha″) were observed at 5.69 and 5.97 ppm in a 69:31 integral ratio (Figure 4b). The former signal was assignable to 1a·(FN)2 on the basis of the NMR studies described above (Figure 2d,f). The latter signal indicated the formation of 1a·(FN·FB) by the 19F NMR spectrum, where new fluorine signals for FN and FB appeared in a 1:1 ratio in the range of −165.4 to −149.0 ppm (Figure 4e). These 1H and 19F NMR signals are different from those of 1a·(FB)2, prepared separately from 1a and FB (6.11 and −167.5 ppm, respectively; Figure S13). Accordingly, the total FN selectivity of 1a was estimated to be 85% from a 1:1 FN and FB mixture. The minor formation of the uncommon hetero dimer is most probably derived from the steric bulkiness of (FN)2 as compared to that of FN·FB in the closed spherical cavity.18 In contrast, capsule 1a exclusively bound two FN molecules from mixtures of FN and perfluoroanthraquinone, used as an analogue of larger perfluoroarenes (i.e., perfluoroanthracene), as well as FN and perfluorobiphenyl (Figure S14).12

Figure 4.

Figure 4

Open in a new tab

(a) Selective binding of FN from a mixture of FN and FB by capsule 1a (left) or N-doped capsule 1b (right) in water. 1H NMR spectra (500 MHz, D2O, r.t.) of products after mixing FN and FB with (b) 1a, (c) 1b, or (d) 1b (measured at 60 °C; the signal assignment is same as 1a in Figure 2a) at r.t. for 1 h. 19F NMR spectra (470 MHz, D2O, r.t.) of products after mixing FN and FB with (e) 1a, (f) 1b, or (g) 1b (measured at 60 °C) at r.t. for 1 h.

To improve the selectivity, nitrogen-doped Pt(II)-linked capsule 1b (Figure 1e) was applied to the same competitive binding experiments in D2O (Figure 4a, right), due to its adaptable cavity.9 The 1H and 19F NMR spectra of the resultant solution showed relatively broadened, simple host and guest signals at room temperature (Figure 4c,f, respectively), in contrast to those of 1a·(FN)2 and 1a·(FN·FB). These broadened signals became sharp at elevated temperature (e.g., 60 °C; Figure 4d,g), suggesting strong host–guest π-stacking interactions in the cavity. The observed, sharp proton and fluorine signals were fully overlapped with those of 1b·(FN)2 prepared directly from 1b and FN (Figure S15),12 indicating the successful discrimination (100% selectivity) of FN from a mixture of FN and FB by 1b. The host–guest composition of 1b·(FN)2 was also supported by the ESI-TOF MS analysis (Figure S16). These results clarified unusual binding ability of capsules 1a,b toward perfluoroarenes as well as their discrimination order to be FN ≫ FB ≫ FD under ambient aqueous conditions.

Binding Selectivity toward Perfluoroalkanes

We next found the efficient and selective binding ability of Pd(II)-linked polyaromatic capsule 1c (Figure 1e), which provides kinetically labile metal-pyridyl bonds at room temperature,19 toward perfluorinated alkanes under ambient aqueous conditions. Whereas the host ability of Pt(II)/Pd(II)-capsules 1a and 1c has been regarded as the same in our previous studies,9e Pt(II)-capsule 1a displayed no binding ability toward perfluoroalkanes, such as FD and perfluoroheptane (FHp), at room temperature (Figure S26c),12 owing to their rigid and bulky frameworks and kinetically inert metal-pyridyl bonds. In contrast, when a mixture of perfluorohexane (FHx), FHp, and perfluorooctane (FOc; 12 μmol each) was stirred in a D2O solution (0.5 mL) of 1c (0.26 μmol) at room temperature for 1 d, host–guest composite 1c·FHp was formed in a quantitative fashion based on 1c (100% selectivity; Figure 5a). After filtration of the resultant mixture, in the 1H NMR spectrum (Figure 5e), the inner phenylene signal (Ha) of 1c remained nearly unchanged (Δδ = +0.01 ppm) yet the inner pyridyl signal (Hf) of 1c was largely shifted (Δδ = –0.40 ppm; Figure S20), as compared with those of empty 1c. The same product was also exclusively obtained from 1c and FHp under the same conditions (Figure 5f). The ESI-TOF MS analysis of the product indicated the selective formation of 1c·FHp (e.g., m/z = 1296.7 for [1c·FHp – 3·NO3–]3+; Figure S22). Only four fluorine signals (FA–D) were observed in the range of −128.5 to −83.9 ppm in the 19F NMR spectra of 1c·FHp (Figure 5h,i), where the inner signals (FB–D) are selectively shifted by up to −3.7 ppm, relative to those of free FHp in CDCl3 (Figure S21). These characteristic host and guest signals suggested that linear FHp is fully accommodated in the spherical cavity of 1c. Such an accommodation of FHp, with a length of 9.9 Å, within 1c was corroborated by the optimized structure of 1c·FHp (Figure 5c). Due to the rigid perfluorinated frameworks, small length differences between FHp and FHx or FOc (Δd = 1.2–1.4 Å; Figure 5a)11 could be distinguished by the present capsule. The competitive binding experiments using 1c and the corresponding alkanes (i.e., hexane, heptane, and octane), in contrast, gave 1:2 host–guest composites in an unselective fashion (Figure S24).12 The flexibility of nonfluorinated alkyl chains has been found in aqueous host–guest systems.20

Figure 5.

Figure 5

Open in a new tab

Selective binding of (a) FHp from a mixture of FHx, FHp, and FOc as well as (b) FD from a mixture of FHp and FD by capsule 1c in water. Optimized structures of (c) 1c·FHp and (d) 1c·FD (R = H). (e) 1H and (h) 19F NMR spectra (500/470 MHz, D2O, r.t.) of 1c·FHp. (f) 1H and (i) 19F NMR spectra (500/470 MHz, D2O, r.t.) of products after mixing FHx, FHp, and FOc with 1c. (g) 1H and (j) 19F NMR spectra (500/470 MHz, D2O, r.t.) of products after mixing FHp and FD with 1c, and (k) its ESI-TOF MS spectra (H2O) and the expanded and simulated signals of [1c·FD – 4·NO3–]4+.

Moreover, the discrimination between linear and cyclic perfluoroalkanes was demonstrated by capsule 1c. Stirring a mixture of FHp and FD (50 equiv. each) with 1c in water at room temperature led to the quantitative formation of 1c·FD based on 1c in absolute selectivity (Figure 5b).14,21 The 1H and 19F NMR spectra of the host–guest product after simple filtration (Figure 5g,j) were consistent with those of 1c·FD, prepared quantitatively from 1c and FD in water (Figure S25a).12 The ESI-TOF MS spectrum of the product also indicated the selective formation of 1c·FD (Figure 5k). Observed prominent molecular ion peaks at m/z = 975.6 (n = 4) and 1321.4 (n = 3) were assignable to the [1c·FD – n·NO3–]n+ species. The optimized structure of 1c·FD supported that the cyclic structure of FD is complementary in shape to the spherical structure of the capsule cavity (Figure 5d).18,22 The present competitive studies therefore elucidated the binding preference of 1c toward perfluoroalkanes to be FD ≫ FHp ≫ FHx.

Binding Selectivity toward Oligofluoro and Nonperfluoro Compounds

As final competitive studies, we revealed the binding selectivities of capsules 1a and 1c toward perfluorobiphenyls with/without substituted hydrogen atoms as well as aromatic and alkyl compounds with/without perfluorination, respectively. Whereas the diameters and volumes of decafluorobiphenyl (DFB), 2,2′,3,3′,4,5,5′,6,6′-nonafluorobiphenyl (NFB), and 2,2′,3,3′,5,5′,6,6′-octafluorobiphenyl (OFB) are nearly comparable (i.e., Δd = 0.2 Å and ΔV = 11 Å3),11 the recognition of such small differences could be performed in the polyaromatic cavity of 1a. From a mixture of DFB, NFB, and OFB (5.0 μmol each) in a D2O solution of 1a (5.0 μmol) at 80 °C, the exclusive binding of OFB by 1a within 1 h was revealed by NMR and MS analyses (Figures 6a and S27a,b).12 After removal of free per/oligofluoro compounds by filtration, the 1H NMR spectrum of the resultant mixture showed a new single signal at 2.81 ppm (Figure S27a), derived from bound OFB in the cavity, besides the host aromatic signals (Ha–i; Figure 6d). The signal was shifted highly upfield (Δδ = −4.39 ppm) as compared with that of free OFB in CDCl3, due to the aromatic shielding effect. These signals were consistent with those of separately prepared 1a·OFB and no proton signals for empty 1a, 1a·DFB, and 1a·NFB were observed in the spectrum. The high selectivity was also indicated by the 19F NMR and ESI-TOF MS analyses, in which two signals were found at −146.7 and −139.2 ppm for bound OFB (Figure 6e) and two peaks were detected at m/z = 978.8 and 1326.8 for [1a·OFB – n·NO3–]n+ (n = 4, 3; Figure S27b), respectively.12 The observed, uncommon selectivity toward OFB is most probably attributed to effective host–guest CH-π interactions, rather than the size complementarity, which are supported by its large proton signal shift.

Figure 6.

Figure 6

Open in a new tab

Selective binding of (a) OFB from a mixture of DFB, NFB, and OFB as well as (b) HN from a mixture of FN and HN by capsule 1a in water. (c) Preferential binding of FD from a mixture of FD and HD by 1c in water. 1H and 19F NMR spectra (500/470 MHz, D2O, r.t.) of products after mixing (d,e) DFB, NFB, and OFB with 1a at 80 °C and (f,g) FD and HD with 1c at r.t. for 1 h, respectively.

As expected, capsule 1a bound naphthalene (HN) over FN in water at room temperature for 1 h (Figure 6b). The formation of 1:2 host–guest composite 1a·(HN)2 in a quantitative fashion based on 1a was confirmed by NMR (Figure S28a) and MS studies.12 The result is explainable by attractive host–guest and guest–guest π-π interactions as well as host–guest CH-π interactions in the cavity.9,23 These multiple interactions cannot be generated within 1a·(FN)2, whereas the hydrophobicity of FN is higher than that of HN (Figure S30). A similar NMR study also revealed the quantitative formation of 1c·HHp from a mixture of 1c, FHp, and heptane (HHp), due to host–guest CH-π interactions (Figure S28b).12 In contrast, as unexpected, Pd(II)-linked capsule 1c bound FD over nonperfluorinated trans-decaline (HD) under similar conditions to give 1c·FD and 1c·HD in a 89:11 ratio (Figure 6c). The host–guest products, without remaining empty 1c, were confirmed by 1H and 19F NMR analyses (Figure 6f,g).24

Theoretical Studies on Interactions with Perfluorocarbons

Finally, the observed, unusual perfluorophilicity in the polyaromatic cavity was studied with theoretical calculation. The noncovalent interaction (NCI) plots25,26 of 1a·(FN)2, generated from its crystal structure, displayed effective π-π interactions between the distal anthracene panel and FN, FN and FN, and FN and the distal anthracene panel in the cavity (Figures 7a and S33).12 The interactions between FN and FN spread throughout the aromatic frameworks. These multiple π-π interactions most likely allow capsules 1a,b to bind FN over FD and FB in a quantitative and selective fashion. The optimized structure (DFT) of 1c·FD indicated that spheroidal FD (trans-isomer; ∼250 Å3) exists nearly in the center of the cavity with multiple host–guest interactions (Figure 7b). The focused NCI plots clarified the presence of attractive CF···π interactions between FD and the eight anthracene panels of 1c (3.2–3.4 Å; Figures 7c and S34). In addition, effective CH···F hydrogen-bonding interactions with the eight pyridyl α-hydrogens of 1c (3.2–3.4 Å) were found in the cavity, besides Pd(II)···F interactions (3.2 Å; Figure 7d).27,28 The same plots of 1c·HD (trans-isomer) suggested the existence of six CH-π (anthracene) interactions between HD and 1c (3.2–3.4 Å, Figure S35), owing to its one-sided accommodation in the cavity, without hydrogen-bonding interactions. Natural energy decomposition analysis also suggested that host–guest interactions of 1c with FD are stronger than those of 1c with HD in the cavity (ΔE = –0.3 kJ/mol; Figure S32). Therefore, the present unique selectivity toward FD over HD most probably stems from multiple host–guest CH···F interactions, which have never been demonstrated with previous synthetic hosts,3,6,7 in the confined polyaromatic cavity of 1c.

Figure 7.

Figure 7

Open in a new tab

Selected NCI plots (isosurface value 0.5) of (a) 1a·(FN)2, and (b) 1c·FD and (c,d) its focused structures (R = R′ = H),12 displaying attractive interaction regions highlighted in yellow.

Conclusions

In this study, the highly efficient and selective binding of perfluorocarbons has been accomplished by modulable polyaromatic capsules under ambient aqueous conditions. A Pt(II)-linked polyaromatic capsule bound two molecules of perfluoroarene (i.e., perfluoronaphthalene) from a mixture with the corresponding perfluorinated alkane in an exclusive fashion (quantitative selectivity). The crystallographic analysis of the host–guest composite indicated the generation of effective D-A-A-D stacking interactions in the well-defined, spherical cavity. The size-selective binding toward larger perfluoroarenes (e.g., perfluorinated naphthalene versus benzene) was demonstrated by the capsule in high selectivity as well as by the analogous interior-N-doped capsule in perfect selectivity, due to the adaptable cavity. Whereas the Pt(II)-capsule provides no binding ability toward perfluoroalkanes, an isostructural Pd(II)-linked capsule, with more flexible frameworks, bound perfluorinated linear/cyclic alkanes under ambient conditions. Notably, from a complex mixture of linear perfluoroalkanes (i.e., perfluorinated hexane, heptane, and octane), the Pd(II)-capsule quantitatively bound perfluoroheptane in perfect selectivity. A cyclic perfluoroalkane was bound by the capsule over the linear ones. The present studies revealed for the first time that (i) polyaromatic cavities are useful tools to bind perfluorocarbons with high efficiency and selectivity as well as (ii) minor cavity modulation is an effective strategy to switch the binding selectivity.

The binding preference toward arenes and alkanes is usually higher than the corresponding perfluorinated ones, owing to the highly repulsive nature of fluorine. Nevertheless, the present polyaromatic capsule exceptionally bound one molecule of perfluorodecaline over decaline under ambient conditions (i.e., ∼90% selectivity). This unique result indicates that, in perspective, further rational host designs could create new molecular tools to exclusively bind various perfluorocarbons with/without functional groups, even from further complex mixtures in water.

Acknowledgments

This work was supported by JSPS KAKENHI (grant no. JP22H00348/JP23K17913). Synchrotron X-ray diffraction data was obtained by Dr. Masahiro Yamashina (Science Tokyo) under the approval of the Photon Factory Program Advisory Committee (Proposal 2024G540). Theoretical calculations were performed using computers at the Research Center for Computational Science, Okazaki, Japan (23-IMS-C063, 24-IMS-C060). R.S. thanks the JSPS for a Research Fellowship for Young Scientists.

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.5c00904.

  • Experimental procedures, and NMR, MS, X-ray, and calculation data (PDF)

The authors declare no competing financial interest.

Supplementary Material

ja5c00904_si_001.pdf (17.7MB, pdf)

References

  1. a Kirsch P.Modern Fluoroorganic Chemistry, Synthesis, Reactivity, Applications; Wiley VCH: Germany, 2004. [Google Scholar]; b Reichenbächer K.; Süssa H. I.; Hulliger J. Fluorine in Crystal Engineering– “the Little Atom that Could”. Chem. Soc. Rev. 2005, 34, 22–30. 10.1039/B406892K. [DOI] [PubMed] [Google Scholar]
  2. Améduri B., Ed.; RSC: UK, 2022; .Perfluoroalkyl Substances, Synthesis, Applications, Challenges and Regulations [Google Scholar]
  3. Recent reviews:; a Brown C. J.; Toste F. D.; Bergman R. G.; Raymond K. N. Supramolecular Catalysis in Metal–Ligand Cluster Hosts. Chem. Rev. 2015, 115, 3012–3035. 10.1021/cr4001226. [DOI] [PubMed] [Google Scholar]; b Murray J.; Kim K.; Ogoshi T.; Yao W.; Gibb B. C. The Aqueous Supramolecular Chemistry of Cucurbit[n]urils, Pillar[n]arenes and Deep-Cavity Cavitands. Chem. Soc. Rev. 2017, 46, 2479–2496. 10.1039/C7CS00095B. [DOI] [PMC free article] [PubMed] [Google Scholar]; c Grommet A. B.; Feller M.; Klajn R. Chemical Reactivity under Nanoconfinement. Nat. Nanotechnol. 2020, 15, 256–271. 10.1038/s41565-020-0652-2. [DOI] [PubMed] [Google Scholar]; d Percastegui E. G.; Ronson T. K.; Nitschke J. R. Design and Applications of Water-Soluble Coordination Cages. Chem. Rev. 2020, 120, 13480–13544. 10.1021/acs.chemrev.0c00672. [DOI] [PMC free article] [PubMed] [Google Scholar]; e Liu W.; Stoddart J. F. Emergent Behavior in Nanoconfined Molecular Containers. Chem. 2021, 7, 919–947. 10.1016/j.chempr.2021.02.016. [DOI] [Google Scholar]; f Escobar L.; Ballester P. Molecular Recognition in Water Using Macrocyclic Synthetic Receptors. Chem. Rev. 2021, 121, 2445–2514. 10.1021/acs.chemrev.0c00522. [DOI] [PubMed] [Google Scholar]; g Takezawa H.; Fujita M. Molecular Confinement Effects by Self-Assembled Coordination Cages. Bull. Chem. Soc. Jpn. 2021, 94, 2351–2369. 10.1246/bcsj.20210273. [DOI] [Google Scholar]; h McTernan C. T.; Davies J. A.; Nitschke J. R. Beyond Platonic: How to Build Metal-Organic Polyhedra Capable of Binding Low-Symmetry, Information-Rich Molecular Cargoes. Chem. Rev. 2022, 122, 10393–10437. 10.1021/acs.chemrev.1c00763. [DOI] [PMC free article] [PubMed] [Google Scholar]; i Moree L. K.; Faulkner L. A. V.; Crowley J. D. Heterometallic Cages: Synthesis and Applications. Chem. Soc. Rev. 2025, 53, 25–46. 10.1039/D3CS00690E. [DOI] [PubMed] [Google Scholar]
  4. Host functions of M2L4 cages and capsules (reviews):; a Debata N. B.; Tripathy D.; Chand D. K. Self-Assembled Coordination Complexes from Various Palladium(II) Components and Bidentate or Polydentate Ligands. Coord. Chem. Rev. 2012, 256, 1831–1945. 10.1016/j.ccr.2012.04.001. [DOI] [Google Scholar]; b Han M.; Engelhard D. M.; Clever G. H. Self-Assembled Coordination Cages Based on Banana-Shaped Ligands. Chem. Soc. Rev. 2014, 43, 1848–1860. 10.1039/C3CS60473J. [DOI] [PubMed] [Google Scholar]; c Pullen S.; Tessarolo J.; Clever G. H. Increasing Structural and Functional Complexity in Self-Assembled Coordination Cages. Chem. Sci. 2021, 12, 7269–7293. 10.1039/D1SC01226F. [DOI] [PMC free article] [PubMed] [Google Scholar]; d Lewis J. E. M. Molecular Engineering of Confined Space in Metal-Organic Cages. Chem. Commun. 2022, 58, 13873–13886. 10.1039/D2CC05560K. [DOI] [PubMed] [Google Scholar]; e Catti L.; Sumida R.; Yoshizawa M. Aqueous Polyaromatic Receptors for Biomolecules with High Selectivity. Coord. Chem. Rev. 2022, 460, 214460. 10.1016/j.ccr.2022.214460. [DOI] [Google Scholar]
  5. a Druliner J. D.; Wasserman E. Synthesis and Characterization of Cyclodextrin/Perfluoroalkane Inclusion Compounds. J. Fluorine Chem. 1995, 72, 75–78. 10.1016/0022-1139(94)03179-4. [DOI] [Google Scholar]; b Tatsuno H.; Ando S. Structure and Dynamics of Perfluoroalkane/β-Cyclodextrin Inclusion Compounds As Studied by Solid-State 19F MAS and 1H → 19F CP/MAS NMR Spectroscopy. J. Phys. Chem. B 2006, 110, 25751–25760. 10.1021/jp064579k. [DOI] [PubMed] [Google Scholar]; c Raju R. K.; Hillier I. H.; Burton N. A.; Vincent M. A.; Doudou S.; Bryce R. A. The Effects of Perfluorination on Carbohydrate-π Interactions: Computational Studies of the Interaction of Benzene and Hexafluorobenzene with Fucose and Cyclodextrin. Phys. Chem. Chem. Phys. 2010, 12, 7959–7967. 10.1039/c002058c. [DOI] [PubMed] [Google Scholar]
  6. Purse B. W.; Rebek J. Jr Encapsulation of Oligoethylene Glycols and Perfluoro-n-alkanes in a Cylindrical Host Molecule. Chem. Commun. 2005, 722–724. 10.1039/b408878f. [DOI] [PubMed] [Google Scholar]
  7. Takezawa H.; Murase T.; Resnati G.; Metrangolo P.; Fujita M. Recognition of Polyfluorinated Compounds Through Self-Aggregation in a Cavity. J. Am. Chem. Soc. 2014, 136, 1786–1788. 10.1021/ja412893c. [DOI] [PubMed] [Google Scholar]
  8. Intensive studies on the binding of PFCs with a functional group (e.g., PFOA and PFOS):; a Lai Z.; Zhou Y.; Bai S.; Sun Q. Opportunity and Challenge of Advanced Porous Sorbents for PFAS Removal. ChemSusChem 2025, 18, e202401229 10.1002/cssc.202401229. [DOI] [PubMed] [Google Scholar]; b Li Y.-Q.; Zhao H.; Han E.; Jiang Z.; Bai Q.; Guan Y.-M.; Zhang Z.; Wu T.; Wang P. Dynamic Selection in Metallo-organic Cube CdII8L4 Conformations Induced by Perfluorooctanoate Encapsulation. Chem. Sci. 2025, 16, 364–370. 10.1039/D4SC07105K. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. a Kishi N.; Li Z.; Yoza K.; Akita M.; Yoshizawa M. An M2L4 Molecular Capsule with an Anthracene Shell: Encapsulation of Large Guests up to 1 nm. J. Am. Chem. Soc. 2011, 133, 11438–11441. 10.1021/ja2037029. [DOI] [PubMed] [Google Scholar]; b Yamashina M.; Sei Y.; Akita M.; Yoshizawa M. Safe Storage of Radical Initiators within a Polyaromatic Nanocapsule. Nat. Commun. 2014, 5, 4662. 10.1038/ncomms5662. [DOI] [PubMed] [Google Scholar]; c Yamashina M.; Sartin M.; Sei Y.; Akita M.; Takeuchi S.; Tahara T.; Yoshizawa M. Preparation of Highly Fluorescent Host-Guest Complexes with Tunable Color upon Encapsulation. J. Am. Chem. Soc. 2015, 137, 9266–9269. 10.1021/jacs.5b06195. [DOI] [PubMed] [Google Scholar]; d Yamashina M.; Matsuno S.; Sei Y.; Akita M.; Yoshizawa M. Recognition of Multiple Methyl Groups on Aromatic Rings by a Polyaromatic Cavity. Chem.–Eur. J. 2016, 22, 14147–14150. 10.1002/chem.201603032. [DOI] [PubMed] [Google Scholar]; e Dobashi H.; Catti L.; Tanaka Y.; Akita M.; Yoshizawa M. N-Doping of Polyaromatic Capsules: Small Cavity Modification Leads to Large Change in Host-Guest Interactions. Angew. Chem., Int. Ed. 2020, 59, 11881–11885. 10.1002/anie.202004168. [DOI] [PubMed] [Google Scholar]
  10. a Yoshizawa M.; Yamashina M. Coordination-Driven Nanostructures with Polyaromatic Shells. Chem. Lett. 2017, 46, 163–171. 10.1246/cl.160852. [DOI] [Google Scholar]; b Kondo K.; Klosterman J. K.; Yoshizawa M. Aromatic Micelles as a New Class of Aqueous Molecular Flasks. Chem.–Eur. J. 2017, 23, 16710–16721. 10.1002/chem.201702519. [DOI] [PubMed] [Google Scholar]; c Yazaki K.; Catti L.; Yoshizawa M. Polyaromatic Molecular Tubes: from Strategic Synthesis to Host Functions. Chem. Commun. 2018, 54, 3195–3206. 10.1039/C8CC00799C. [DOI] [PubMed] [Google Scholar]; d Yoshizawa M.; Catti L. Bent Anthracene Dimers as Versatile Building Blocks for Supramolecular Capsules. Acc. Chem. Res. 2019, 52, 2392–2404. 10.1021/acs.accounts.9b00301. [DOI] [PubMed] [Google Scholar]; e Kishida N.; Matsumoto K.; Tanaka Y.; Akita M.; Sakurai H.; Yoshizawa M. Anisotropic Contraction of a Polyaromatic Capsule and its Cavity-induced Compression Effect. J. Am. Chem. Soc. 2020, 142, 9599–9603. 10.1021/jacs.0c02932. [DOI] [PubMed] [Google Scholar]; f Tsutsui T.; Catti L.; Yoza K.; Yoshizawa M. An Atropisomeric M2L4 Cage Mixture Displaying Guest-Induced Convergence and Strong Guest Emission in Water. Chem. Sci. 2020, 11, 8145–8150. 10.1039/D0SC03223A. [DOI] [PMC free article] [PubMed] [Google Scholar]; g Sasafuchi H.; Ueda M.; Kishida N.; Sawada T.; Suzuki S.; Imai Y.; Yoshizawa M. Remote Optical Chirality Transfer via Helical Polyaromatic Capsules upon Encapsulation. Chem 2025, 11, 102332. 10.1016/j.chempr.2024.09.031. [DOI] [Google Scholar]
  11. Sizes and volumes of perfluorocarbons and other were estimated by DFT calculation (B3LYP/6–31G*, Spartan 18 software (version 1.4.5)). FN: 168 Å3, FD: 251 Å3, FHx: 193 Å3, FHp: 228 Å3, FOc: 269 Å3, DFB: 216 Å3, OFB: 205 Å3. The cavity volume of 1a has been determined based on the crystal structure9 with VOIDOO calculation:Kleywegt G. J.; Jones T. A. Detection, Delineation, Measurement and Display of Cavities in Macromolecular Structures. Acta Crystallogr. Sect. D 1994, 50, 178–185. 10.1107/S0907444993011333. [DOI] [PubMed] [Google Scholar]
  12. See the Supporting Information.
  13. The supplementary crystallographic data (CCDC 2403669 and 2429529) can be obtained free of charge from the Cambridge Crystallographic Data Centre. Whereas a couple of the hydrophilic side chains of 1a were severely disordered, the encapsulated (FN)2 structure could be clearly assigned in the cavity of 1a (Figure S31). The crystallographic analysis of 1c•FD also indicated multiple host–guest CF•••π and CH•••F interactions in the cavity (Figure S36).
  14. After removal of excess free FN and FD by filtration, the bound FN molecules could be fully extracted from the resultant, aqueous 1a•(FN)2 solution with organic solvent (e.g., diethyl ether), through a typical extraction procedure, as revealed by 1H/19F NMR and GC MS analyses (Figure S7c). 1c•FD was exclusively obtained from a 6.0 equiv. mixture of bulkier FD and FHp (each; Figure S26a). The bound FD was also extracted from 1c•FD in a similar way (Figure S26d).
  15. The 1H and 19F NMR spectra of 1a•(FN)2 and 1c•FD in D2O (0.5 mM based on 1a/1c) remained intact even under high dilution conditions (i.e., 0.05 and 0.005 mM; Figure S2c and S26b).12 The repetitive selective binding and quantitative extraction of FN could be performed at least three times using 1a and a mixture of FN and FD (Figure S7e).12
  16. Pausch T.; Clopot S.; Jordan D. N.; Weingart O.; Janiak C.; Schmidt B. M. Fluorinated Squareimines for Molecular Sieving of Aromatic over Aliphatic Compounds. Angew. Chem. Int. Ed 2024, 63, e202418877 10.1002/anie.202418877. [DOI] [PubMed] [Google Scholar]
  17. a Yamashina M.; Kusaba S.; Akita M.; Kikuchi T.; Yoshizawa M. Cramming versus Threading of Long Amphiphilic Oligomers into a Polyaromatic Capsule. Nat. Commun. 2018, 9, 4227. 10.1038/s41467-018-06458-w. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Shuto M.; Sumida R.; Yuasa M.; Sawada T.; Yoshizawa M. A Closed Cavity Strategy for Selective Dipeptide Binding by a Polyaromatic Receptor in Water. JACS Au 2023, 3, 2905–2911. 10.1021/jacsau.3c00484. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Sumida R.; Tanaka Y.; Niki K.; Sei Y.; Toyota S.; Yoshizawa M. Cyclic Monoterpenes Trapped in a Polyaromatic Capsule: Unusual Selectivity, Isomerization, and Volatility Suppression. Chem. Sci. 2021, 12, 9946–9951. 10.1039/D1SC01987B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fujita M.; Ibukuro F.; Yamaguchi K.; Ogura K. A Molecular Lock. J. Am. Chem. Soc. 1995, 117, 4175–4176. 10.1021/ja00119a036. [DOI] [Google Scholar]
  20. a Liu S.; Russell D. H.; Zinnel N. F.; Gibb B. C. Guest Packing Motifs within a Supramolecular Nanocapsule and a Covalent Analogue. J. Am. Chem. Soc. 2013, 135, 4314–4324. 10.1021/ja310741q. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Dumitrescu D.; Legrand Y.-M.; Petit E.; van der Lee A.; Barboiu M. Adaptive Binding and Selection of Compressed 1,ω-Diammonium-alkanes via Molecular Encapsulation in Water. Chem. Sci. 2015, 6, 2079–2086. 10.1039/C4SC03945A. [DOI] [PMC free article] [PubMed] [Google Scholar]; c Niki K.; Tsutsui T.; Yamashina M.; Akita M.; Yoshizawa M. Recognition and Stabilization of Unsaturated Fatty Acids by a Polyaromatic Receptor. Angew. Chem., Int. Ed. 2020, 59, 10489–10492. 10.1002/anie.202003253. [DOI] [PubMed] [Google Scholar]
  21. The relative binding constants of 1a toward FB and 1c toward FD were estimated to be ∼1 × 106 and 5.7 × 105 M–1, by competitive binding experiments (Figure S8b,c and Table S1b), respectively.12,15
  22. As a preliminary study, the treatment of a mixture of FD and perfluoroadamantane (FAd) including perfluoro(1-methyladamantane) (6.0 equiv. each) with 1c in D2O gave 1c•FD in quantitative yield under the same conditions (Figure S26e).12
  23. Selective binding of nonperfluorinated arenes (recent works):; a Zhang D.; Ronson T. K.; Lavendomme R.; Nitschke J. R. Selective Separation of Polyaromatic Hydrocarbons by Phase Transfer of Coordination Cages. J. Am. Chem. Soc. 2019, 141, 18949–18953. 10.1021/jacs.9b10741. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Sainaba A. B.; Venkateswarulu M.; Bhandari P.; Arachchige K. S. A.; Clegg J.; Mukherjee P. S. An Adaptable Water-Soluble Molecular Boat for Selective Separation of Phenanthrene from Isomeric Anthracene. J. Am. Chem. Soc. 2022, 144, 7504–7513. 10.1021/jacs.2c02540. [DOI] [PubMed] [Google Scholar]; c Kishida N.; Tanaka Y.; Yoshizawa M. CH-π Multi-Interaction-Driven Recognition and Isolation of Planar Compounds in a Spheroidal Polyaromatic Cavity. Chem.–Eur. J. 2022, 28, e202202075 10.1002/chem.202202075. [DOI] [PMC free article] [PubMed] [Google Scholar]; d Prajapati D.; Bhandari P.; Hickey N.; Mukherjee P. S. Water-Soluble Pd6L3 Molecular Bowl for Separation of Phenanthrene from a Mixture of Isomeric Aromatic Hydrocarbons. Inorg. Chem. 2023, 62, 9230–9239. 10.1021/acs.inorgchem.3c01156. [DOI] [PubMed] [Google Scholar]
  24. Selective binding of nonperfluorinated alkanes (recent works):; a Ogoshi T.; Shimada Y.; Sakata Y.; Akine S.; Yamagishi T. Alkane-Shape-Selective Vapochromic Behavior Based on Crystal-State Host-Guest Complexation of Pillar[5]arene Containing One Benzoquinone Unit. J. Am. Chem. Soc. 2017, 139, 5664–5667. 10.1021/jacs.7b00631. [DOI] [PubMed] [Google Scholar]; b Ogoshi T.; Saito K.; Sueto R.; Kojima R.; Hamada Y.; Akine S.; Moeljadi A. M. P.; Hirao H.; Kakuta T.; Yamagishi T. Separation of Linear and Branched Alkanes Using Host-Guest Complexation of Cyclic and Branched Alkane Vapors by Crystal State Pillar[6]arene. Angew. Chem., Int. Ed. 2018, 57, 1592–1595. 10.1002/anie.201711575. [DOI] [PubMed] [Google Scholar]
  25. The (NCI) plots was obtained by the DFT calculation (CAM-B3LYP+D3BJ/LanL2DZ (Pd), 6–31G(d,p) (others) level of theory).
  26. a Johnson E. R.; Keinan S.; Mori-Sánchez P.; Contreras-García J.; Cohen A. J.; Yang W. Revealing Noncovalent Interactions. J. Am. Chem. Soc. 2010, 132, 6498–6506. 10.1021/ja100936w. [DOI] [PMC free article] [PubMed] [Google Scholar]; b Contreras-García J.; Johnson E. R.; Keinan S.; Chaudret R.; Piquemal J.-P.; Beratan D. N.; Yang W. NCIPLOT: A Program for Plotting Noncovalent Interaction Regions. J. Chem. Theory Comput. 2011, 7, 625–632. 10.1021/ct100641a. [DOI] [PMC free article] [PubMed] [Google Scholar]; https://www.lct.jussieu.fr/pagesperso/contrera/nciplot.html.
  27. Stanek K.; Czarniecki B.; Aardoom R.; Rüegger H.; Togni A. Remote C-F-Metal Interactions in Late-Transition-Metal Complexes. Organometallics 2010, 29, 2540–2546. 10.1021/om100261e. [DOI] [Google Scholar]
  28. Rosenberg R. E. The Strength of Hydrogen Bonds between Fluoro-Organics and Alcohols, a Theoretical Study. J. Phys. Chem. A 2018, 122, 4521–4529. 10.1021/acs.jpca.8b01148. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ja5c00904_si_001.pdf (17.7MB, pdf)

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.

Articles from Journal of the American Chemical Society are provided here courtesy of American Chemical Society

RESOURCES