Non-Covalent Interactions Mimic the Covalent: An Electrode-Orthogonal Self-Assembled Layer

Charge-transfer events central to energy conversion and storage and molecular sensing occur at electrified interfaces. Synthetic control over the interface is traditionally accessed through electrode-specific covalent tethering of molecules. Covalent linkages inherently limit the scope and the potential stability window of molecularly tunable electrodes. Here, we report a synthetic strategy that is agnostic to the electrode’s surface chemistry to molecularly define electrified interfaces. We append ferrocene redox reporters to amphiphiles, utilizing non-covalent electrostatic and van der Waals interactions to prepare a self-assembled layer stable over a 2.9 V range. The layer’s voltammetric response and in situ infrared spectra mimic those reported for analogous covalently bound ferrocene. This design is electrode-orthogonal; layer self-assembly is reversible and independent of the underlying electrode material’s surface chemistry. We demonstrate that the design can be utilized across a wide range of electrode material classes (transition metal, carbon, carbon composites) and morphologies (nanostructured, planar). Merging atomically precise organic synthesis of amphiphiles with in situ non-covalent self-assembly at polarized electrodes, our work sets the stage for predictive and non-fouling synthetic control over electrified interfaces.


S7
Polycrystalline Pt working electrodes were purchased from CH instruments (2 mm diameter, 0.0314 cm 2 area). Pt electrodes were polished with alumina slurry and sonicated in ultrapure water.
A monolayer of benzenethiol was prepared on a polycrystalline Au disk electrode according to a previously reported procedure. 1 Commercially available benzenethiol (Sigma-Aldrich, 99%) was used without further purification. The Au electrode was immersed for 1-2 minutes in a freshly prepared 1 mM solution of benzenethiol in ultrapure water, stirring with a using magnetic bead for ~30 seconds. The Au electrode was then removed and placed in a glass vial containing 15 mL ethanol (200 proof) and stirred at 300-400 rpm for 2-3 min to remove excess unbound benzenethiol. The prepared electrode was placed in an electrochemical cell with 0.5 M KOH, and CVs were recorded at a scan rate of 10 mV s -1 with a Hg/HgO reference electrode and a Pt mesh counter electrode (Fig. S23) (Fig. S24).
Glassy carbon foil working electrodes were purchased from Goodfellow (0.3 cm 2 geometric surface area, 0.5 mm thickness). Glassy carbon foil electrodes were rinsed with ultrapure water and connected to the working electrode lead using stainless steel alligator clips.
Highly oriented pyrolytic graphite (HOPG) was purchased from SPI (brand grade SPI-2, 10 mm × 10 mm × 1 mm). HOPG working electrodes were sonicated for 1 min in CH2Cl2 and dried. Both sides of the HOPG surface were peeled off using Scotch tape as previously reported. 3,4 Excess visible graphite flasks on top layer were removed using Kimwipes wet with ultrapure water. HOPG working electrodes (1×1 cm 2 geometric surface area) were connected to the working electrode lead using stainless steel alligator clips, and 0.5 cm 2 of the electrode was submerged in the electrolyte. Ten CV cycles were recorded for each experiment at scan rate of 20 mV s -1 and 100 mV s -1 to obtain reproducible data.
Edge-plane pyrolytic graphite (EPG) electrodes were purchased from BASi (3 mm diameter, 0.0707 cm 2 area). EPG electrodes were polished with alumina slurry and sonicated in ultrapure water.
Au nanoparticles were electrodeposited on a glassy carbon rotating disk electrode (Pine Research, 5 mm diameter, 0.196 cm 2 geometric surface area, no rotation) according to a previously reported procedure. 5 A glassy carbon electrode was polished for 3 minutes with alumina slurry and 3 minutes with diamond slurry and sonicated in ultrapure water. To electrodeposit Au nanoparticles, S8 chronoamperometry was performed at -0.5 V vs Ag/AgCl for 60 seconds in 0.1 mM NaAuCl4 with N2-saturated 0.1 M H2SO4 as an electrolyte. The prepared electrode was rinsed with ultrapure water. The surface coverage of the Au nanoparticles was calculated to be 32.5% based on integration of the oxidative stripping wave in N2-saturated 0.1 M HCl (Fig. S25).

Electrolyte and Stock Solution Preparation.
High purity NaClO4 hydrate (99.99% trace metals basis, Sigma Aldrich) and Millipore Type 1, 18.2 MΩ, water were used for all aqueous electrolyte preparation. The same stock solution of 0.1 M NaClO4 was used for all comparison studies, e.g., scan rate dependence, concentration dependence. To prepare the C18-Fc stock solution, approximately 2 to 5 mg of C18-Fc or C18(C12)-Fc was added to 1 mL of ultra-pure water. The stock solution was manually shaken at room temperature to dissolve, avoiding any sonication or heating. The stock solution was stored in refrigerator and used for electrochemical measurements within 12 hours to avoid decomposition. We note that 75 μM of each monomer was examined because the limited solubility of the species in aqueous media precluded investigation at the identical concentration values utilized for MeCN in Figure 1.
For the H/D exchange SEIRAS experiment, 4 mg of C18-Fc was dissolved in 1 mL of D2O and stirred for 4 days at room temperature, resulting in a stock solution with a concentration of 6.96 × 10 -3 M. Separately, 0.61 g NaClO4 (anhydrous) was dissolved in 50 mL of D2O, resulting in electrolyte with a concentration of 0.1 M. In the spectroelectrochemical cell, 12 mL of 0.1 M NaClO4 in D2O was purged with N2 for 15 min prior to measurement. 60 to 125 µL of C18-Fc stock solution was added into the cell to obtain concentrations of 35 µM to 200 µM.

Surface-Enhanced Infrared Absorption Spectroscopy (SEIRAS)
. SEIRA spectra were recorded in an attenuated total reflection (ATR) configuration using a Nicolet iS50 FTIR spectrometer equipped with a HgCdTe (MCT) detector (Ref/01 Gain Setting) and a PIKE VeeMax III accessory (incident angle of 60˚). The Nicolet spectrometer was operated in series mode with optical velocity of 1.8988 cm s -1 . Spectra were sequentially acquired with a spectral resolution of 4 cm -1 at every 19.2(8) s interval. A single beam spectrum (32 scans) collected at the starting potential in the absence of any substrates (i.e., in the presence of 0.1 M NaClO4 in water) was used as the reference spectrum. All ATR-SEIRA spectra are reported in absorbance units defined as A=−log (I/I0), where I and I0 stand for the sample and reference single-beam spectra, respectively. Data analysis was conducted using the OMNIC version 9.12.928 software. The PIKE Jackfish J2 spectroelectrochemical cells were used.

Preparation of Au Films for Surface-Enhanced Infrared Absorption Spectroscopy.
Au films were prepared on undoped Si prisms (PIKE) using the "double deposition method" as previously detailed. 7,8 The geometric surface area of the SEIRAS film exposed to the electrolyte was 0.71 cm 2 . Following the deposition, the Au-coated Si prism was assembled into the PIKE Jackfish J2 spectroelectrochemical cell and cleaned using an electrochemical procedure. Prior to use, the Au film was cycled in 0.1 M H2SO4 from -0.10 V to 1.65 V vs Ag/AgCl at 100 mV s -1 for 5 continuous cycles and -0.10 V to -1.20 V vs Ag/AgCl at 100 mV s -1 for 5 continuous cycles. Following, the Au film was washed 3 to 5 times with ultrapure water. The electrolyte in the cell was replaced with 0.1 M NaClO4 and purged with N2 for at least 15 minutes. Prior to data collection, the Au film was cycled in 0.1 M NaClO4 from -0.10 V to 1.65 V vs Ag/AgCl at 100 mV s -1 for 5 continuous cycles and -0.10 V to -1.20 V vs Ag/AgCl at 100 mV s -1 for 5 continuous cycles. The electrolyte was replaced with fresh 0.1 M NaClO4 and purged with N2 for at least 15 minutes. Around 3 to 5 cycles of cyclic voltammetry from -0.10 V to 0.80 V vs Ag/AgCl at 2 mV s -1 were performed to obtain a stabilized signal from the Au film in 0.1 M NaClO4. Before recording the IR spectrum, the Au film was held for at least 200 sec at -0.10 V vs Ag/AgCl. All experiments were conducted under N2. All background spectra were recorded at −0.10 V vs Ag/AgCl over various time intervals.

Solid-Phase IR Measurements.
Spectra were recorded using a Nicolet iS50 FTIR spectrometer equipped with a HgCdTe (MCT) detector (Ref/01 Gain Setting). The Nicolet spectrometer was operated in single scan mode with optical velocity of 0.9494 cm s -1 . Spectra were acquired with a spectral resolution of 4 cm -1 with a single beam spectrum (32 scans). KBr (60-100 mg, FT-IR grade, ≥99% trace metals basis) pellets were prepared with a custom-built wrench-operated pellet press. Background spectra were recorded with visually transparent and stable pellets.

Cu Underpotential Deposition.
Cu underpotential deposition on Au disk electrodes and Au SEIRAS-active films used in this study were characterized by a previously reported procedure. 9 9. Numerical Frequency Calculations. Density functional theory calculations to compute numerical frequencies were performed using ORCA 5.0.1. The B3PW91 functional and its associated extended basis set were used. 10 The initial structure of C18-Fc was built in Avogadro 1.2.0. The geometries of these structures were optimized, and numerical frequencies were calculated in ORCA ( Figure S12 and Table S6). Normal modes were visualized from the output Hessian using Avogadro.

Determination of Diffusion Coefficients for C18-Fc, C18(C12)-Fc, and C2-Fc.
A 15 mL solution of 0.1 M TBAClO4 in acetonitrile was prepared and sparged with N2 in an electrochemical cell for 15 minutes. 1 mM of the analyte was added, and the electrolyte solution was stirred and sparged with N2 until the solution reached a final volume of 10 mL. Cyclic voltammetry measurements were taken at scan rates of 0.05, 0.1, 0.2, 0.5, 1, 2, and 5 V s -1 . The electrolyte solution was stirred for 2 minutes and allowed to settle for 3 minutes between each scan. Using the Randles-Ševčík equation, 11,12 the diffusion coefficient was calculated to be 1.87 × 10 -5 cm 2 s -1 for C2-Fc, 0.625 × 10 -5 cm 2 s -1 for C18-Fc, and 1.01 × 10 -5 cm 2 s -1 for C18(C12)-Fc by plotting peak current ip (A) versus the square root of scan rate ν 1/2 (V ½ s -½ ) and measuring the slope, 13 where n = 1 electron, F = Faraday constant (96485 C mol -1 ), A = 0.031415 cm 2 , C 0 = 1 × 10 -6 mol cm -3 , R = gas constant (8.314 J mol -1 K -1 ), and T = 298 K. Data for slopes are reported in Fig. S3. We note that 1 mM of each monomer was examined to obtain sufficient voltametric signal above the background capacitive feature, including those reported in Figure 1a.  Table S2. These values were used to calculate the surface tension of C18-Fc and measured using a DSA 100 Drop Shape Analyzer KRÜSS GmbH system, with a pendant drop method. C18-Fc 3.9 mg was dissolved in 1 mL ultrapure water (stock concentration of 6.78 × 10 -3 M) to prepare 200 M in ultrapure water and 0.1 M NaClO4. Measured surface tensions at 27.3C are tabulated in Table S2. The experimental values for total charge transferred were calculated by integrating the oxidative peak with respect to potential scanned over that region, which was normalized for the scan rate within the Gamry Echem Analyst software workstation. The total charge transfer was normalized for electrochemically active surface area by dividing by the area of the gold disk electrode (0.0314 cm 2 ).

Estimation of the Charge Integration for a Monolayer
Following the literature approach of estimating surface coverage via measuring the cross-sectional area of the surface-bound species 16 , the geometry optimized structure of C18-Fc (Fig. S12) was used to model the cross-sectional area as a rectangle with calculated length from ferrocene to the ammonium group, 10.33 × 10 -8 cm, and width as the diameter of ferrocene, 6.6 × 10 -8 cm. Thus, each molecule occupies 6.82 × 10 -15 cm 2 , which corresponds to a surface coverage of 2.44 × 10 -10 mol cm -2 . By multiplying by Faraday's constant, the estimated charge transferred for a monolayer of C18-Fc is 23.5 µC cm -2 .
13. Rinse Test. Au disk electrode (CH Instruments, 2 mm diameter, 0.0314 cm 2 area) electrode was polished with alumina slurry in ultrapure water and annealed as mentioned in Section 3. A stock solution of C18-Fc was prepared according to Section 4. 2.7 mg of C18-Fc was dissolved in 1 mL ultrapure water to make a stock solution of 4.7 × 10 -3 M. Two separate electrochemical cells were set up each containing 10 mL of N 2 -saturated 0.1 M NaClO4 electrolyte. In the first cell, CVs of 0.1 M NaClO4 were recorded, followed by the addition of an aliquot 0.44 mL of stock solution to obtain 200 µM C18-Fc. After recording the CV, the Au disk electrode was carefully taken out of the cell and suspended in a glass vial containing 15 mL ultrapure water without touching the bottom of the vial or the stir bar. The water was stirred for 10 min at approximately S11 450 rpm. This electrode was then placed in the second cell with 0.1 M NaClO4. CVs were recorded at a scan rate of 2 mV s -1 with Ag/AgCl as the reference and Pt mesh as the counterM electrode. This electrode was then placed back in the first cell that contained 200 µM C18-Fc and CVs were recorded in the same conditions. CVs are shown in Figures 1d and S14.

Synthetic Scheme for C18-Fc, C18(C12)-Fc, and C2-Fc Monomer Synthesis.
Scheme S1. Synthetic scheme for compounds 1 to 18. Detailed synthesis procedures are described in Section 15. All reported yields are isolated yields.

Synthetic Procedures
Preparation of 2-aminooctadecanoic acid (1). Diethyl 2-acetamidomalonate (21 g, 97 mmol) was added into 150 mL ethanol and was stirred for 30 minutes. 1bromohexadecane (35 g, 116 mmol) was added and the reaction mixture was refluxed for 24 h. After cooling, the reaction mixture was poured into ice-cold water and the precipitate was filtered out and washed with cold water. This solid was dissolved in 15 mL DMF and 100 mL HCl (36%) and stirred for 36 h. The reaction mixture was cooled to room temperature and poured into ethanol/water (2/1) and neutralized with ammonium hydroxide to obtain a solid compound. For better yield, this crude compound was crushed, dried, and washed with hexanes and used for the next reaction without further purifications. Crude yield: 21 g, 74% Data consistent with literature-reported precedent. 17  mmol) was dissolved in 120 mL tert-butanol in a 1 L round bottom flask. Around 10 g of NaOH was dissolved in 180 mL of water and added into the reaction mixture. Di-tert-butyl dicarbonate (Boc2O, 17.2 mL, 75 mmol) was added dropwise to the reaction mixture at 0-4 °C, maintaining pH ~13 for 2 hours and stirring for 12 h. After completion of the reaction, citric acid was added portion-wise, and the mixture was maintained at pH ~3 and stirred for 30 minutes. The reaction mixture was extracted with ethyl acetate/water (500 mL × 3), and the ethyl acetate layer was dried over Na2SO4 and then concentrated in vacuo. 20 mL of MeCN was added to the crude product and was kept at room temperature to obtain a light-yellow crystalline product. Yield: 16 g, 80%. 1

Preparation of 2-amino-N,N-dimethyloctadecanamide (4):
Compound 3 (9 g, 21.1 mmol) was dissolved in 120 mL of anhydrous dichloromethane and 30 mL trifluoroacetic acid was added dropwise at ~0 °C under nitrogen. The resulting mixture was stirred at room temperature for 6 h and monitored using TLC. After completion, the reaction mixture was diluted with 150 mL dichloromethane and quenched with saturated NaHCO3 in water at 0 °C (caution: gas evolution!) and extracted with dichloromethane (500 mL × 3) and water. The combined organic layer was dried over Na2SO4 and evaporated in vacuo to obtain a white solid compound. This solid compound was used for the next reaction without further purification. Yield: 6.2 g, 90%. 1

Preparation of N,N-dimethyloctadecane-1,2-diamine (5):
Compound 4 (10 g, 30.6 mmol) was mixed with anhydrous diethyl ether (150 mL). Lithium aluminum hydride (LiAlH4, 3.5 g, 91.8 mmol) was added portion-wise to the reaction mixture at ~0 °C. The reaction was stirred for 8 h and was monitored using TLC. After completion, the reaction mixture was diluted with diethyl ether (100 mL) and quenched with cold water and then by 15% NaOH solution (Caution: exothermic and gas evolution!). The obtained emulsion was extracted with diethyl ether (500 mL × 3) and water. The combined organic layer was dried over MgSO4 and concentrated in vacuo to obtain a viscous light yellow oily crude product at room temperature which solidified upon refrigeration. This product is used in the next reaction without further purification. Yield: 7.6 g, 79%. ESI-MS Calcd. for C20H44N2 [M] + , 312.3504 found 312.3502 m/z. Preparation of 12-Oxooctadecanoic acid (6): 12-Hydroxystearic acid (6 g, 20 mmol) was mixed with 10 mL dimethyl sulfoxide and Na2Cr2O7 (3.6 g, 14 mmol). In the reaction mixture, conc. H2SO4 (2.67 mL, 50 mmol) was added dropwise (Caution: exothermic reaction) and stirred at 75 °C for 2 hours and an additional 12 h at room temperature. The reaction mixture was poured into ice-cold water and stirred for 30 minutes. The reaction mixture was extracted with CH2Cl2 and water (5 × 500 mL), and the organic layer was dried over Na2SO4 and concentrated in vacuo to obtain a green-colored compound. The crude product was purified by silica gel chromatography using CH2Cl2:hexanes (1:5). Yield: 4.3 g, 72%. 1 (7): 6 (2 g, 6.7 mmol), dimethylamine hydrochloride (0.6 g, 7.4 mmol), hexafluorophosphate azabenzotriazole tetramethyl uronium (HATU, 2.8 g, 7.4 mmol) and diisopropylethylamine (DIPEA, 5.8 mL, 33.5 mmol) were dissolved in anhydrous DMF (20 mL) and stirred for 12 h. The reaction mixture was extracted with ethyl acetate and water (250 mL × 4), dried over Na2SO4, and concentrated in vacuo. The crude product was purified by silica gel chromatography using hexanes:ethyl acetate (2:1). Yield: 1.7 g, 78%. 1   Sodium cyanoborohydride (NaCNBH3, 867 mg, 13.8 mmol) was added into reaction mixture and stirred for 36 hours. The reaction was monitored using TLC, and after completion, the product was concentrated in vacuo. The residue was dissolved in CH2Cl2 and extracted with water (250 mL × 2). The crude product was purified by silica gel chromatography using CH2Cl2:CH3OH (20:1). Yield: 1.6 g, 94%. Data are consistent with literature-reported precedent. 19 (9): 8 (1.5 g, 4.6 mmol) was dissolved in anhydrous diethyl ether (40 mL). Lithium aluminum hydride (LiAlH4, 0.52 g, 13.8 mmol) was added portion-wise to the reaction mixture at ~0 °C. The reaction mixture was stirred for 5 h and was monitored using TLC. After completion, the reaction mixture was diluted with diethyl ether (100 mL) and quenched with cold water and then by 15% NaOH solution (Caution: exothermic and gas evolution!). The obtained emulsion was extracted with diethyl ether (500 mL × 2) and water. The combined organic layer was dried over MgSO4 and concentrated in vacuo. The compound was purified by silica gel chromatography using CH2Cl2:CH3OH (20:1) to obtain a viscous light yellow oily product at room temperature. Yield: 0.97 g, 68%. 1

Moles of Material on Electrode (×10 -4 , mole/cm 2 ) b
Electroactive material on Electrode (×10 -9 , mol/cm 2 ) c a stock solution concentrations in CH2Cl2 (0.8 mL), ethanol (0.15 mL) and Nafion perfluorinated resin, 5 wt %, (0.05 mL). b moles on electrode calculated from aliquot used for Zn(II)Pc (5 µL), Co(II)Pc (10 µL), and Fe(III)OEPCl (5 µL) from stock concentration of ink, which was dropcast onto glassy carbon electrode. c electroactive material calculated by integration of the redox feature associated with ZnPc (II/I), CoPc (II/I) and FeOEPCl (III/II) (Fig. S26-S28).   Figure S30. Peak assignments made based on literature precedent as referenced and described in the main text. Fig. S1. Plot of the intensity of scattered light (kilo counts per sec, kCnt s -1 ) obtained with various concentrations of C18-Fc in water. Intersection of two lines corresponds to critical micelle concentration.   S3. Peak current of cyclic voltammograms obtained at varying scan rates for 1 mM C18-Fc (blue triangles, anodic peak current; green triangles, cathodic peak current), 1 mM C2-Fc (black squares, anodic peak current; red circles, cathodic peak current), and 1 mM C18(C12)-Fc (purple diamonds, anodic peak current; yellow triangles, cathodic peak current) in 0.1 M TBAClO4 in MeCN, working electrode as Au disk electrode and Ag/AgCl reference electrode. Data used to estimate the diffusion coefficients for the three molecules.     While, in principle, a ΔEp of zero is expected for a non-interacting, surface-bound redoxactive moiety exhibiting fast electron transfer kinetics at the scan rates examined, 28,29 literature reports suggest that the electronic interaction of Au with Fc positioned at short distances to the Au electrode (within one to two methylene units from the S-Au linkage) leads to observed CV peak distortions and separation. [30][31][32][33][34][35][36][37][38][39][40] While these effects convolute the elucidation of a well-defined electron transfer rate to C18-Fc from the Au electrode via the Laviron formalism 28,29,[41][42][43][44][45] and Marcus Theory, 46 the similarity between the ΔEp observed in this work and covalently-bound ferrocene to Au suggest that C18-Fc are similarly immobilized at the Au surface.    to the geometric surface area of each of the Au electrode materials examined. Cu underpotential deposition on Au estimates the electrochemically active surface area. 47,48 We observe that the integrated charges for the Cu UPD feature normalized for the geometric surface area of each Au material (SEIRAS-active Au film versus the Au disk electrode) over the range of 0.34 to 0.50 V vs SHE are roughly similar. These results demonstrate that the integrated charge of 24 μC cm -2 reported for the C18-Fc wave observed on the SEIRAS-active Au film in Figure 2c is representative of the surface area of the nanostructured Au film.    (Figure S13), suggesting that the self-assembled layer does not grow upon the application of multiple potential cycles. Therefore, we attribute the slight increase in the charge not due to multilayer formation but rather due to structural changes of the SEIRAS film, leading to an overall roughening of the SEIRASactive film surface. The simultaneously collected SEIRA spectra reveal near-identical spectroscopic features with the first scan on an absolute basis, Figure S20b. The differential spectra, Figure S20c (integrations shown in Figure S21) are nearly identical to that observed in the initial, first scan, Figure 2b. The sole difference is the integration of the ν(CH) of the aliphatic tail, Figure S21c; it is reversible with respect to application of the potential. These results suggest that, after the first scan, the interfacial structure of the self-assembled layer converges to a structure in which the aliphatic tail region only reorients in concert with Fc oxidation (Scheme 3). The spectroscopic data are consistent with electrochemical data; C18-Fc localizes at the Au interface upon addition to the bulk solution via electrostatic templating and persists upon the application of positive potential due to the secondary hydrophobic interaction between the aliphatic tails on the timescale of the slow CV scan.   Figure S19, second consecutive scan. Blue represents the oxidative trace; red represents the reverse, reductive trace.     Electrode modified using a dropcast method with 5 µL ink prepared from 1.3 mg FeOEPCl/mL stock solution (see Table S7).  Table S7).

Fig. S28.
Cyclic voltammogram of glassy carbon disk electrode modified with zinc phthalocyanine (ZnPc) recorded in electrolyte solution of 0.1 M TBAP in acetonitrile at 100 mV s -1 with negative direction of scan. Electrode modified using a dropcast method with 5 µL ink prepared from 7.9 mg ZnPc/mL stock solution (see Table S7).