Rapidly Adaptive All‐covalent Nanoparticle Surface Engineering

Abstract Emerging nanotechnologies demand the manipulation of nanoscale components with the same predictability and programmability as is taken for granted in molecular synthetic methodologies. Yet installing appropriately reactive chemical functionality on nanomaterial surfaces has previously entailed compromises in terms of reactivity scope, functionalization density, or both. Here, we introduce an idealized dynamic covalent nanoparticle building block for divergent and adaptive post‐synthesis modification of colloidal nanomaterials. Acetal‐protected monolayer‐stabilized gold nanoparticles are prepared via operationally simple protocols and are stable to long‐term storage. Tunable surface densities of reactive aldehyde functionalities are revealed on‐demand, leading to a wide range of adaptive surface engineering options from one nanoscale synthon. Analytically tractable with molecular precision, interfacial reaction kinetics and dynamic surface constitutions can be probed in situ at the ensemble level. High functionalization densities combined with rapid equilibration kinetics enable environmentally adaptive surface constitutions and rapid nanoparticle property switching in response to simple chemical effectors.


General experimental procedures
Unless stated otherwise, all reagents were purchased from commercial sources and used without further purification. Dry solvents were obtained by means of a MBRAUN MB SPS-800TM solvent purification system, where solvents were passed through filter columns and dispensed under an argon atmosphere. Flash column chromatography was performed using Geduran® Si60 (40-63 µm, Merck, Germany) as the stationary phase. Thin-layer chromatography (TLC) was performed on pre-coated silica gel plates (0.25 mm thick, 60F254, Merck, Germany) and observed under UV light (lmax 254 nm) or visualized by staining with a basic potassium permanganate solution, followed by heating. Nanoparticle micrographs were obtained using a JEM 2010 transmission electron microscope (TEM) on samples prepared by deposition of one drop of nanoparticle suspension on holey carbon films supported on a 300 mesh Cu grid (Agar Scientific®). Nanoparticle diameters were measured automatically using the software ImageJ. The images were first converted to black and white images using the "Threshold" function; the area of each nanoparticle was measured using the "Analyze particles" function; particles on edges were excluded. 1 H, 13 C, and 19 F NMR spectra were recorded on Bruker Avance II 300, 400 and 500 instruments, at a constant temperature of 25 °C. All 19 F and 13 C spectra were recorded using proton decoupling. Chemical shifts are reported in parts per million (ppm) from low to high field; 1 H and 13 C chemical shifts are referenced to the literature values for chemical shifts of residual non-deuterated solvent, with respect to tetramethylsilane; 19 F Chemical shifts are referenced to CF3Cl (0.00 ppm) as external standard. Standard abbreviations indicating multiplicity are used as follows: bs (broad singlet), d (doublet), dd (doublet of doublets), dt (doublet of triplets), m (multiplet), s (singlet), t (triplet), tt (triplet of triplets), q (quartet), quint (quintuplet), J (coupling constant). All spectra were analyzed using MestReNova (Version 9.0.0-12.0.02). All melting points were determined using a Stuart SMP30 Melting Point Apparatus and are reported uncorrected. A Beckman Coulter Avanti J-25 centrifuge was used equipped with the JA-25.50 rotor. Thermogravimetric analysis is performed using a Staton Redcroft STA-780 simultaneous TG-DTA: room temperature to 900 ºC, which is connected to a Rheometric Scientific STA SID System Interface.

2-(3-Fluoro-4-(pentyloxy)phenyl)-1,3-dioxolane (5)
Compound 6 (0.380 g, 1.80 mmol) was dissolved in CH2Cl2 (10 mL) in a round-bottomed flask. The solvent was removed under reduced pressure to afford an oil. Ethylene glycol (40 mL) was added to the flask. The mixture was stirred to give a suspension, and then ptoluenesulfonic acid (0.270 g, 1.44 mmol) was added to the solution. The flask was heated to 80 ºC in the water bath of a rotary evaporator and carefully evacuated for short periods to remove water while retaining ethylene glycol. The reaction was monitored by 1 H and 19 F NMR spectroscopy. Once the reaction had gone to completion, the solution was dissolved in CH2Cl2 (150 mL) and washed with saturated aqueous NaHCO3 (3 × 150 mL) and brine (3 × 150 mL). The organic layer was dried over MgSO4, then volatiles were removed under reduced pressure to obtain the final product 5 as a white solid (0.375 g, yield: 82%).

1-(3-Fluoro-4-(pentyloxy)phenyl)-N-(4-fluorobenzyl)methanimine (S10)
Compound 6 (0.300 g, 1.43 mmol) and 4-fluorobenzylamine (0.180 g, 1.43 mmol) were dissolved in MeOH (6 mL). The mixture was heated at 50 °C overnight. Volatiles were then removed under reduced pressure and the residue obtained was re-dissolved in CH2Cl2 and dried over MgSO4. Volatiles were again removed under reduced pressure and the solid obtained was re-dissolved in hexane. The solution was cooled to -18 °C in a freezer, during which time a pale-yellow solid precipitated from the solution. The mixture was filtered to remove hexane and obtain the final product as a pale-yellow solid (0.170 g, yield 38%).

2-(3-fluoro-4-(2-methoxyethoxy)phenyl)-1,3-dioxolane (7)
Compound 8 (1.00 g, 1.00 mmol) was dissolved in CH2Cl2 (10 mL) in a round-bottomed flask. The solvent was removed under reduced pressure to afford an oil. Ethylene glycol (50 mL) was added to the flask. The mixture was stirred to give a suspension, and then ptoluenesulfonic acid (133 mg, 700 µmol) was added to the solution. The flask was heated to 80 ºC in the water bath of a rotary evaporator and carefully evacuated for short periods to remove water while retaining ethylene glycol. The reaction was monitored by 1 H and 19 F NMR spectroscopy. Once the reaction had gone to completion, the solution was dissolved in CH2Cl2 (150 mL) and washed with saturated aqueous NaHCO3 (3 × 150 mL) and brine (3 × 150 mL). The organic layer was dried over MgSO4, then volatiles were removed under reduced pressure to obtain a colourless oil. The residue was purified by flash column chromatography (SiO2, CH2Cl2/triethylamine 98:2 v/v) to obtain the final product 7 as a white solid (0.81 g, yield: 66%).

Ex situ NMR Characterization of AuNP-1: oxidative ligand desorption
Ligand desorption using a mild oxidising agent allows analysis of the released molecular species in bulk solution. A colloidal solution of AuNP-1 (2.0 mg) in CDCl3 was treated with iodine (2 mg) then 1 H and 19 F NMR spectra were recorded immediately ( Figure S3).

Nanoscale characterization of AuNP-1
Thermal gravimetric analysis (TGA) and differential thermal analysis (DTA) were performed by heating AuNP-1 powder, under a stream of air (24 mL min -1 ), 20-900 °C with a 5 min isotherm at 50 ºC. A blank was recorded under the same conditions and the mass subtracted to eliminate the buoyancy effect. Thermal decomposition of the surface-bound monolayer resulted in progressive mass loss as temperature increased above the onset temperature (Tm) at 325 °C. Assuming an isotropic gold core (d = 5.6 nm, Figure S5) the organic mass lost between 50-900 °C was used to estimate molar weight and number of ligands per nanoparticle ( Figure S4 and Table S1).   Figure S5. Representative TEM images and histogram of size distribution as found through analysis of multiple images for AuNP-1 (<d> = 5.6 ± 0.5 nm). The reaction was allowed to cool to room temperature, then quenched by adding Et2O (5 mL) to achieve nanoparticle precipitation. The mixture was transferred to glass vials then centrifuged (1446 ×g rcf, 4 °C, 10 min). The supernatant was removed using a glass pipette. The black solid was subsequently washed with MeOH using the following procedure: nanoparticles were dispersed in MeOH (7 mL), sonicated for 15 min, centrifuged (1312 ×g rcf, 5 °C, 10 min) and the supernatant was removed using a glass pipette. The same operation was performed in Et2O (7 mL) and the entire process was repeated two further times. Evaporation under reduced pressure afforded AuNP-2 as a black solid (25.8 mg). Mean diameter (TEM, Figure S9): 4.9 ± 0.6 nm (dispersity: 12 %).

Ex situ NMR Characterization of AuNP-2: oxidative ligand desorption
Ligand desorption using a mild oxidising agent allows analysis of the released molecular species in bulk solution. A colloidal solution of AuNP-2 (5.6 mg) in CDCl3 was treated with iodine (2 mg), then 1 H and 19 F spectra were recorded immediately ( Figure S7).

Nanoscale characterization of AuNP-2
Thermal gravimetric analysis (TGA) and differential thermal analysis (DTA) were performed by heating AuNP-2 powder, under a stream of air (24 mL min -1 ), over the range 20-900 °C with a 10 min isotherm at 50 ºC. A blank was recorded under the same conditions and the mass subtracted to eliminate the buoyancy effect. Thermal decomposition of AuNP-2 surfacebound monolayer resulted in progressive mass loss as temperature increased above the onset temperature (Tm) at 315 °C. Assuming an isotropic gold core (d = 4.9 nm, Figure S9) the organic mass lost between 50-900 °C was used to estimate molar weight and number of ligands per nanoparticle ( Figure S8 and Table S2).   Figure S9. Representative TEM images and histogram of size distribution as found through analysis of multiple images for AuNP-2 (<d> = 4.9 ± 0.6 nm). Counts diameter / nm S17

Identification and elimination of monolayer impurities
Butylated hydroxytoluene (BHT), a radical scavenger, was added to avoid cluster-catalyzed oxidation of acetals by reactive oxygen species that were either endogenous to the THF solvent or generated during the synthetic procedure. [3] Nevertheless, when characterizing an early batch of AuNP-1 by 19 F NMR spectroscopy ( Figure S10), oxidative ligand desorption (procedure as described in Section 3.1) revealed a signal at -134.55 ppm (blue arrow, Figure S10b), which corresponds to neither the acetal disulfide 12 nor the analogous aldehyde S4. Figure S10. 19 F NMR spectra (470 MHz, CDCl3) of: a) AuNP-1; b) oxidative ligand desorption performed on a solution of purified AuNP-1; c) acetal model compound 5; d) aldehyde model compound 6; e) mixture of model acyclic acetal S11 or S12 (-134.55 ppm) and cyclic acetal 5 (-134.68 ppm).
In order to identify the source of this unknown peak at -134.55 ppm, acetal formation was studied using model compound 6. p-Toluenesulfonic acid (47 mg, 0.25 mmol) was added to a solution of compound 6 (75 mg, 0.32 mmol) in ethylene glycol (18 mL). The reaction mixture was heated to 50 °C under reduced pressure for 3 h. The reaction was cooled to room temperature and extracted with CH2Cl2 (3 × 50 mL) containing Et3N (1% v/v). The organics were washed with a saturated solution of NaHCO3 (2 × 40 mL) then dried over MgSO4. Volatiles were removed under reduced pressure to obtain a pale-yellow oil (57 mg). Analysis by 19 F NMR spectroscopy ( Figure S10e) revealed two signals, corresponding to the expected acetal model compound 5 (at -134.68 ppm) and an unknown compound at the chemical shift (-134.55 ppm) observed for the impurity in the nanoparticle experiment ( Figure  10b). Although isolation of a pure sample of the unknown compound was unsuccessful, further characterization of the mixture obtained from the model compound experient by 1 H NMR spectroscopy ( Figure S11c) identified the impurity as an acyclic acetal compound, hemi-acetal S11 or bis-ethan-1-ol acetal S12. This assignment was reinforced by incubation of the mixture with excess of CF3CO2H at 25 ºC, which resulted in complete disappereance of the resonances for the unknown species after 2 h, with the only new peaks corresponding to the aldehyde model compound 6 and traces of cyclic acetal model compound 5 ( Figure S11d). To avoid formation of the acyclic acetal impurity, it is crucial to limit the duration of the acetal formation procedure to less than 2 h and to maintain a reaction temperature above 80 ºC. Following this procedure carefully, eliminated the acyclic acetal impurity from all further batches of acetal functionalized AuNPs. S19

Exhaustive and partial on-nanoparticle acetal hydrolysis
A stock solution of CF3CO2H was freshly prepared for each experiment in the reaction solvent mixture with 4-fluorotoluene as internal standard (5.00 mM) and concentration measured by 19 F NMR.

Partial acetal hydrolysis from AuNP-1
A colloidal solution of AuNP-1 containing 4-fluorotoluene as internal standard (5.00 mM) was prepared in THF/D2O (9:1 v/v, 600 µL) giving 5.90 mM in terms of surface-bound acetal 1. This solution was heated to 50 °C, then an aliquot of the CF3CO2H stock solution (13.5 µL) was added, giving a final concentration of AuNP-1 (5.77 mM) and CF3CO2H (23.2 mM). This mixture was incubated at 50 °C and the reaction was followed by 19 F NMR spectroscopy. A new broad nanoparticle-bound signal appeared at -133.07 ppm, corresponding to nanoparticle-bound aldehyde 3 ( Figure S14). After 2 h, the solution was removed from heating, and triethylamine (2.4 μL) was added to neutralise CF3CO2H. Nanoparticles were precipitated by adding Et2O/MeOH (4:1 v/v, 10 mL). The black solid recovered was resuspended in Et2O/MeOH (4:1 v/v, 15 mL), sonicated for 10 min, then recollected by centrifugation (1446 ×g rcf, 4 °C, 10 min). This operation was repeated a further twice. Traces of volatile solvents were removed from the purified residue under a stream of compressed air.

S22
The composition of the nanoparticle-bound monolayer was determined by deconvolution of the relevant signals in the in situ 19 F NMR spectrum ( Figure S14), revealing 60% acetal and 40% aldehyde ligands (AuNP-10.630.4).

Partial acetal hydrolysis from AuNP-2
A stock solution of CF3CO2H was freshly prepared in distilled DMF/D2O 9:1 v/v with 4fluorotoluene as internal standard (5.00 mM) and concentration measured by 19 F NMR A colloidal solution of AuNP-2 containing 4-fluorotoluene (5.00 mM) was prepared in distilled DMF/D2O (9:1 v/v, 600 µL) giving 7.33 mM in terms of surface-bound acetal 2. Then, an aliquot of the CF3CO2H stock solution (1.022 M, 11.7 μL) was added, giving a final concentration of AuNP-2 (7.26 mM) and CF3CO2H (20.1 mM). This mixture was incubated at room temperature and the reaction was followed by 19 F NMR spectroscopy. A new broad nanoparticle-bound signal appeared at -132.95 ppm, corresponding to nanoparticle-bound aldehyde 4 ( Figure S15). After 100 min, the nanoparticles were precipitated by adding Et2O/MeOH (4:1 v/v, 10 mL) to stop the hydrolysis reaction. The black solid was recovered by centrifugation (1446 ×g rcf, 4 °C, 10 min), then resuspended in Et2O/MeOH (4:1 v/v, 15 mL), sonicated for 10 min and recollected by centrifugation. This operation was repeated further twice. Traces of volatile solvents were removed from the purified residue under a stream of compressed air. The composition of the nanoparticle-bound monolayer was determined by deconvolution of the relevant signals in the in situ 19 F NMR spectrum ( Figure S15), revealing 60% acetal and 40% aldehyde ligands (AuNP-20.640.4).

Hydrolysis of acetal model compounds
A stock solution of CF3CO2H was freshly prepared for each experiment in distilled DMF/D2O 97:3 v/v with 4-fluorotoluene as internal standard (5.00 mM) and concentration measured by 19 F NMR

Hydrolysis of acetal model compound 5
A solution of compound 5 containing 4-fluorotoluene (5.00 mM) as internal standard was prepared in freshly distilled DMF/D2O (97:3 v/v, 600 µL). This mixture was heated to 50 °C, then an aliquot of the CF3CO2H stock solution (1.059 M, 11.5 µL) added, giving final concentrations of 5 (4.86 mM) and CF3CO2H (20.4 mM). This mixture was incubated at 50 °C and the reaction was followed by 19 F NMR spectroscopy. A new signal appeared at -133.97 ppm, corresponding to aldehyde 6 ( Figure S16). No further changes were observed after 1.5 h.

Hydrolysis of acetal model compound 7
A solution of compound 7 containing 4-fluorotoluene (5.00 mM) as internal standard was prepared in freshly distilled DMF/D2O (97:3 v/v, 600 µL). This mixture was heated to 50 °C, then an aliquot of the CF3CO2H stock solution (1.020 M, 12.5 µL) added, giving final concentrations of 7 (5.50 mM) and CF3CO2H (20.4 mM). This mixture was incubated at 50 °C and the reaction was followed by 19 F NMR spectroscopy. A new signal appeared at -133.67 ppm, corresponding to aldehyde 8 ( Figure S17). No further changes were observed after 1.5 h.

Kinetic studies of acetal hydrolysis 7.1 Experimental protocol for kinetic experiments
Reactions were carried out in mixtures of freshly distilled DMF and D2O, at 50 °C in an NMR tube. The reaction solvent mixture was first prepared, and 4-fluorotoluene added as an internal standard of known concentration (5.00 mM). Concentrations of all fluorine-containing species were determined by quantitative 19 F NMR spectroscopy by comparison to the signal for 4-fluorotoluene.
A stock solution of CF3CO2H was freshly prepared in the reaction solvent containing 4fluorotoluene as internal standard (5.00 mM) and concentration measured by quantitative 19 F NMR spectroscopy. A solution of AuNP or model compound of concentration ca. 5 mM was prepared in the reaction solvent (600 µL). This solution was incubated at 50 °C for 10 min before hydrolysis was initiated by adding an aliquot of the CF3CO2H stock solution so as to reach the desired concentration of acid. Reactions were followed by quantitative 19 F NMR spectroscopy until the acetal peak was no longer detectable. Increasing the water concentration to 10% D2O/DMF, the hydrolysis of AuNP-2 was complete after only 100 min at 50 °C ( Figure S18). Quantitative analysis afforded a rate constant ( Table  S4) that is consistent with a first-order dependence on water concentration. Likewise, nanoparticle-bound acetal hydrolysis kinetics also showed a first-order dependence on acid concentration ( Figure S19, Table S5).

One-step dynamic covalent modification of acetal-functionalized nanoparticles with nucleophilic modifiers
Concentrations of all fluorine-containing species were determined by quantitative 19 F NMR in the presence of 4-fluorotoluene as an internal standard of known concentration. A stock solution of CF3CO2H was prepared in 9:1 v/v distilled DMF/D2O with 4-fluorotoluene as internal standard (5.00 mM) and concentration measured by 19 F NMR.

Rapid reversible solvophilicity switching in one phase
AuNP-4 (ca. 1 mg) was placed in a glass vial with water (ca. 1 mL), in which they remained insoluble. Solid NaHSO3 was added (ca. 2 mg) followed by sonication for 5 min, resulting in the dissolution of all nanoparticle material, consistent with the formation of negatively charged bisulfite adduct AuNP-15. Rapid decomposition back to the starting aldehyde was achieved on addition of solid NaHCO3 (4 mg per 1 mg of AuNPs), resulting in complete re-precipitation of all nanoparticle material after 5 min. After washing with water to remove the excess of salts, the resulting black solid was solubilized organic solvent (CH2Cl2) indicating the recovery of the starting material AuNP-4.

Rapid reversible biphasic switching
AuNP-4 (ca. 2 mg) was dissolved in CH2Cl2 (ca. 1 mL) in a glass vial and water (ca. 1 mL) was added with the consequent formation of two phases (left-hand image, Figure 3b). Solid NaHSO3 (ca. 5.0 mg) was then added and the mixture was shaken by hand for 5 min, resulting in the transfer of the nanoparticles to the aqueous phase, indicating the formation of negatively charged bisulfite adduct AuNP-15 (middle image, Figure 3b). Finally, solid NaHCO3 (ca. 10 mg) was added, the biphasic mixture was shaken by hand (5 min), resulting in the transfer of the nanoparticles back to the organic phase, indicating the recovery of the starting material AuNP-4 (right-hand image, Figure 3b).

NMR characterization of reversible biphasic switching
AuNP-4 (ca. 5 mg) was dissolved in CDCl3 (600 µL), and a 1 H NMR spectrum recorded ( Figure  S26a). The mixture was transferred to a glass vial and D2O (600 µL) added with the consequent formation of two phases. Solid NaHSO3 (11.0 mg) was then added and the mixture was shaken by hand for 5 min, followed by 5 min sonication, resulting in formation of an emulsion. Evaporating the organic solvent under a stream of air produced a clear dark reddish/brownish solution of AuNPs in D2O. Analysis by 1 H NMR spectroscopy indicated formation of the bisulfite adduct AuNP-15 ( Figure S26b).
To the solution of AuNP-15 in D2O was then added CDCl3 (600 µL), followed by solid NaHCO3 (19 mg). The mixture was shaken by hand (5 min) followed by 5 min of sonication, resulting in formation of an emulsion. Both solvents were evaporated under a stream of air. The black solid was washed with deionised water to eliminate the excess of salts, dried under a stream of air, then solubilized in CDCl3 (600 µL). The 1 H NMR spectrum of the resulting solution ( Figure S26c) indicated the presence of AuNP-4, confirming the reversibility of the switching process.              Figure S44. 1 H NMR spectrum of compound S10 (400 MHz, CDCl3). Figure S45. 13 C NMR spectrum of compound S10 (101 MHz, CDCl3).