Multi-Functionalization of Solid Support via Zn(II)-Mediated Chirality-Directed Self-Assembly

Establishing a strategy for realizing programmed self-assembly is critical in manufacturing materials with functional hybrid structures. In this work, we introduce a robust methodology for enabling multi-component self-assembly using the concept of chirality-directed self-assembly. A specific combination of heterochiral Zn(II) methylene bis(oxazoline) (BOX) complexes can be selectively generated when combinations of enantiomers of chiral BOX ligands are mixed in the presence of Zn(Oac)2. The resulting Zn(II) BOX complexes, unlike non-covalent bonds, are highly stable and stay intact at elevated temperatures, yet can be reversibly disintegrated under mild conditions using EDTA. This approach can be easily applied to multi-functionalize various solid supports enabling the one-pot generation of multi-functional hybrid struc-


Introduction
Multi-component self-assembly is critical in the construction of functional hybrid structures from various building blocks, constructing structurally complex and functional architectures for various applications. 1 The process of self-assembly, 2 where building units of a system are organized into an ordered and/or functional structure, has attracted researchers from a diverse range of disciplines that vary from chemistry and materials science to engineering and technology. 3 The assembly through covalent bonds 4 has the advantage of being independent of and unaffected by solvent conditions, such as solvent polarity and ionic strength. 5 However, the generation of covalent bonds often raises challenges due to the compatibility of reagents and reaction conditions for functional groups within building units. 6 This issue often results in significant limitations in the diversity of assembling functional groups. In contrast, using non-covalent bonds 7 reduces the risk of their reaction compatibility due to their mild assembly conditions. In addition, highly reversible bonds allow multi-functional structures disassemble back into individual building units, enabling the recycling of these small units. 8 Unfortunately, their inherent weaker bond strengths, which are highly sensitive to environmental conditions such as solvent polarity, temperature, ionic strength, and pH, often severely limit their range of applications. 9 To this end, reversible metal-ligand complexes 10 have higher stability under a broader range of reaction conditions/environments and can be assembled in situ under mild conditions. Most examples utilize combinations of metal and achiral ligands for self-assembly, and a lack of selectivities among achiral ligands often leads to challenges in controlling the overall shapes and sizes of supramolecules. To this end, we recently reported a versatile approach to preparing resin-immobilized catalysts through metal-ligand complexes. 11 Our unique approach utilizes the concept of chirality-directed self-assembly 12 where chiral bis(oxazoline) (BOX) ligands with complemental chirality selectively form a heterochiral Zn(II) complex, 13 anchoring a catalyst onto the resin. In other words, when equimolar amounts of chiral BOX ligands with opposite enantiomers were mixed in the presence of Zn(OAc) 2 , exclusive formation of the heterochiral Zn(II) complex was observed. The high selectivity toward the formation of the heterochiral Zn(II) BOX complex arises from reduced steric crowding of the phenyl rings on the chiral BOX ligand around the Zn(II) center. 14 The Zn(II) center on the homochiral Zn(II) complex is destabilized by adopting a distorted tetrahedral geometry to accommodate two phenyl substituents in the same quadrant ( Figure 1). The loaded catalyst can be altered in situ through ligand exchange for 2-step sequential reactions using the same resin. The report demonstrated that resinbound catalysts could be easily recycled under mild conditions without damaging resin functional groups, helping to reduce the generation of microparticle wastes.
The concept of chirality-directed self-assembly was also employed to quantitatively generate distinct Janus dendrimers in situ from various dendron subunits attached to chiral BOX ligands. 15 The highly selective formation of the heterochiral Zn(II) complex at the focal point of the Janus dendrimer enables the exclusive formation of Janus dendrimer upon mixing two distinct dendritic domains, each functionalized by opposite enantiomer of BOX ligands. The heterochiral Zn(II) BOX complex at the focal point was stable under elevated temperature. However, the complex can be reversibly disintegrated under mild conditions with the use of EDTA.
These previous studies have led us to believe that resin multi-functionalization is possible if we identify sets of chiral BOX ligands that selectively form multiple Zn(II) BOX complexes on the resin surface. 1,5,16 In other words, to realize the multi-functionalization of the solid support, the interaction between the BOX ligand immobilized on the solid support and added ligands must be highly selective (Scheme 1A). Lack of such selectivity led to a non-selective complexation, resulting in the generation of wasteful unbound catalysts (Scheme 1B); therefore, difficult to control the relative ratio of the functional groups on the resin. Hence, our goal here is to identify the optimum sets of chiral BOX ligands generating highly selective Zn(II) BOX complexes to establish a versatile one-pot approach for the multi-functionalization of solid support.

Results and Discussion
Comprehensive studies on the relative stability of various homo-and heterochiral Zn(II) BOX complexes have not been reported thus far (Scheme 1A vs. 1B). Therefore, we decided to elucidate the efficiency in generating heterochiral Zn(II) complexes to identify the correlation between the size of chiral substituents and selectivity toward forming heterochiral Zn(II) complexes (Table 1). For this investigation, the ratio of heterochiral vs. homochiral Zn(II) BOX complexes was measured using 1 H NMR spectroscopy in CDCl 3 (Figure 2).
The exclusive formation of the heterochiral Zn(II) complex (SS,RR)-Zn(II) BOX (Ph/Ph) was observed when equimolar amounts of (S,S)-BOX (Ph) and (R,R)-BOX (Ph) were used ( Table 1, entry 1; Figure 2, top). The ligand exchange process was instantaneous, and no homochiral Zn(II) complexes nor free ligands remained in the solutions. By contrast, when the steric bulkiness was moved further away from the chiral center on the oxazoline ring, as steric interactions among chiral substituents in the homochiral Zn(II) complex is reduced, selectivity toward the heterochiral Zn(II) complex was reduced. For instance, the benzyl-substituted chiral BOXʼs Zn(II) complex reached equilibrium at ambient temperature with 25 % homochiral Zn(II) complexes remaining in the solution (Table 1, entry 2; Figure 2, middle). 17 Interestingly, similar selectivity was obtained with methyl-substituted chiral BOX (Table 1,   In addition, when equimolar amounts of (RR,RR)-Zn(II) BOX (Bn) and (SS,SS)-Zn(II) BOX (Me) were mixed, the outcome of selectivity toward generating the heterochiral Zn (II) complex ((SS,RR)-Zn(II) BOX (Me/Bn)) was identical to the cases with (SS,RR)-Zn(II) BOX (Me) and (SS,RR)-Zn(II) BOX (Bn) (entry 4). 18 These results indicate that steric effect at the C1 position on the chiral substituent is significant in the outcome of chiral discrimination processes.
Next, to probe stability differences among homochiral Zn(II) complexes (Scheme 1B), a free ligand titration experiment was conducted as shown in Table 2 To achieve the quantitative multi-functionalization of the solid support (Scheme 1A), combinations of chiral ligands must be carefully chosen to avoid generating wasteful unbound complexes. To this end, based on the results above, the heterochiral Zn(II) complexes showed higher stabilities than all of the homochiral Zn(II) complexes studied here. In-     In other words, both complexes were intact at elevated temperatures, and no disintegration of complexes was observed. These experiments suggest that chirality-driven self-assembly can load multiple functional groups quantitatively through metal-ligand complexes onto the resin without generating wasteful unbound ligands.
To probe the applicability of the chirality-driven self-assembly for the multi-functionalization of resin depicted in Scheme 1A, we decided to use (R,R)-BOX-immobilized Wang resin, 20 which we have previously used to demonstrate a single functionalization of resin through self-assembly. As previously demonstrated, the functional loading capacity of the (R,R)-BOX-immobilized Wang resin was calculated by the standard Fmoc-quantification approach and determined to be 0.28 mmol/g. 21 The generation of heterochiral Zn(II) BOX on the resin was achieved by mixing (R,R)-BOX-immobilized Wang resin with free (S,S)-BOX ligands in the presence of an equimolar amount of Zn(OAc) 2 . The formation of heterochiral Zn(II) complexes on the resin was monitored through FTIR analysis. Any unbound free ligand and free Zn(II) complexes can be detected by 1 H NMR spectroscopy by analyzing the supernatant solution.
When the equimolar BOX ligand, (S,S)-BOX (Ph), was added to the (R,R)-BOX (Ph)-functionalized Wang resin along with Zn(OAc) 2 , no free ligands or unbound homochiral Zn(II) BOX (Ph) complexes were detected in the supernatant solution (CDCl 3 ) ( Table 3,    The FTIR analysis of EDTA-treated resin shows the C=N stretch at 1655 cm −1 , indicating that the (R,R)-BOX (Ph) covalently attached to the resin was not affected by the EDTA treatment. In fact, the heterochiral Zn(II) BOX can be regenerated on the EDTA-treated resin by mixing an equimolar amount of (S,S)-BOX (Ph) along with Zn(OAc) 2 . In sharp contrast to the formation of the heterochiral Zn(II) complex on the resin, the efficiency in forming homochiral Zn(II) complex, using (R,R)-BOX (Ph), on the resin was not efficient as almost half of the added ligand formed unbound homochiral Zn(II) complex in solution (entry 3). These results support that using chiral BOX ligands with complemental chirality led to quantitative loading of the functional group on the resin.
Encouraged by the results above, a quantitative formation of two distinct heterochiral Zn(II) complexes on the (R,R)-BOX (Ph)-functionalized Wang resin was investigated. When 0.5 equiv of each (S,S)-BOX (Ph) and (S,S)-BOX (Me) were mixed with the (R,R)-BOX (Ph)-functionalized resin in the presence of Zn(OAc) 2 , these added free ligands quickly disappeared from the supernatant solution (Table 4, entry 1). The FTIR analysis of the resin indicated a generation of two heterochiral Zn(II) complexes on a resin with two C=N stretches, 1598 cm −1 ((SS,RR)-Zn(II) BOX (Ph/Ph)) and 1593 cm −1 ((SS,RR)-Zn(II) BOX (Me/Ph)), shifted from 1655 cm −1 (uncomplexed (R,R)-BOX (Ph) on a resin) (Scheme 6D vs. 6B and 6C). We did not observe any thermal stability differences on the resin between (SS,RR)-Zn(II) BOX (Ph/Ph) and (SS,RR)-Zn(II) BOX (Me/Ph) through a temperature range tested (up to 160°C in d 6 -DMSO). In addition, no ligand exchange took place when additional (S,S)-BOX (Ph) or (S,S)-BOX (Me) was further mixed with the above dual-function-alized resin. For example, when 10 equiv of (S,S)-BOX (Me) was mixed with (SS,RR)-Zn(II) BOX (Ph/Ph)/(SS,RR)-Zn(II) BOX (Me/Ph) dual-functionalized resin, no (S,S)-BOX (Ph) was released into the supernatant solution after 72 h at ambient temperature. In addition, both FTIR C=N stretch signals (1598 cm −1 and 1593 cm −1 ) remained the same after the resulting resins were washed several times.
These observations have led us to believe the alteration of the ratio of loading functional group can be possible by simply adjusting the mixture ratio of chiral BOX ligands. Indeed, when a ligand mixture consisting of 1 : 3 ratio of (S,S)-BOX (Me) and Fmoc-functionalized (S,S)-BOX (Ph) was mixed with (R,R)-BOX (Ph)-immobilized resin in the presence of Zn(OAc) 2 , the ratio of bound ligands (> 98 % saturation on loading capacity) was identical (1 : 3) to the added free ligand mixture (entry 3). An identical result was obtained when the same mixture was used in excess molar amount (entry 4). These results indicate that the ratio of loaded  functional groups can be precisely controlled by altering the mixture ratio of added functional groups.

Conclusions
In summary, the chirality-directed self-assembly of heteroleptic Zn(II) complexes can be used as a strategy to load multiple functional groups on resin. Due to the high selectivity toward forming heteroleptic Zn(II) complexes, this straightforward process can be used without purification because there are no by-products formed. The selectivity toward the formation of heterochiral Zn(II) core complexes was not influenced by the size of the chiral substituents on the BOX ligand. The resin-generated heteroleptic Zn(II) complexes are stable at elevated temperatures; however, they can be easily disassembled through treatment by EDTA for resin recyclization. In theory, this methodology enables to functionalize a range of solid supports such as graphene, metal nanoparticles, and various polymer-based materials. Further studies are currently in progress, including theoretical calculations of various homo-and heterochiral BOX Zn(II) complexes.

Experimental Section
General Information: Solvents and materials were obtained from commercial suppliers (Fisher Scientific, Sigma-Aldrich, and Alfa Aesar) and used without further purification. (R,R)-or (S,S)-BOX ligands were prepared according to the published procedure. 24 (R,R)-or (S,S)-BOX (Ph): To a solution of diethyl malonimidate dihydrochloride (90 mmol) in dichloromethane (600 mL) was added the amino alcohol (98 mmol, 2.2 equiv) at room temperature. The reaction mixture was stirred under reflux for 48 h, and water (150 mL) was added to the reaction mixture. The crude product was extracted using dichloromethane (3 × 100 mL), and the combined organic layers were washed with brine (2 × 200 mL). Solvents were evaporated under reduced pressure, and the crude product was purified by column chromatography with DCM/MeOH