Decorating polymer beads with 10¹⁴ inorganic-organic [2]rotaxanes as shown by spin counting

The study explores whether polymetallic compounds can be covalently attached to polystyrene beads at very high loading, creating highly paramagnetic microparticles. Prior work established polystyrene beads as cores for inorganic shells (e.g., magnetite, ZnS, MOFs) and as supports for paramagnetic species used in catalysis and dynamic nuclear polarization (DNP). The authors propose two potential directions enabled by such attachment: (i) decomposition of the attached polymetallic units to yield core–shell particles with controlled metal shells; and (ii) creation of beads densely decorated with highly paramagnetic molecules for applications such as DNP, contrast agents, or ferrofluids. The immediate research objective is to demonstrate the feasibility and quantify the extent of attaching hybrid inorganic–organic [2]rotaxanes containing {Cr–Ni} rings to azide-functionalized polystyrene beads using CuAAC click chemistry, and to count the number of attached species by EPR spin counting.

• Caruso et al. (1999) demonstrated magnetite growth on polystyrene via polyelectrolyte coating, establishing polymer beads as cores for magnetic particles. Similar core–shell structures have been prepared with ZnS for photonic crystals and with MOF coatings for pollutant extraction.
• Beads functionalized with paramagnetic molecules have been used in catalysis and as DNP agents; magnetic polymer beads find uses in sensing, imaging, and separations. Radical functionalization has probed bead macromolecular structure and swelling.
• CuAAC click chemistry is a robust method for linking alkyne and azide functionalities. Heterometallic {Cr7Ni} rings have been well studied as supramolecular building blocks and incorporated into hybrid inorganic–organic rotaxanes. The earliest [2]rotaxane synthesis involved polymer-supported macrocycles, highlighting the relevance of solid-supported rotaxane chemistry.

• Design and synthesis of axle/template: A secondary ammonium thread, 1, (HOC6H4CH2CH2)2NH was prepared by coupling a primary amine and an alkyl bromide each bearing terminal phenols. The ammonium distance to phenols was tuned to allow selective functionalization while minimizing esterification during ring formation.
• Formation of hybrid inorganic–organic [2]rotaxane: Thread 1 was reacted with CrF3·4H2O and basic Ni carbonate (2NiCO3·3Ni(OH)2·4H2O) in pivalic acid at 140 °C for 24 h to yield rotaxane 2, comprising an eight-metal {Cr7Ni} ring templated around the ammonium center, capped by sixteen pivalate groups. Compound 2 was purified chromatographically and structurally confirmed by single-crystal X-ray diffraction (monoclinic P21/c).
• Introduction of click handles and spin label: Rotaxane 2 was mono-esterified with 4-pentynoic acid (Steglich conditions: DCC/DMAP in THF, 50 °C) to give alkyne-bearing 3, then further esterified with 4-carboxy-TEMPO (DCC/DMAP in CH2Cl2, RT) to yield spin-labeled rotaxane 4. Rotaxane 4 was characterized by ESI-MS, elemental analysis, and single-crystal XRD (monoclinic C/2c).
• Bead functionalization and CuAAC attachment: Commercial polystyrene microspheres were azide-functionalized following literature procedures, producing larger beads 5 (115 ± 35 µm) and smaller beads 6 (10 ± 1 µm). Heterogeneous CuAAC reactions were performed between bead azides and alkyne-terminated rotaxanes using [Cu(MeCN)4]PF6 in CH2Cl2:DMF (3:2), under argon at RT for 72 h, producing decorated beads 7 (from 5 + 4) and 8 (from 6 + 4). Beads were extensively washed with CH3CN, CH3OH, and CH2Cl2 to remove unreacted rotaxane and catalyst.
• Reaction monitoring and accessible site estimation: FT-IR spectroscopy monitored the azide band (~2090 cm−1), which decreased significantly post-reaction; residual azide signal was attributed to inaccessible interior sites. The number of solvent-accessible azide sites was estimated assuming spherical beads, uniform site density, partial swelling allowing reaction to a depth of 0.5 µm, and quantitative azide formation and click reactions. Lower-bound estimates: ≥2 × 10^13 accessible sites for 115 µm beads (5) and ~8 × 10^10 for 10 µm beads (6); upper bounds if all sites reacted: 7 × 10^14 (5) and 3 × 10^11 (6).
• Structural metrics from XRD: In 2 and 4, the {Cr7Ni} rings form octagons with disordered divalent site, carboxylate bridges outside the ring, and inner fluoride bridges forming H-bonds to the protonated axle nitrogen. The alkyne terminus to central ammonium N distance is ~1.36 nm, enabling CuAAC.
• EPR spectroscopy and spin counting: CW Q-band (~34 GHz) EPR spectra were recorded at 5 K and 280 K on 4 (powder, solution calibrants) and beads 7 and 8. The {Cr7Ni} S = 1/2 ground-state resonance (~g = 1.78) appears at low T and is unchanged upon bead attachment. Nitroxide resonances are observed at all temperatures; on beads, signals are sharper due to dilution on the surface. Calibration curves were constructed from 4 in toluene at RT using concentrations 0.005–0.06 mM (a discontinuity was noted around 0.07 mM and higher, so excluded). Spin counting was performed by measuring integrated EPR signal from increasing numbers of beads (1–15) and mapping intensities onto the calibration.
• Magnetometry: Magnetic measurements (Quantum Design MPMS3) were performed on azide-only beads 5, decorated beads 7 (TEMPO-bearing) and 11 (non-TEMPO rotaxane 10 clicked onto 5), and on molecular reference 12 ([Pr2NH2][Cr7NiF8(O2C’Bu)16]). Susceptibility: 300–2 K at 1000 Oe; magnetization: 2 K to 7 T. Data were scaled per bead using bead counts per mg (e.g., ~831 beads/mg for 7).

• Successful covalent decoration of azide-functionalized polystyrene beads (115 ± 35 µm) with hybrid inorganic–organic [2]rotaxanes containing {Cr7Ni} rings via CuAAC, yielding green beads 7; analogous decoration achieved on 10 ± 1 µm beads (8).
• FT-IR shows a strong decrease of the azide band (~2090 cm−1) post-click, consistent with reaction at accessible sites; residual azide attributed to inaccessible interior regions.
• EPR spectroscopy: The {Cr7Ni} S = 1/2 signal at g ≈ 1.78 appears only at low temperature and is unperturbed upon bead attachment. Nitroxide signals are present at all temperatures; on beads they are sharper with partial 14N hyperfine resolution, indicating surface dilution.
• Spin counting (Q-band, RT) on 115 µm beads (7) indicates on the order of 10^14 nitroxide spins per bead. From Table 2, multi-bead measurements give averages ~0.9–1.1 × 10^14 spins per bead; the overall conclusion is ~10^14 spins/bead. This aligns within an order of magnitude with the lower-bound estimate from accessible sites (≥2 × 10^13) and below the upper bound (7 × 10^14).
• For 10 µm beads (8), bulk measurements yield ~7.2 × 10^10 spins per bead, close to the estimated accessible-site capacity (~8 × 10^10) and within an order of magnitude of the upper bound (3 × 10^11).
• Magnetic measurements per bead show 7 and 11 have susceptibilities at high T matching those expected for ~10^14 molecules of the {Cr7Ni} ring reference (12). At low T, 7 matches ~10^14 molecules of 12, while 11 fits better with ~3 × 10^14 molecules, suggesting similar magnitude loading; differences may indicate weak interactions but are inconclusive.
• X-ray crystallography confirms the structures of rotaxanes 2, 4, and 10, with the alkyne handle suitably positioned (~1.36 nm from the ring center N) for click coupling.

The work demonstrates the feasibility of densely attaching polymetallic hybrid inorganic–organic [2]rotaxanes to polymer beads through CuAAC click chemistry, directly addressing the central question of whether such high-loadings of polymetallic complexes can be achieved on polymer supports. EPR spin counting of the TEMPO labels provides a quantitative measure of loading, showing ~10^14 rotaxanes per 115 µm bead and ~7.2 × 10^10 per 10 µm bead. These quantitative results are consistent with estimates based on solvent-accessible azide site densities and bead swelling assumptions. The preservation of the {Cr7Ni} EPR signature upon attachment indicates that the polymetallic rings remain intact on the beads. Magnetometry further corroborates the loading levels, with per-bead susceptibilities consistent with on the order of 10^14 rings. Collectively, these results validate the approach and suggest that such decorated beads could form the basis for novel paramagnetic microparticles, with potential relevance to DNP, imaging contrast agents, separations, or as precursors to metal-coated core–shell particles after decomposition of the attached complexes.

Highly paramagnetic microparticles have been prepared by decorating azide-functionalized polystyrene beads with hybrid inorganic–organic [2]rotaxanes using CuAAC click chemistry. Quantification by EPR spin counting and corroborating magnetometry indicates ~10^14 rotaxanes per 115 µm bead, in good agreement with estimates of accessible surface sites. The chemistry is general and flexible (applicable to different bead sizes and rotaxane variants) and preserves the integrity of the {Cr7Ni} rings. This platform opens routes to unconventional nanostructures and microparticles with high densities of paramagnetic centers. Future work will explore applications such as dynamic nuclear polarization, contrast agents, ferrofluids, and conversion of attached complexes into controlled metal shells for core–shell particles.

• EPR spin counting approached the detection limit (~10^12 spins/mT linewidth), leading to limited accuracy at low signal levels. A calibration discontinuity around ~0.07 mM necessitated restricting calibrant concentrations to 0.005–0.06 mM.
• Sample positioning sensitivity for measurements with very few beads; bead size distribution (80–150 µm) introduces variability in per-bead loading.
• Estimation of accessible sites relies on assumptions: spherical beads, uniform site density, swelling providing a reactive depth of ~0.5 µm, and quantitative azide formation and click conversion; residual azides likely remain inaccessible.
• Magnetic measurements on bead samples suffer from handling issues (static), no diamagnetic correction applied, and significant experimental uncertainties; evidence of potential weak interactions is inconclusive.
• Some esterification of phenolic groups during ring formation can occur, potentially impacting functionalization efficiency.