M₂⁴⁺ paddlewheel clusters as junction points in single-chain nanoparticles

Abstract

We report the formation of well-defined copper and molybdenum single-chain nanoparticles, exhibiting metal clusters (M₂⁴⁺) as junction points. Upon addition of dimetallic precursor complexes [Cu₂(OAc)₄] and [Mo₂(OAc)₄], carboxylic acid functionalized polymer chains collapse into nanoparticles containing a M₂⁴⁺ (M = Cu, Mo) centered paddlewheel folding moiety. This synthetic strategy allows a high crosslinking per metal unit and the incorporation of an adjustable number of dimetallic centers into a single polymer chain. The successful formation of M₂-SCNPs is evidenced by size exclusion chromatography and diffusion-ordered NMR spectroscopy. Detailed insight into the dimetallic folding unit is provided by ¹H/²H NMR-, IR-, Raman- and UV-Vis spectroscopy as well as comparative analyses of molecular dicopper and dimolybdenum model-complexes. For the first time, M₂⁴⁺ paddlewheel folding motifs, including quadruple bonded dimetallic units, as in the case of Mo₂⁴⁺, are utilized as structure forming elements in the realm of SCNP chemistry.

---

Introduction

Single-chain nanoparticles (SCNPs) are versatile and multi-functional macromolecular architectures, which have recently attracted particular attention due to their application as highly specialized enzyme-mimetic catalytic systems.^1–3 SCNPs consist of single polymer chains, which are folded intramolecularly into nanoparticles by an external trigger.^4–6 The collapse of single-chains can be induced by a variety of chemical reactions, yet recent approaches focus on coordinative metal to ligand bond formation. Employing metal-ions as cross-linkers within the chain does not only lead to the formation of compact nanoparticles, yet introduces unique properties to the resulting metal-SCNPs, such as catalytic activity, magnetism or luminescence.^7–11 Up to date, a series of metal-crosslinked SCNP systems has been established and to some extend examined in initial catalytic studies, demonstrating promising results, regarding enhanced catalytic activity, selectivity as well as catalyst recyclability.^1,12–17 Essential for metal-SCNP formation are suitable metal precursor complexes, exhibiting free coordination sites or leaving groups, which are readily replaceable by the ligand moieties in the polymer chains. Up to now, the literature mainly reports the use of monometallic complexes for SCNP formation, resulting in cross-points, consisting of single metal centers.

First results to incorporate dimetallic species into a SCNP system were achieved by the group of Berda. Here, a diiron-cluster was attached to a polymer chain, which – via subsequent light triggered [4 + 4] cycloaddition of anthracene functions in the polymer – was collapsed into a nanoparticle. In this way, a polymeric scaffold is formed around the diiron-cluster, creating a synthetic [FeFe] hydrogenase mimic.¹⁸ However, in this approach the cluster is not part of the actual SCNP formation, instead attached to the side chain.

Furthermore, the group of Lemcoff described the synthesis of metal-SCNPs exhibiting a μ-chloro bridged dinuclear binding motif.^9,15 The application of polycycloocta-1,5-diene (pCOD) allowed the incorporation of the dimetallic precursors [RhCl(C₂H₄)₂]₂ or [IrCl(COE)₂]₂via ligand exchange reaction. This approach enabled, for the first time, the introduction of dinuclear metal complexes as structure forming elements in SCNPs. A similar μ-chloro bridged dinuclear nickel crosslinking motif was proposed by the group of Hübner, utilizing pyridine-functionalized polymer chains.¹⁹ However, in the reported examples the two metal ions within the same folding unit are locally separated from each other. Therefore, a different accessible pathway for the implementation of dimetallic clusters into SCNP systems, enabling a close special distance and interaction between the metal ions, is highly desirable.

Herein, we report a synthetic strategy to incorporate dimetallic clusters (Cu₂⁴⁺, Mo₂⁴⁺) as junction points into single-chains, thus generating metal-SCNPs. In doing so, not only a high crosslinking within the polymer chain is achieved, yet unique M₂⁴⁺ paddlewheel folding motifs, including multiple bonded metal centers, are introduced into the field of SCNPs (Fig. 1).

Fig. 1 Schematic illustration of M₂-SCNP formation via an intramolecular chain collapse upon dimetal cluster coordination (Cu₂⁴⁺, Mo₂⁴⁺).

Experimental

Materials

4-Vinylbenzoic acid (>97%, TCI), 4-ethylbenzoic acid (99%, Alfa Aesar) and copper(II) acetate monohydrate (99.9%, Alfa Aesar) were used as received. Styrene (99%, Merck) was passed through a column of basic alumina (Acros) and stored at −19 °C. [Mo₂(OAc)₄], [Mo₂(OAc)₄]-d₁₂ ²⁰ and 2,2,6,6-tetramethyl-1-(1-phenylethoxy)piperidine¹² were synthesized according to literature procedures. Reactions and characterization methods of all molybdenum compounds were performed under exclusion of moisture and oxygen in flame-dried Schlenk-type glassware or in an argon-filled MBraun glovebox.

Hydrocarbon solvents (THF, n-pentane) were dried using an MBraun solvent purification system (SPS-800). THF was additionally distilled under nitrogen from potassium before storage over 4 Å molecular sieves. Prior to use, dimethyl sulfoxide (DMSO) was distilled under nitrogen from CaH₂ and methanol (MeOH) was distilled under nitrogen from magnesium. Dimethyl formamide (DMF) and dichloromethane (DCM) were purchased as analytical grade and used as received.

Instruments

NMR spectra were recorded on a Bruker Avance 400 (400 MHz) spectrometer. ¹H and ¹³C{¹H} NMR chemical shifts are reported relative to tetramethylsilane (TMS). ²H NMR measurements were performed in non-deuterated THF. Abbreviations include singlet (s), broad singlet (bs), doublet (d), triplet (t), quartet (q) and unresolved multiplet (m). The diffusion coefficients were measured with Diffusion Ordered NMR Spectroscopy (DOSY) experiments, using a Bruker Smart VT unit to control the temperature which was calibrated to be exactly 298 K.

Size-exclusion chromatography (SEC) measurements were performed on a PL-SEC 50 Plus (Polymer Laboratories, Varian), running on THF (HPLC-grade). The applied columns were a PLgel Mixed C guard column (50 × 7.5 mm), followed by three PLgel Mixed C linear columns (300 × 7.5 mm, 5 μm beadsize) and a differential refractive index (RI) detector. The device was operated at 35 °C column temperature with a flow rate of 1 mL min⁻¹. To obtain M_n and Đ values, the integration of the polymer peak was carried out from low elution times to approximately 40 minutes due to overlap with an SEC system peak. No baseline correction was performed. Consequently, M_n and Đ values are estimates.

IR spectra were obtained on a Bruker Tensor 37 FTIR spectrometer, equipped with a room temperature DLaTGS detector and a diamond ATR (attenuated total reflection) unit. Abbreviations used in the compounds’ analysis include very strong (vs), strong (s), medium (m), weak (w) and very weak (vw).

Raman spectra were obtained on a Bruker MultiRam spectrometer. Abbreviations used in the compounds’ analysis include very strong (vs), strong (s), medium (m), weak (w) and very weak (vw).

Elemental analyses were carried out with an Elementar Micro Cube.

Ultraviolet–visible (UV-Vis) spectra were recorded on a VARIAN Cary 50 Scan UV-Visible Spectrophotometer. Spectra were recorded in THF at room temperature and collected between 200 and 800 nm. Samples were baseline corrected with respect to the pure solvent.

Copolymerization of 4-vinyl-benzoic acid and styrene (P_Acid)

2,2,6,6-Tetramethyl-1-(1-phenylethoxy)piperidine (9.00 mg, 34.4 μmol, 1.00 equiv.) (synthesized according to literature¹²), styrene (2.64 g, 25.3 mmol, 735 equiv.) and 4-vinylbenzoic acid (178 mg, 1.21 mmol, 35.0 equiv.) were dissolved in 1.45 mL of degassed DMF. While stirring, the solution was degassed with N₂ for 30 min. Subsequently, the reaction mixture was heated at 125 °C for 9 h. Cooling to ambient temperature and opening the flask to air, quenched the reaction. After diluting with DCM, the polymer P_Acid was precipitated twice into cold MeOH, resulting in a white powder. Characterization via¹H NMR revealed a monomer ratio of approximately 1 : 14, regarding the higher amount to styrene (0.65 mmol of benzoic acid functionality in 1.00 g P_Acid).

SEC characterization (THF, RI) resulted in a number average molecular weight of M_n = 32 200 g mol⁻¹ and a polydispersity of Đ = 1.2.

¹H NMR (400 MHz, THF-d₈): δ [ppm] = 11.2 (bs, COOH), 7.91–7.51 (m, CH-benzoic acid, H^a), 7.25–6.27 (m, CH-aromatic), 2.59–2.83 (m, H-backbone).

IR (ATR): [small nu, Greek, tilde] [cm⁻¹] = 3082 (vw), 3060 (w), 3026 (s), 3002 (vw), 2923 (s), 2849 (w), 1734 (m), 1690 (m), 1603 (m), 1582 (vw), 1493 (s), 1451 (s), 1423 (w), 1369 (vw), 1315 (vw), 1285 (w), 1179 (w), 1156 (vw), 1086 (vw), 1028 (w), 907 (vw), 856 (vw), 756 (s), 698 (vs), 540 (m).

Raman (solid state): [small nu, Greek, tilde] [cm⁻¹] = 3055 (s), 3002 (vw), 2978 (vw), 2903 (m), 2851 (w), 1604 (s), 1584 (m), 1451 (w), 1329 (vw), 1183 (m), 1156 (w), 1032 (m), 1002 (vs), 798 (w), 662 (m), 224 (w).

Synthesis of Cu₂-SCNPs

4.60 mg of copper(II) acetate hydrate (11.5 μmol, 0.25 equiv.) was dissolved in 25 mL of THF (c = 4.6 × 10⁻⁴ mol L⁻¹). 70.0 mg of copolymer P_Acid (45.5 μmol benzoic acid functionality, 1.00 equiv.) was dissolved in 100 mL of THF. While stirring the metal salt solution, the polymer solution was added dropwise with a dropping rate of ∼5 mL h⁻¹. After complete addition (20 h), the blue solution was stirred for additional 4 h. The solution was concentrated under reduced pressure and subsequently precipitated into cold MeOH. The solid was filtered off and dried under vacuum, obtaining Cu₂-SCNPs.

SEC characterization (THF, RI) resulted in a number average molecular weight of M_n = 24 500 g mol⁻¹ and a polydispersity of Đ = 1.2.

¹H NMR (400 MHz, THF-d₈): δ [ppm] = 7.95–7.51 (m, CH-benzoic acid), 7.59–5.82 (m, CH-aromatic), 2.36–0.62 (m, H-backbone) – The multiplet resonance at δ = 7.95–7.51 ppm (CH-benzoic acid) is significantly broadened, compared to P_Acid, due to the coordination of paramagnetic Cu(II) ions. At δ = 11.2 ppm a singlet resonance is detected for R-COOH, indicating free carboxylic acid moieties, most likely due to an axial coordination of the acidic functionalities to the Cu₂⁴⁺ moieties.

IR (ATR): [small nu, Greek, tilde] [cm⁻¹] = 3083 (vw), 3060 (w), 3026 (s), 3001 (vw), 2923 (s), 2849 (w), 1733 (w), 1691 (w), 1619 (m), 1603 (m), 1583 (vw), 1565 (vw), 1493 (s), 1452 (s), 1407 (s), 1374 (vw), 1181 (w), 1155 (vw), 1070 (vw), 1028 (w), 907 (w), 857 (vw), 757 (s), 698 (vs), 540 (m).

Synthesis of Mo₂-SCNPs

7.00 mg of molybdenum(II) acetate (16.3 μmol, 0.25 equiv.) was dissolved in 50 mL of THF (c = 3.3 × 10⁻⁴ mol L⁻¹). 100 mg of copolymer P_Acid (65.0 μmol benzoic acid functionality, 1.00 equiv.) was dissolved in 120 mL of THF. While stirring the metal salt solution, the polymer solution was added dropwise with a dropping rate of ∼3 mL h⁻¹. After complete addition (40 h), the yellow solution was stirred for additional 8 h. The solution was concentrated under reduced pressure and subsequently precipitated into cold MeOH. The solid was filtered off and dried under vacuum, yielding Mo₂-SCNPs.

¹H NMR (400 MHz, CDCl₃): δ [ppm] = 8.16–7.54 (m, CH-benzoic acid), 7.37–6.13 (m, CH-aromatic), 2.55–0.89 (m, H-backbone). Additional NMR measurements in THF-d₈ revealed no free carboxylic acid proton resonances.

Raman (solid state): [small nu, Greek, tilde] [cm⁻¹] = 3055 (s), 3000 (vw), 2977 (vw), 2910 (m), 2851 (w), 1606 (s), 1583 (m), 1513 (m), 1451 (w), 1411 (s), 1401 (s), 1332 (vw), 1184 (m), 1156 (w), 1032 (m), 1002 (vs), 795 (w), 622 (m), 408 (w), 225 (w).

IR (ATR): [small nu, Greek, tilde] [cm⁻¹] = 3082 (vw), 3060 (w), 3026 (s), 3001 (vw), 2922 (s), 2849 (w), 1602 (m), 1556 (w), 1493 (s), 1451 (s), 1401 (m), 1181 (w), 1154 (vw), 1068 (vw), 1028 (w), 907 (w), 845 (vw), 756 (s), 698 (vs), 540 (m).

UV-Vis: λ_max. [nm] = 433.

For the synthesis and analysis of the Cu₂- and Mo₂-model complexes, as well as all original data, refer to the ESI.†

Results and discussion

Polymer synthesis and SCNP formation

To synthesize cluster folded SCNPs, the dimeric tetra-carboxylates [[M₂(O₂CR)₄] (M = Cu, Mo)] were chosen as suitable metal precursors. These complexes form characteristic paddlewheel structures, in which the carboxylates act as bidentate bridging ligands of the dimetallic center.^21–23 The bridging ligands cause a close proximity of the metal ions, thus allowing quadruple bonded clusters in the case of molybdenum.²⁴ Of specific interest are the readily accessible tetraacetates [Cu₂(OAc)₄] and [Mo₂(OAc)₄], as both complexes are known to readily substitute their acetate ligands by addition of stronger carboxylic acids at room temperature,^25–27 ideal reaction conditions for SCNP formation. This synthetic strategy allows a high crosslinking of up to four linkages per dimetallic unit. Furthermore, [Cu₂(OAc)₄] has already been successfully employed as a crosslinking agent in polymeric systems, modulating the mechanical properties of metallopolymers.²⁸

Thus, the precursor polymer chains require benzoic acid moieties to allow, via simple carboxylate exchange reactions, the incorporation of dimetallic M₂⁴⁺ units. Provided that at least two acid linkers in the polymer coordinate to the dimetallic unit, a chain collapse is anticipated. Benzoic acid moieties were incorporated into the polymer chains by statistical copolymerization of styrene and 4-vinylbenzoic acid via nitroxide-mediated polymerization (NMP),^29,30 resulting in P_Acid (Scheme 1). The copolymer composition was determined by ¹H NMR spectroscopy (ESI, Fig. S4†), distinguishing the aromatic proton resonances of both monomers. Hence, a monomer ratio of 1 : 14, with regard of the higher amount to styrene, was identified, resulting in approximately 7 mol% of carboxylic acid units (approx. 20 units per chain) for P_Acid, in good accordance to the feed ratio. Size exclusion chromatography (SEC) analysis [THF, RI] of P_Acid indicated a M_n of 32 200 g mol⁻¹ and a narrow distribution of Đ ≈ 1.2.

Scheme 1 Nitroxide-mediated polymerization (NMP) of styrene and 4-vinylbenzoic acid results in copolymer P_Acid. Subsequent addition of either [Cu₂(OAc)₄]·2H₂O or [Mo₂(OAc)₄] under high dilution, leads to SCNP formation via carboxylate exchange, with dimetallic clusters as cross-linkers (M₂-SCNPs). The cross-links of the Mo₂-SCNPs consist of carboxylate bridged, quadruple bonded Mo–Mo units, whereas no metal–metal bond is present in the dicopper species. The depicted folding unit illustrates a replacement of all four acetate ligands by the benzoic acid moieties of P_Acid and neglects axial coordination of additional donor molecules.

For the subsequent SCNP formation, P_Acid was dissolved in THF and added to a solution of the metal precursors [Cu₂(OAc)₄]·2H₂O or [Mo₂(OAc)₄]. Working under high dilution is necessary to prevent network formation of the polymer chains by intermolecular reaction. After isolation of the polymeric structures, the collapse into nanoparticles and the structural motif of the cross-links, were evidenced. The folding of copolymer P_Acid into a nanoparticle upon [Cu₂⁴⁺] coordination was initially confirmed by SEC analysis [THF, RI]. In comparison to P_Acid, the SEC trace for the Cu₂-SCNPs is shifted towards longer retention times (Fig. 2), indicating a smaller hydrodynamic radius, and thus the exclusive formation of SCNP nanostructures (Đ ≈ 1.2). However, SEC analysis of the Mo₂-SCNPs was not feasible, as Mo₂⁴⁺ compounds are known to be highly reactive towards moisture and air, hence prohibiting a meaningful analysis.

Fig. 2 SEC (THF, RI) traces of P_Acid and Cu₂-SCNPs. The shift of the trace of the Cu₂-SCNPs towards longer retention times indicates a decrease of the hydrodynamic radius upon metal coordination, thus evidencing [Cu₂⁴⁺]-crosslinked polymeric nanoparticles.

The decrease of the particles’ mean hydrodynamic radius was additionally confirmed by diffusion ordered NMR spectroscopy (DOSY).³¹ The diffusion coefficient of molecules in solution is dependent on their hydrodynamic radius, hence particles of different sizes can be distinguished. The mean hydrodynamic radius of P_Acid resulted in R_h = 3.9 nm, whereas a radius of R_h = 2.3 nm for the Cu₂-SCNPs and R_h = 2.6 nm for the Mo₂-SCNPs was determined (ESI, Fig. S14–S16†). Thus, a reduction of the hydrodynamic radius by approximately 35–40% evidenced a successful SCNP formation for both dimetallic-crosslinked structures.

Interestingly, in the ¹H NMR spectrum (THF-d₈) of the Cu₂-SCNPs a resonance at δ = 11.2 ppm, belonging to the acidic protons (R-COOH) of P_Acid, is still detected (approximately 30% compared to P_Acid, Fig. S6†). Remaining carboxylic acid moieties in the polymer chains derive most likely from additional axial coordination of the acidic functionalities to the Cu₂ moiety (for details, refer to section ‘model complexes’). In contrast to the Cu₂-SCNPs, no resonance for acidic protons is detected in the ¹H NMR spectrum of the Mo₂-SCNPs (ESI, Fig. S9†). Given the stoichiometry applied (0.25 mol% of metal precursor), a substitution of all four acetate ligands by the benzoic acid moieties of P_Acid is presumed, considering the error range of NMR spectroscopy. To quantify the actual number of substituted acetate groups during the folding process, deuterated [Mo₂(OAc)₄]-d₁₂ was applied as a metal precursor. ²H NMR spectroscopy of the corresponding Mo₂-SCNPs revealed no remaining deuterium resonances, indicating a complete substitution of all deuterated acetate ligands (ESI, Fig. S12 and S13†). However, analogous experiments with deuterated [Cu₂(OAc)₄]-d₁₂ were not feasible. NMR analytics were hindered by the paramagnetic Cu₂⁴⁺ moieties, which is in agreement with the corresponding dicopper tetra carboxylate model complexes (see next section).

Model complexes

After evidencing the particles’ size reduction, a detailed look into the dimetallic linkages was crucial for understanding the composition and structure of the folding moieties. Therefore, corresponding molecular dimetallic model complexes of both copper and molybdenum were synthesized. To simulate the benzoic acid moieties within the polymer chain, 4-ethylbenzoic acid was chosen as the most suitable ligand (Scheme 2).

Scheme 2 Synthesis of the Cu₂-complex 1′via a carboxylic exchange reaction in THF. In addition to the four bridging carboxylates, two additional carboxylic acids coordinate axially to the Cu₂⁴⁺ unit. Subsequent crystallization from hot DMSO leads to an axial ligand replacement (Cu₂-model complex 1).

Under similar reaction conditions as for the M₂-SCNP synthesis, 4-ethylbenzoic acid reacted with [Cu₂(OAc)₄]·2H₂O in THF, resulting in the compound [Cu₂(p-O₂CPhEt)₄(p-HO₂CPhEt)₂] (1′), in which six ligand molecules coordinate in different coordination modes to one dicopper unit (solid state structure, Fig. 3, top).³² Four ligands arrange as carboxylates in a typical bidentate bridged paddlewheel structure. In addition, two carboxylic acid molecules coordinate axially to both sides of the dicopper unit via their carbonyl functionality to saturate the coordination sphere. Most likely, this structural arrangement explains the remaining carboxylic acid functionalities observed in the ¹H NMR spectrum of the Cu₂-SCNPs. However, subsequent crystallization from DMSO leads to the substitution of both axially binding 4-ethyl benzoic acid moieties and the formation of [Cu₂(p-O₂CPh-Et)₄(dmso)₂] (1) (ESI, Fig. S1†). A Cu–Cu distance of 2.602 Å is detected for complex 1 in the solid state, in agreement to structurally similar dicopper complexes,³³ indicating no direct bond between the two copper atoms.

Fig. 3 Molecular structures of the M₂-model complexes 1′ (copper, top) and 2 (molybdenum, bottom) in the solid state. Most of the hydrogen atoms and non-coordinating solvent molecules (DMSO) are omitted for clarity. The model complexes illustrate the assumed folding motifs in the Cu₂-SCNPs and Mo₂-SCNPs.

The analogous reaction with [Mo₂(OAc)₄], resulted in the homoleptic compound [Mo₂(p-O₂CPh-Et)₄] (2), in which additional THF molecules saturate the axial coordination sites of the dimolybdenum Mo₂⁴⁺ unit. The molecular structure of 2 (crystallized from a DMSO solution) in the solid state (Fig. 3, bottom) confirms the expected, highly symmetric, tetracarboxylato-bridged paddlewheel structure with two axially coordinating solvent molecules. The Mo-Mo quadruple bond length is 2.115 Å, which is a typical value in a bidentate bridged coordination geometry (2.06–2.13 Å).³⁴

The synthesis of both model complexes proceeded at ambient temperature, yielding exclusively homoleptic paddlewheel structures (neglecting axial solvent coordination) in high yields. Therefore, a similar reactivity is expected for the formation of the Cu₂-SCNPs and Mo₂-SCNPs. The complexes 1, 1′ and 2 are assumed to depict the actual folding motif of the SCNP linkages, exhibiting a tetracarboxylate dimetallic folding unit. However, a complete substitution of all four acetate ligands of the precursor complexes by the benzoic acid moieties of P_Acid might be hampered, due to steric hindrance in the collapsed nanoparticles after a certain folding degree. Hence, to a small extend, heteroleptic structures, containing unsubstituted acetate ligands, are likely to be present in the SCNP structures.

Characterization of the folding unit

At first, vibrational spectroscopy analysis was employed to determine the structural arrangement within the SCNP folding units, comparing the M₂-complexed nanoparticles with the dimetallic model complexes (1 and 2). In a bridged bidentate coordination of a carboxylate unit (RCO₂⁻), the two C–O bonds are equivalent and the corresponding frequencies of the symmetric and anti-symmetric stretching modes are detected as strong bands in the region of 1400–1600 cm⁻¹, mainly depending on the ligand design and the metal ions.^20,33,35

Applying IR spectroscopy, a characteristic band for the C=O stretching mode of the benzoic acid moieties (R-COOH) at [small nu, Greek, tilde] = 1690 cm⁻¹ is detected for copolymer P_Acid (Fig. 4, black). In the IR spectrum of the Cu₂-SCNPs, the band at 1690 cm⁻¹ (C=O) decreased significantly, yet did not disappear completely. This is in good agreement with the NMR data of the Cu₂-SCNPs in which approximately 30% of free carboxylic acid moieties are still observed. Additionally, two characteristic bands appear at [small nu, Greek, tilde]_a = 1619 and [small nu, Greek, tilde]_s = 1374 cm⁻¹ for the asymmetric and symmetric C–O stretching modes (Fig. 4, blue).³⁶ In conclusion, IR spectroscopy clearly indicates a bridged carboxylate coordination of the polymers’ acid moieties to the dicopper units. The IR spectrum of the Cu₂-model complex 1 further supports this assumption, exhibiting similar bands at [small nu, Greek, tilde]_a = 1619 and [small nu, Greek, tilde]_s = 1398 cm⁻¹ (ESI, Fig. S20†).

Fig. 4 IR spectra of P_Acid (black) and Cu₂-SCNPs (blue, doted). New characteristic bands, confirming a bridged carboxylate coordination at [small nu, Greek, tilde] = 1619 and 1374 cm⁻¹ (asymmetric and symmetric C–O stretching modes), were detected for the Cu₂-SCNPs.

IR spectroscopy of the Mo₂-SCNPs indicates, in accordance to ¹H NMR spectroscopy, a coordination of nearly all carboxylic acid moieties, as the C=O stretching mode of free carboxylic acids has almost vanished. In addition, an intense band at [small nu, Greek, tilde]_s = 1401 cm⁻¹ appears, associated with the symmetric C–O stretching mode of a bridged bidentate carboxylate (ESI, Fig. S19†). However, due to overlaying bands, a band for the asymmetric stretching mode cannot be observed. Therefore, Raman spectroscopy was applied as a complementary vibrational spectroscopic method for the Mo₂-SCNPs. In the spectrum of the Mo₂-SCNPs, two bands at [small nu, Greek, tilde]_a = 1513 cm⁻¹ and [small nu, Greek, tilde]_s = 1411/1401 cm⁻¹ (asymmetric and symmetric C–O stretching modes) are visible, indicating a bridged carboxylate coordination (Fig. 5, orange). The relatively small distance between these two bands (approx. 100 cm⁻¹) is typical for a dimolybdenum species.³³ Additionally, the characteristic band at [small nu, Greek, tilde] = 408 cm⁻¹ is associated with the quadruple bonded Mo–Mo stretching mode,^33,37 proving an intact Mo–Mo quadruple bond within the Mo₂-SCNP structure. Similar results are received for the Mo₂-model compound 2 ([small nu, Greek, tilde]_a = 1515 cm⁻¹; [small nu, Greek, tilde]_s = 1415/1398 cm⁻¹; [small nu, Greek, tilde]_Mo–Mo = 403 cm⁻¹; Fig. S24†), yet the axial coordination of different solvent molecules (THF, DMSO) needs to be considered.

Fig. 5 Raman spectra of P_Acid (black) and Mo₂-SCNPs (orange, doted). For the Mo₂-SCNPs characteristic bands for a bridged carboxylate coordination at [small nu, Greek, tilde] = 1513 and 1411/1401 cm⁻¹ (asymmetric and symmetric C–O stretching modes) are observed. Additionally, a band at [small nu, Greek, tilde] = 408 cm⁻¹ is detected, related to the Mo–Mo stretching mode.

The M₂-SCNPs and their respective model complexes were further analyzed by UV-Vis spectroscopy. Upon metal coordination, the resulting M₂-SCNPs exhibit either a blue (Cu₂) or a dark yellow/orange (Mo₂) color. UV-Vis measurements of the Cu₂-SCNPs in THF showed a broad absorption between 550–800 nm (ESI, Fig. S25†), characteristic for Cu(II) carboxylates and typically considered as the result of d–d transitions,³⁶ yet recent studies also consider different contributions.³⁸ UV-Vis measurements of the Cu₂-model complex (1) and the starting material [Cu₂(OAc)₄]·2H₂O showed similar trace shapes and absorption maxima (ESI, Fig. S25 and S27†), validating the assumption of a successful implementation of Cu(II) ions into the polymeric structure.

Analogous UV-Vis measurements of the Mo₂-SCNPs show a pronounced absorption band between 350–500 nm, with an absorption maximum at approximately 430 nm (Fig. 6). This absorption originates from the electronic transition between the δ orbitals of the Mo₂ unit and π* orbitals of the aromatic ligands. It represents a fully-allowed metal to ligand charge transfer (MLCT), which can be tuned over a wide wavelength range, strongly dependent on the π acceptor-ligands (e.g. O₂CR).^27,39–41 In comparison, the starting material [Mo₂(OAc)₄] does not show any absorption in this area, as no π acceptor ligands are present (ESI, Fig. S28†). Thus, the characteristic absorption of the Mo₂-SCNPs at λ = 430 nm proves the substitution of acetate by the π acceptor ligand moieties of the polymer chains, allowing a δ–π* transition. UV-Vis measurement of the Mo₂-model complex (2) shows an almost identical absorption between 350–500 nm (Fig. 6), further evidencing the proposed structural motif within the Mo₂-SCNPs and the presence of a quadruple bonded dimolybdenum unit.

Fig. 6 UV-Vis spectra of P_Acid, Mo₂-SCNPs and the Mo₂-model complex 2. A local maximum absorbance at 430–435 nm is observed for the Mo₂ containing compounds.

The combined results of vibrational (IR, Raman) and UV-Vis spectroscopy clearly point towards a successful incorporation of the dimetallic clusters and the formation of a paddlewheel folding motif, as proposed by the model complexes 1 and 2. Thus, a bridged carboxylate dimetallic moiety is postulated as the actual junction point in the SCNPs.

Conclusions

Herein, we present the synthesis and in-depth characterization of polymeric single-chain nanoparticles, exhibiting M₂⁴⁺ (M = Cu, Mo) dimetallic units as novel crosslinking junctions (M₂-SCNPs). Reacting the dimeric precursor complexes [Cu₂(OAc)₄] and [Mo₂(OAc)₄] with carboxylic acid functionalized polymer chains, multiple cross-links within the chains are formed, leading to intramolecular collapse. The successful formation of nanoparticles was evidenced by SEC and DOSY, whereas the dimetallic folding unit was analyzed by ¹H/²H NMR, IR-, Raman- and UV-Vis spectroscopy. In addition, the synthesis of dicopper and dimolybdenum model complexes (1, 1′ and 2) provided detailed insight into the paddlewheel structured folding motif. The described M₂-SCNPs employ metal clusters as structure forming elements, introducing a unique M₂⁴⁺ paddlewheel folding motif to SCNP chemistry. Importantly, Mo₂-SCNPs display the first nanostructured polymeric system exhibiting quadruple bonded metal centers, thus creating novel soft matter materials with promising reactivity and characteristics.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

C. B.-K. and P. W. R. acknowledge support from the SFB 1176 (project A2) funded by the German Research Council (DFG). H. R.'s and N. K.'s PhD studies are additionally supported by the Fonds der Chemischen Industrie (FCI). C. B.-K. acknowledges key support from the Australian Research Council (ARC) in the context of a Laureate Fellowship as well as by the Queensland University of Technology (QUT). C. B.-K. additionally acknowledges continued support by the Helmholtz association via the STN and BIFTM programs.

References

1. H. Rothfuss, N. D. Knöfel, P. W. Roesky and C. Barner-Kowollik, J. Am. Chem. Soc., 2018, 140, 5875.
2. Y. Liu, P. Turunen, B. F. M. de Waal, K. G. Blank, A. E. Rowan, A. R. A. Palmans and E. W. Meijer, Mol. Syst. Des. Eng., 2018, 3, 609.
3. J. P. Cole, A. M. Hanlon, K. J. Rodriguez and E. B. Berda, J. Polym. Sci., Part A: Polym. Chem., 2017, 55, 191.
4. S. Mavila, O. Eivgi, I. Berkovich and N. G. Lemcoff, Chem. Rev., 2016, 116, 878.
5. J. A. Pomposo, J. Rubio-Cervilla, A. J. Moreno, F. Lo Verso, P. Bacova, A. Arbe and J. Colmenero, Macromolecules, 2017, 50, 1732.
6. A. M. Hanlon, C. K. Lyon and E. B. Berda, Macromolecules, 2016, 49, 2.
7. A. Latorre-Sánchez and J. A. Pomposo, Polym. Int., 2016, 65, 855.
8. Y. Bai, X. Feng, H. Xing, Y. Xu, B. K. Kim, N. Baig, T. Zhou, A. A. Gewirth, Y. Lu, E. Oldfield and S. C. Zimmerman, J. Am. Chem. Soc., 2016, 138, 11077.
9. S. Mavila, C. E. Diesendruck, S. Linde, L. Amir, R. Shikler and N. G. Lemcoff, Angew. Chem., Int. Ed., 2013, 52, 5767.
10. J. Rubio-Cervilla, E. González and J. Pomposo, Nanomaterials, 2017, 7, 341.
11. Y. Liu, S. Pujals, P. J. M. Stals, T. Paulöhrl, S. I. Presolski, E. W. Meijer, L. Albertazzi and A. R. A. Palmans, J. Am. Chem. Soc., 2018, 140, 3423.
12. N. D. Knöfel, H. Rothfuss, J. Willenbacher, C. Barner-Kowollik and P. W. Roesky, Angew. Chem., Int. Ed., 2017, 56, 4950.
13. A. Sanchez-Sanchez, A. Arbe, J. Colmenero and J. A. Pomposo, ACS Macro Lett., 2014, 3, 439.
14. T. Terashima, T. Mes, T. F. A. De Greef, M. A. J. Gillissen, P. Besenius, A. R. A. Palmans and E. W. Meijer, J. Am. Chem. Soc., 2011, 133, 4742.
15. S. Mavila, I. Rozenberg and N. G. Lemcoff, Chem. Sci., 2014, 5, 4196.
16. E. Huerta, P. J. M. Stals, E. W. Meijer and A. R. A. Palmans, Angew. Chem., Int. Ed., 2013, 52, 2906.
17. S. Thanneeru, J. K. Nganga, A. S. Amin, B. Liu, L. Jin, A. M. Angeles-Boza and J. He, ChemCatChem, 2017, 9, 1157.
18. C. A. Tooley, S. Pazicni and E. B. Berda, Polym. Chem., 2015, 6, 7646.
19. K. Freytag, S. Säfken, K. Wolter, J. C. Namyslo and E. G. Hübner, Polym. Chem., 2017, 8, 7546.
20. T. A. Stephenson, E. Bannister and G. Wilkinson, J. Chem. Soc., 1964, 2538.
21. P. Sharrock and M. Melník, Can. J. Chem., 1985, 63, 52.
22. P. K. Ross, M. D. Allendorf and E. I. Solomon, J. Am. Chem. Soc., 1989, 111, 4009.
23. D. Lawton and R. Mason, J. Am. Chem. Soc., 1965, 87, 921.
24. F. A. Cotton, C. A. Murillo and R. A. Walton, Multiple Bonds Between Metal Atoms, Springer Science and Business Media, Inc., New York, 3rd edn, 2005.
25. J. Hicks, S. P. Ring and N. J. Patmore, Dalton Trans., 2012, 41, 6641.
26. F. A. Cotton, L. R. Falvello, A. H. Reid and J. H. Tocher, J. Organomet. Chem., 1987, 319, 87.
27. N. D. Knöfel, C. Schweigert, T. J. Feuerstein, C. Schoo, N. Reinfandt, A.-N. Unterreiner and P. W. Roesky, Inorg. Chem., 2018, 57, 9364.
28. V. Yuval, B. Suwon and M. N. Silberstein, J. Polym. Sci., Part A: Polym. Chem., 2018, 56, 1117.
29. C. J. Hawker, G. G. Barclay and J. Dao, J. Am. Chem. Soc., 1996, 118, 11467.
30. G. Moad and E. Rizzardo, Macromolecules, 1995, 28, 8722.
31. E. Blasco, B. T. Tuten, H. Frisch, A. Lederer and C. Barner-Kowollik, Polym. Chem., 2017, 8, 5845.
32. A. C. Sunil, B. C. B. Bezuidenhoudt and J. M. Janse van Rensburg, Acta Crystallogr., Sect. E: Struct. Rep. Online, 2008, 64, m939.
33. J. Catterick and P. Thornton, in Advances in Inorganic Chemistry and Radiochemistry, ed. H. J. Emeléus and A. G. Sharpe, Academic Press, 1977, vol. 20, p. 291.
34. F. A. Cotton, L. M. Daniels, E. A. Hillard and C. A. Murillo, Inorg. Chem., 2002, 41, 2466.
35. F. A. Cotton and D. G. Lay, Inorg. Chem., 1981, 20, 935.
36. M. Mikuriya, R. Indrawati, R. Hashido, S. Matsubara, C. Nakamura, D. Yoshioka, K. Yokota, M. Fukuzaki and M. Handa, Magnetochemistry, 2018, 4, 22.
37. A. J. Hempleman, R. J. H. Clark and C. D. Flint, Inorg. Chem., 1986, 25, 2915.
38. M. Kyuzou, W. Mori and J. Tanaka, Inorg. Chim. Acta, 2010, 363, 930.
39. M. H. Chisholm, T. L. Gustafson and C. Turro, Acc. Chem. Res., 2013, 46, 529.
40. M. H. Chisholm, S. E. Brown-Xu and T. F. Spilker, Acc. Chem. Res., 2015, 48, 877.
41. B. G. Alberding, M. H. Chisholm, C. B. Durr, J. C. Gallucci, Y. Ghosh and T. F. Spilker, Dalton Trans., 2014, 43, 11397.

---

Footnotes

† Electronic supplementary information (ESI) available. CCDC 1873610 and 1873611. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8py01486h

‡ These authors contributed equally to this work.

---