Spontaneous Aryldiazonium Film Formation on 440C Stainless Steel

Nov 7, 2014 - I&EC Process Design and Development · - I&EC Fundamentals · - Product Research & ..... (26-32) A satellite peak, labeled “Fe2O3 sat. ...
1 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Langmuir

Spontaneous Aryldiazonium Film Formation on 440C Stainless Steel in Nonaqueous Environments Leo J. Small,* Michael R. Hibbs, and David R. Wheeler Sandia National Laboratories, Albuquerque, New Mexico 87185, United States ABSTRACT: The ability of three aryldiazonium salts to spontaneously assemble onto the surface of type 440C stainless steel is investigated in acetonitrile (ACN) and the model hydraulic fluids tributyl phosphate (TBP) and hexamethyldisiloxane (HMDS). Competition between native oxide formation and organic film growth at different diazonium salt concentrations is monitored by electrochemical impedance spectroscopy. At 1 mM diazonium salt, 70% of total assembly is complete within 10 min, though total surface coverage by organics is limited to ≈0.15 monolayers. Adding HCl to the electrolyte renders native oxide formation unfavorable, yet the diazonium molecules are still unable to the increase surface coverage over 1 M−10 μM HCl in solution. X-ray photoelectron spectroscopy confirms preferential bonding of organic molecules to iron over chromium, while secondary ion mass spectroscopy reveals the ability of these films to self-heal when mechanically removed or damaged. Aging the diazonium salts in these nonaqueous environments demonstrates that up to 90% of the original diazonium salt concentration remains after 21 days at room temperature, while increasing the temperature beyond 50 °C results in complete decomposition within 24 h, regardless of solvent−salt combination. It is concluded that the investigated diazonium molecules will not spontaneously form a continuous monolayer on 440C stainless steel immersed in ACN, TBP, or HMDS.

■

INTRODUCTION Electrografting aryldiazonium salts provides a convenient means by which designer molecules may be covalently bonded to surfaces.1−4 In a typical reaction, an electron is transferred from the surface to the positively charged nitrogen group, generating nitrogen gas and a phenyl radical. With a given probability, the phenyl radical may bond to the surface. Many reports have shown the ability of aryldiazonium salts to be electrografted spontaneously or under dc bias to carbon (graphite, glassy carbon), 5−7 semiconductors (silicon, GaAs),8,9 and a variety of engineering and precious metals such as iron, nickel, copper, platinum, and gold.10−15 Typically, electroreduction is performed in acidic to neutral aqueous solutions or in acetonitrile (ACN). Aryldiazonium films have been shown to be effective at preventing corrosion on iron and steel surfaces.11,12,16,17 Previous studies have utilized bias directed assembly onto iron, mild steel, and 304L or 316 stainless steel in dilute sulfuric acid or in ACN.11,16−18 Spontaneous assembly has only been demonstrated for iron surfaces immersed in dilute sulfuric acid or ACN.11,13 As a deposition method, spontaneous assembly of diazonium films is attractive, as it provides a mechanism whereby corrosion-inhibiting films may self-heal, provided the diazonium salt is stable and contained in the working fluid. Yet to be explored is the ability of aryldiazonium salts to spontaneously graft to stainless steel surfaces. Owing to their chromium content, stainless steels are distinct from iron in their ability to rapidly form a dense native oxide. As diazonium salts do not efficiently graft onto oxide surfaces, diazonium film growth may be limited by native oxide formation.19 Some © XXXX American Chemical Society

reports, however, have shown it possible to graft diazoniums to iron oxide or silicon oxide nanoparticles.20,21 Under the reducing conditions used in previous studies on 304L and 316 stainless steel, oxide formation was of minimal concern. The three distinct diazonium salts used in this study are outlined in Figure 1. 1 is commercially available and provides a convenient means by which to characterize surface coverage via a subsequent reduction of the nitro groups to amino groups. 2 provides a waxy, hydrophobic surface, while 3 results in a hydrophobic, lipophobic surface. All films provide an electrically insulating surface which have been shown to impede electron transfer from electrode to electrolyte.7

■

EXPERIMENTAL SECTION

Synthesis of 4-Nitrophenyldiazonium Tetrafluoroborate, 1. Compound 1 was purchased from Sigma-Aldrich and was recrystallized from water before its use. 1H NMR (500 MHz, DMSO-d6): δ 8.92 (d, J = 9.3 Hz, 2H), 8.71 (d, J = 9.3 Hz, 2H). Synthesis of 4-Hexylphenyldiazonium Tetrafluoroborate, 2. Compound 2 was prepared according to a previously published procedure.22 1H NMR (500 MHz, CDCl3): δ 8.46 (d, J = 8.9 Hz, 2H), 7.54 (d, J = 8.9 Hz, 2H), 2.74 (t, J = 7.8 Hz, 2H), 1.58 (m, 2H), 1.27 (m, 6H), 0.84 (t, J = 7.0 Hz, 3H), Synthesis of 4-(Heptadecafluorooctyl)phenyldiazonium Tetrafluoroborate, 3. 4-(Heptadecafluorooctyl)aniline (3.00 g, 5.87 mmol) was dissolved in hexafluoroisopropanol (20 mL) and trifluoroacetic acid (0.9 mL). The solution was cooled to 0 °C, and Received: September 10, 2014 Revised: November 5, 2014

A

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

used for the electrochemical test: 0.1 M tetrabutylammonium tetrafluoroborate (n-Bu4NBF4, 99% Sigma-Aldrich) in acetonitrile, 0.5 M n-Bu4NBF4 in TBP, or 0.5 M tetrabutylammonium acetate (>97% Sigma-Aldrich) in the HDMS solution. The working electrodes were precision ground 1/8 in. diameter hardened 440C stainless steel rods (Rockwell C55 hardness, McMaster-Carr, Santa Fe Springs, CA, composition in wt %: 16−18% Cr, 0−0.75% Ni, 0.95−1.2% C, 0−1% Mn, 0−0.5% Cu, 0−0.75% Mo, 0−1% Si, 0−0.04% S and P, balance Fe). The working electrode was polished successively with 220, 320, and 400 silicon carbide, 1.0 μm alumina polish, and 0.05 μm alumina polish to attain a flat, smooth, mirror finish. After each polishing stage the electrode was sonicated in deionized water for 5 min, rinsed in ethanol, and dried under nitrogen. To define the working electrode area, each electrode was sleeved with polyolefin heat shrink tubing. Directly prior to use, working electrodes were polished with 0.01 μm alumina, rinsed in ethanol, and dried under nitrogen. A platinum wire counter electrode was used in all solvents. In acetonitrile, the reference electrode consisted of a silver wire in 10 mM AgNO3 in 0.1 M nBu4NBF4. For TBP and HMDS, a platinum pseudo-reference electrode was used. For determination of surface coverages, electrodes were sonicated in similar solvent for 15 min just prior to electrochemical measurements. For a typical measurement, the freshly prepared stainless steel electrode was immediately placed in the electrochemical test cell, and the open circuit potential (OCP) and impedance spectra were recorded. After the electrode was rinsed in the solvent (no salt) used for the electrochemical test, then it was placed in the diazonium solution of the corresponding solvent. Periodically, the electrode was removed from the diazonium-containing solution, rinsed in similar solvent, dried under nitrogen, and electrochemically interrogated. Electrochemical impedance spectroscopy (EIS) measurements were recorded at the OCP with 10 mV RMS over 100 kHz−10 mHz using a Solartron Modulab instrument. The resulting data were fit with equivalent circuit models using the complex nonlinear least-squares method in the software ZPlot (Scribner Associates, South Pines, NC). The dc measurements were recorded using an Epsilon EC potentiostat (BASi, Inc., West Lafeyette, IN). X-ray photoelectron spectroscopy (XPS) was performed using a Kratos Axis Ultra DLD instrument with a monochromatic Al Kα (1486.6 eV) source. Full survey spectra were collected with an analyzer pass energy of 160 eV and a step size of 1 eV. High-resolution spectra were collected with an analyzer pass energy of 20 eV and step sizes of 0.1 eV. The analyzer was used in hybrid mode with a large spot size of 300 μm by 700 μm, elliptically. Data from control samples of steel exposed to neat TBP or TBP with 5 mM NaBF4 were fitted using a Gaussian−Lorentizan fitting routine. Subsequently, spectra from steel exposed to diazoniums solutions were fitted, constraining peak locations to within ±0.2 eV, and peak width (fwhm) to within ±0.1 eV of those in the control spectra. The Gaussian−Lorentzian ratio for a given peak was unchanged across all samples.

Figure 1. Schemes for spontaneous surface functionalization using 1 (4-nitrophenyl), 2 (4-hexylphenyl), and 3 (4-heptadecafluorooctylphenyl). a solution of butyl nitrite (0.66 g, 6.46 mmol) in hexafluoroisopropanol (2 mL) was added dropwise. After allowing the reaction to stir for 45 min at 0 °C while protecting from light, it was poured into a rapidly stirred solution of HBF4 (125 mL, 10%). After mixing thoroughly, the mixture was filtered and the solid material was rinsed with cold deionized water and then dried under vacuum at room temperature over P2O5 (protected from light). The crude product was recrystallized from acetone/ethanol (1:1) to yield 3 as an off-white solid (2.32 g, 65%). 1H NMR (500 MHz, DMSO-d6): δ 8.92 (d, J = 9.2 Hz, 2H), 8.38 (d, J = 9.2 Hz, 2H). 1, 2, and 3 were dissolved at 0.1, 1, or 10 mM in either ACN (>99.96%, Biotech grade, Sigma-Aldrich), TBP (99% Sigma-Aldrich), or a solution of 80 vol % HMDS (NMR grade, >99.5%), 10 vol % acetone, and 10 vol % 2-propanol (cosolvents were necessary to dissolve the diazonium salts), hereafter referred to as “HMDS”. All solvents were deaerated with ultrahigh-purity argon. 1 H NMR spectra of the compounds were obtained on a Bruker 500 MHz spectrometer using 5 mm o.d. tubes. Sample concentrations were about 5 wt % (w/v) in CDCl3 for 3 and DMSO-d6 for 1 and 2. The aging study was performed by preparing solutions (4−10 mM) of 1, 2, and 3 in the solvents described above under argon. Neopentyl alcohol was used as an internal standard (also 4−10 mM), and the solutions were protected from light and stored at room temperature, in glassware, and never exposed to stainless steel. Solutions of 1, 2, and 3 in TBP were also stored in ovens at 50 and 75 °C. At various time intervals, samples were taken via syringe through septa on the sample vials, and the samples were mixed with the appropriate solvent for 1 HNMR analysis. Degradation of the diazonium compounds was determined by comparing the ratios of the areas of the peaks from the aryl protons to the peaks from the methylene in the internal standard. Electrochemical measurements occurred in a standard threeelectrode cell kept under argon blanket. Depending on the solvent used to spontaneously deposit the diazonium molecules, the corresponding solvent in the following deaerated electrolytes was

■

RESULTS Electrochemical Evaluation. Impedance spectra and equivalent circuit fits typical of the steel electrodes are presented in Figure 2. The equivalent circuit used to describe the impedance response consists of the bulk solution resistance, Rs, the interfacial capacitance, modeled by a constant phase element, CPEfilm, the charge transfer resistance, Rfilm, and a Warburg element, Wi, characteristic of the diffusion of ions in solution. The capacitive or diffusion-limited low-frequency response suggests an interface not undergoing active dissolution, as would be seen under acidic conditions. Rather, the same response and circuit model have been used for glassy carbon and precious metals in nonaqueous environments.23,24 Across all samples, the magnitude of both the impedance and phase increased with respect to time. Specifically, CPEf ilm decreased in magnitude, while distribution of capacitances in CPEfilm became more uniform, and Rfilm increased in magnitude. B

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

attained within 10 min. As shown in Figure 3D, the choice of solvent has little effect on the evolution of Rfilm with respect to time, for a given diazonium salt. Figures 4A,B demonstrate that higher diazonium salt concentrations lead to Rfilm growing more rapidly, though the

Figure 2. Impedance response and fits to equivalent circuit (inset) for stainless steel at the OCP in 0.5 M n-Bu4NBF4 in TBP after being immersed in TBP with 1 mM 3 for varying times.

At long times, Wi could no longer be fitted to the impedance response. This behavior is consistent with the formation of an electrically insulating film on the electrode surface. To evaluate the effect of diazonium salt, concentration, and solvent on the growth of the observed film, the circuit element Rfilm was evaluated. The effect of solvent and diazonium salt on Rfilm is shown in Figure 3. In all cases, Rfilm is greater when a diazonium salt is added to solution, versus a neat solution (control). This suggests that the diazonium salts effectively increase Rfilm, though there is variability as to which diazonium salt yields the greatest Rfilm. All data asymptotically approach a maximum Rfilm over the course of 8 h, though 70% of the change in Rfilm is

Figure 4. Influence of concentration of (A) 1 and (B) 3 in TBP on the evolution of Rfilm over time. (C) Voltammograms typical of the irreversible reduction of 1 to 4-aminophenyl. “Control” indicates a steel electrode conditioned similarly, but not exposed to diazonium salts. 1 V/s scan rate. (D) Surface coverage of 1 determined by integration of reduction peak area in (C).

same limiting value of Rfilm is reached in all cases after 8 h, for the concentrations tested. This behavior is consistent with a finite number of surface sites being grafted at different rates, without multilayer film formation. The number of moles of 1 grafted to the steel surface may be determined by reducing of the nitro groups to amino groups and counting the charge passed in the reaction RNO2 + 6H+ + 6e− → RNH 2 + 2H 2O

(1)

This reaction is performed in 0.1 M KCl in 9:1, by volume, deaerated water:ethanol. At the potentials required to perform reaction 1 in an aqueous environment, reduction of any native oxide (e.g., Cr2O3 and Fe2O3) may occur. To avoid including any charge from oxide reduction in the total charge counted, cyclic voltammograms were run at 1 V/s, allowing the fast reduction of NO2 while avoiding the relatively slow reduction of any oxide. As shown in the Figure 4C, the first scan reveals a reduction peak consistent with NO2 reduction, while the second scan matches the first scan of an electrode immersed in

Figure 3. Rfilm as a function of time for steel immersed in 1 mM 1, 2, or 3 in (A) ACN, (B) TBP, and (C) HMDS. (D) Effect of solvent on Rfilm for films of 3 assembled in (A−C). C

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

solvent without diazonium salt. Integration of the area of the reduction peak allows for calculation of the number of moles of 1 grafted to the surface, as plotted in Figure 4D for films of 1 assembled in ACN, TBP, and HMDS. After 8 h, 0.279−0.442 nmol cm−2 had assembled onto the surface, equivalent to 10− 15% coverage on a pristine α-Fe (110) surface, assuming one atom of 1 per iron lattice site. The change in behavior of the steel from “active” to “passive” may be evaluated by the OCP. In Figure 5A, it is seen that the

Figure 5. Effect of films of 1, 2, or 3 assembled in ACN on the response of a steel electrode measured in 0.1 M n-Bu4NBF4 in ACN. (A) OCP after varying assembly times on the same electrode. (B) Suppression of charge transfer to 5 mM ferrocene for films assembled 8 h.

open circuit potential rises with increasing time, indicative of increasing passivity. The increase in OCP in Figure 5A tracks with the corresponding Rfilm in Figure 3A, suggesting that the addition of diazonium salts to the solvent altered the change in OCP with respect to time. The ability of the diazonium-coated steel electrodes to block electron transfer between electrode and the ferrocene/ferrocenium couple is shown in Figure 5B. X-ray Photoelectron Spectroscopy. XPS data were acquired for steel immersed in 1 mM 1, 2, and 3 in TBP and plotted in Figure 6. To evaluate the effect of BF4− on surface chemistry, control samples were immersed in neat TBP (control) or 5 mM NaBF4 in TBP. The presence of 1 on the steel surface was verified in the N 1s spectra plotted in Figure 6A, where peaks at 400.2 and 406.5 eV characteristic of elemental nitrogen and NO2 were observed.25 Similarly, the presence of 3 was confirmed via the C−F fluorocarbon bonding peak at 689.0 eV in the F 1s spectra, shown in Figure 6B.9,12 Additionally, the peak at 685.0 eV present in all spectra except that of neat TBP confirms the presence of BF4−. Characteristic spectra from Fe 2p3/2, Cr 2p3/2, and O 1s are plotted in Figure 6C−E, while the peak locations and widths are tabulated in Figure 6F. Both Fe and Cr spectra reveal peaks characteristic of metal (Fe0, Cr0) and oxide (Fe2O3 and Cr2O3).26−32 A satellite peak, labeled “Fe2O3 sat.”, characteristic of metal−ligand charge transfer is seen in the Fe 2p3/2 spectra. CrOOH, which forms from Cr2O3 after exposure to water vapor, is seen in the Cr 23/2 spectra.27,33 In the O 1s spectra peaks characteristic of oxide, OH−, and H2O are observed.27,28,32 The peak OH− includes any protonated oxide, not only CrOOH. The bonding states of each element are quantified in Table 1. Here it is seen that the presence of diazonium salts increase the percentage of iron bound as Fe0 from 14% to 17−18%, while the amount of chromium bound as Cr0 remains unchanged. From this it is concluded that the diazonium molecules preferentially bind to Fe, preventing subsequent oxidation. The most hydrophobic diazonium salt, 3, shows the lowest amount

Figure 6. Characteristic XPS spectra for steel immersed in 1 mM 1, 2, or 3 in TBP for 8 h. (A) N 1s, (B) F 1s, (C) Fe 2p3/2, (D) Cr 2p3/2, and (E) O 1s. (F) Summary of peak parameters. “NaBF4” indicates 5 mM NaBF4 in TBP.

of H2O adsorbed and the largest amount of oxide (O2−), consistent with the elevated hydrophobicity of a fluorinated surface. ToF-SIMS. Time-of-flight secondary ion mass spectroscopy (ToF-SIMS) was performed on the entire cross section of three steel samples, and intensity maps of the results are presented in Figure 7. Each column indicates an individual sample, while the rows plot intensities from fluorocarbon (A−C), hydrocarbon (D−F), and steel substrate (G−I) components. The first sample (A,D,G) was immersed in 1 mM 1 in TBP for 8 h. As expected, no fluorocarbon signal is observed, though uniform hydrocarbon and substrate signals are seen. The second sample (B, E, H) was immersed in 1 mM 3 in TBP. Here relatively uniform signals are seen for fluorocarbons, hydrocarbons (the benzene ring in 3), and the substrate. The final sample (C, F, I) was immersed in 1 mM 3, rinsed in ethanol, scratched with a diamond scribe, and then immediately placed in 1 mM 1 in TBP. The fluorocarbon signal appears everywhere, except at the scratch locations, confirming removal of 3. The hydrocarbon signal has partially filled in the scratch locations, while a uniform substrate signal is observed. Thus, it is concluded that the diazonium films will partially heal when scratched or mechanically removed. Diazonium Stability. To evaluate the possibility of diazonium decomposition affecting the observed assembly process, solutions of 1, 2, or 3 in each solvent were stored in glassware in the dark, without stainless steel, and sampled periodically over 3 weeks. Neopentyl alcohol was chosen as an internal standard because of its inertness to diazonium compounds, its solubility in all three solvents, and because its D

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Table 1. Relative Peak Areas from Fe 2p3/2, Cr 2p3/2, and O 1s XPS Data, As Shown in Figure 6 for Steel Immersed for 8 h in 1 mM 1, 2, and 3 in TBP, 5 mM NaBF4 in TBP, or Neat TBP (Control) Fe0/(Fe0 + Fe2O3) 1 2 3 5 mM NaBF4 control

0.18 0.17 0.17 0.14 0.14

Cr0/(Cr0 + Cr2O3 + CrOOH) Cr2O3/(Cr0 + Cr2O3 + CrOOH) 0.17 0.18 0.17 0.17 0.18

0.57 0.59 0.57 0.58 0.56

O2−/(O2− + OH− + H2O)

H2O/(O2− + OH− + H2O)

0.19 0.19 0.22 0.19 0.18

0.37 0.29 0.27 0.27 0.30

Figure 7. SIMS intensity maps of 1/8 in. diameter stainless steel electrode surface for films assembled at 1 mM concentration in TBP. (A−C) Intensity of fluorocarbons. (D−F) Intensity of hydrocarbons. (G−I) Intensity of steel substrate. (A, D, G) Film of 1. (B, E, H) Film of 3. (C, F, I) Film of 3 scratched and then filled in with 1. 1

H NMR peaks do not overlap with those of the solvents. As an example, Figure 8A shows 1H NMR spectra for 3 in ACN aged at room temperature for 0 and 14 days. It can clearly be seen that the two doublets at 8.92 and 8.38 ppm due to the aryl hydrogen atoms in 3 are significantly smaller after 14 days, relative to the singlet at 3.05 ppm which is due to the methylene group in neopentyl alcohol. In addition, two new doublets at 7.82 and 7.59 ppm are apparent in the spectrum for the aged sample, due to the formation of aryl degradation products. The large peak at 3.3 ppm is due to water and can vary greatly depending on the solvent source. Figure 8B shows the degradation of 2 in all three solvents. The rate of degradation is reasonably similar for each solvent although the sample in HMDS showed no evidence of 2 remaining after 14 days, while the ACN and TBP solutions still showed 9 and 14%, respectively, of 2 remaining after 21 days. The degradation products were not identified although new peaks in the aryl region of the 1H NMR spectra were evident. In the ACN sample, two new doublets appeared in the aryl region of the spectrum (upfield from those of 2) while the samples in TBP and HMDS both showed many new overlapping peaks of varying multiplicity. Similar peaks appeared in the spectra of aged samples of 1 and 3. Figures 8C and 8D show the stability results of all three diazonium compounds in ACN and TBP (1 and 3 were not evaluated in HMDS due to difficulties with solubility). In both cases, the order of stability was clearly 1 > 3 > 2. The most

Figure 8. Stability tests of diazonium salts in different solvents in the dark at room temperature. (A) Example 1H NMR spectra showing decrease in signal of 3 at 8.92 and 8.38 ppm and constant signal of internal standard at 3.05 ppm. (B) Effect of solvent on the stability of 2. (C, D) Stability of 1, 2, and 3 in (C) ACN and (D) TBP.

stable combination was 1 in ACN, which showed only a 10% loss of diazonium compound over the 3 week test. When the stability test was performed using 1, 2, and 3 in TBP aged at 50 or 75 °C, all of the samples showed 100% degradation after 1 day.

■

DISCUSSION The results above demonstrate that the diazonium molecules are not simply physically adsorbed, but rather they have covalently bonded to the stainless steel surface and formed 10− 15% of a monolayer. These submonolayers withstand sonication in a variety of solvents, as shown by the surface coverages calculated in Figures 4C,D, consistent with covalent bonding. Further, the diazonium films prevent a consistent, measurable amount of iron oxidation. Physically adsorbed species (e.g., BF4−) do not prevent iron oxidation, as shown by XPS in Table 1. E

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

If diazonium cations were physically adsorbed to the steel surface, N 1s XPS signals characteristic of the two nitrogens triple bonded would be seen across all surfaces. Instead, the only signal in the N 1s spectra comes from 1, the nitrophenyl surface, implying that the diazonium molecules have decomposed and covalently bonded to the surface. During the decomposition process, N2 gas is generated and an aryl radical formed. Aryl radicals are not stable and will readily react with the surface or species in solution, making the possibility of observing an adsorbed, not covalently bonded aryl radical unlikely.1−3 Once assembled onto the surface, these submonolayer films are not mobile, consistent with a covalently bound layer, as opposed to a physically adsorbed layer. In the SIMS experiment of Figure 7C, F, I, the steel surface was immersed in 3 for 8 h, scratched, and then immersed in 1 for 8 h. The absence of 3 at the scratches is clearly visible in the fluorocarbon signal (Figure 7C). If the diazonium film were mobile, such contrast between the scratched and unscratched areas would not be present. Despite the formation of a submonolayer diazonium film on the steel surface, it is seen that the addition of diazonium molecules to the solvent has a mild influence on the observed behavior as compared to the neat solvent. Instability of the diazonium salts can be ruled out as a contributing factor, as more than 75% of all salts remain in solution after 24 h at room temperature (Figure 8), and concentrations from 0.1 to 10 mM were not found to have an appreciable effect on film assembly (Figure 4). The formation of 10−15% of a monolayer of diazonium film only prevents 3−4% of iron from oxidizing. Thus, it is assumed that the diazonium salts have bonded to this 3−4%. The remaining diazonium molecules are likely bonded to the 1 wt % (≈15 at. %) carbon in the steel. Unfortunately, no differences in XPS peak locations are expected for C−C bonding versus Fe−C bonding in C 1s spectra or Fe−Fe versus Fe−C in the Fe 2p spectra. The majority of changes in the observed properties of the stainless steel are likely the result of native oxide formation. Even under deaerated conditions, iron and chromium oxides are stable to better than ≈10−60 Torr of O2 at room temperature.34 Once a native oxide forms, the diazonium salts will like not assemble. From the data collected, it is unclear whether the kinetics of native oxide formation outpaces diazonium film formation or whether diazonium film formation is inherently thermodynamically unstable on specific steel phases, likely due to increased chromium content. In order to shed light on this problem, varying amounts of hydrochloric acid (37%, Sigma-Aldrich) were added to the diazonium deposition solutions and steel electrodes were exposed to these solutions, recording the OCP over the course of 4 h. In Figure 9A, the OCP after 4 h is shown for steel electrodes immersed in 0.1 M n-Bu4NBF4 in ACN with or without (control) 1 mM 1. For the control solutions, a lower potential is observed at higher acid concentrations, indicative of a more active surface where native oxide formation is unfavorable. The lack of passivity at 1 M HCl is further confirmed by the rusty color of the electrode surface and the yellow-green hue acquired by the solution. At lower concentrations of acid a higher OCP is recorded, suggesting a more passive surface and previously seen oxide formation. Even under conditions where oxide formation is unfavorable, adding diazonium molecules to the solution was insufficient to passivate the stainless steel, as seen in the nearly identical OCPs of Figure 9A and the surface coverages in Figure 9B which do

Figure 9. (A) Effect of HCl concentration in 0.1 M n-Bu4NBF4 in ACN on the OCP of steel electrodes immersed for 4 h with or without (control) 1 mM 1. (B) Surface coverage resulting when diazonium was present in (A). Solutions to which no HCl was added are plotted at 10−6 M HCl.

not exceed that recorded without HCl. At high acid concentrations, the assembly of diazonium films is likely prevented by rapid etching of the electrode surface. At intermediate concentrations of acid (1−10 mM), where passivation via oxide formation is incomplete after 4 h, the diazonium molecules still could not assemble a full monolayer on the surface. From these results, it is concluded that diazonium molecules will not spontaneously form a complete monolayer with which to passivate 440C stainless steel in these solvents. These results reveal a limitation of spontaneous diazonium assembly onto steels in nonaqueous electrolytes not previously discussed in the literature. Rather, previous reports have only shown spontaneous assembly of diazoniums onto pure iron in ACN and bias-directed assembly onto pure iron or nickelcontaining stainless steels (304L or 316) in ACN.11−13,16,17 The addition of chromium to iron, forming stainless steel, undoubtedly alters the spontaneously assembly process in acetonitrile. Bias-directed assembly, on the other hand, may provide the additional energy required to render diazonium film formation favorable.

■

CONCLUSION Three aryldiazonium salts were shown to spontaneously assemble onto the surface of type 440C stainless steel immersed in ACN, TBP, or HMDS, though total surface coverage was limited to ≈0.15 monolayers. The mild influence of diazonium salt type or concentration on the observed assembly process is attributed to this low surface coverage. Nevertheless, XPS confirmed preferential bonding of organic molecules to iron over chromium, and SIMS demonstrated that diazonium molecules will partially self-heal when mechanically removed. Aging the diazonium salts showed their stability at room temperature, with up to 90% of the original diazonium salt concentration remaining after 21 days at room temperature. The addition of HCl to the electrolyte succeeded in rendering native oxide formation unfavorable, yet diazonium molecules were still unable to assemble a full monolayer on the surface with the addition of 1 M−10 μM HCl. It is concluded that the investigated diazonium molecules will not spontaneously form a continuous monolayer on 440C stainless steel immersed in ACN, TBP, or HMDS.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (L.J.S.). F

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Notes

Modified with Nitrophenyl Groups. Electrochem. Commun. 2007, 9, 2412−2417. (16) Berge, F.; Delhalle, J.; Mekhalif, Z. Hybrid coating on steel: ZnNi Electrodeposition and Surface Modification with Organothiols and Diazonium Salts. Electrochim. Acta 2008, 53, 2852−2861. (17) Le, X. T.; Zeb, G.; Jegou, P.; Berthelot, T. Electrografting of Stainless Steel by the Diazonium Salt of 4-Aminobenzylphosphonic Acid. Electrochim. Acta 2012, 71, 66−72. (18) Hinge, M.; Goncalves, E. S.; Pedersen, S. U.; Dassbjerg, K. On the Electrografting of Stainless Steel from Para-Substituted Aryldiazonium Salts and the Thermal Stability of the Grafted Layer. Surf. Coat. Technol. 2010, 205, 820−827. (19) Salmi, Z.; Gam-Derouich, S.; Mahouche-Chergui, S.; Turmine, M.; Chehimi, M. M. On the Interfacial Chemistry of Aryl Diazonium Compounds in Polymer Science. Chem. Pap. 2012, 66, 369−391. (20) Ahmad, R.; Griffete, N.; Lamouri, A.; Mangeney, C. Functionalization of Magnetic Nanocrystals by Oligo(ethylene oxide) Chains Carrying Diazonium and Iniferter End Groups. J. Colloid Interface Sci. 2013, 407, 210−214. (21) Griffete, N.; Ahmad, R.; Benmehdi, H.; Lamouri, A.; Decorse, P.; Mangeney, C. Elaboration of Hybrid Silica Particles Using a Diazonium Salt Chemistry Approach. Colloids Surf., A 2013, 439, 145− 150. (22) Oliveira, C.; dos Santos, E.; Nunes, J. B.; Correia, C. J. Org. Chem. 2012, 77, 8120−8190. (23) Aliaga, C.; Santos, C. S.; Baldelli, S. Surface Chemistry of RoomTemperature Ionic Liquids. Phys. Chem. Chem. Phys. 2007, 9, 3683− 3700. (24) Small, L.; Wheeler, D. Influence of Analysis Method on the Experimentally Observed Capacitance at the Gold-Ionic Liquid Interface. J. Electrochem. Soc. 2014, 161, H260−H263. (25) Hunger, R.; Jaegermann, W.; Merson, A.; Shapira, Y.; Pettenkofer, C.; Rappich, J. Electronic Structure of Methoxy-, Bromo-, and Nitrobenzene Grafted onto Si(111). J. Phys. Chem. B 2006, 110, 15432−15441. (26) Brooks, A.; Clayton, C.; Doss, K.; Lu, Y. On the Role of Cr in the Passivity of Stainless Steel. J. Electrochem. Soc. 1986, 133, 2459− 2464. (27) Hermas, A.; Ogura, M. N. K. Enrichment of ChromiumContent in Passive Layers on Stainless Steel Coated with Polyaniline. Electrochim. Acta 2005, 50, 2001−2007. (28) Hanawa, T.; Hiromoto, S.; Yamamoto, A.; Kuroda, D.; Asami, K. XPS Characterization of the Surface Oxide Film of 316L Stainless Steel Samples that were Located in Quasi-Biological Environments. Mater. Trans. 2002, 43, 3088−3092. (29) Cavallini, M. XPS Study of Mechanically Polished Stainless Steel Surfaces. Mater. Chem. 1980, 5, 29−36. (30) Maurice, V.; Cadot, S.; Marcus, P. XPS, LEED and STM Study of Thin Oxide Films Formed on Cr(110). Surf. Sci. 2000, 458, 195− 215. (31) Belo, M. D. C.; Rondot, B.; Compere, C.; Montemore, M.; Simoes, A.; Ferreira, M. Chemical Composition and Semiconducting Behaviour of Stainless Steel Passive Films in Contact with Artificial Seawater. Corros. Sci. 1998, 40, 481−494. (32) Lorang, G.; Belo, M. D. C.; Simoes, A.; Ferreira, M. Chemical Composition of Passive Films on AISI 304 Stainless Steel. J. Electrochem. Soc. 1994, 141, 3347−3356. (33) Maurice, V.; Cadot, S.; Marcus, P. Hydroxylation of Ultra-thin Films of α-Cr2O3 (0001) Formed on Cr(110). Surf. Sci. 2001, 471, 43−58. (34) Davis, J., Ed.; ASM Metals Handbook, Desk ed., 2nd ed.; CRC Press: Boca Raton, FL, 1998.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank Dr. Michael Brumbach for help acquiring XPS data and Tony Olhassen for TOF-SIMs imaging. This work was supported by Boeing Corp. Sandia National Laboratories is a multiprogram laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under Contract DE-AC04-94AL85000.

■

REFERENCES

(1) Mahouche-Chergui, S.; Gam-Derouich, S.; Mangeney, C.; Chehimi, M. M. Aryl Diazonium Salts: A new Class of Coupling Agents for Bonding Polymers, Biomacromolecules and Nanoparticles to Surfaces. Chem. Soc. Rev. 2011, 40, 4143−4166. (2) Pinson, J.; Podvorica, F. Attachment of Organic Layers to Conductive or Semiconductive Surfaces by Reduction of Diazonium Salts. Chem. Soc. Rev. 2005, 34, 429−439. (3) Belanger, D.; Pinson, J. Electrografting: A Powerful Method for Surface Modification. Chem. Soc. Rev. 2011, 40, 3995−4048. (4) Mesnage, A.; Lefevre, X.; Jegou, P.; Deniau, G.; Palacin, S. Spontaneous Grafting of Diazonium Salts: Chemical Mechanism on Metallic Surfaces. Langmuir 2012, 28, 11767−11778. (5) Downard, A. J. Potential Dependence of Self-Limited Films Formed by Reduction of Aryldiazonium Salts at Glassy Carbon Electrodes. Langmuir 2000, 16, 9680−9682. (6) D’Amours, M.; Belanger, D. Stability of Substituted Phenyl Groups Electrochemically Grafted at Carbon Electrode Surface. J. Phys. Chem. B 2003, 107, 4811−4817. (7) Saby, C.; Ortiz, B.; Champagne, G. Y.; Belanger, D. Electrochemical Modification of Glassy Carbon Electrode Using Aromatic Diazonium Slats. 1. Blocking Effect of 4-Nitrophenyl and 4Carboxyphenyl Groups. Langmuir 1997, 13, 6805−6813. (8) Stewart, M. P.; Maya, F.; Kosynkin, D. V.; Dirk, S. M.; Stapleton, J. J.; McGuiness, C. L.; Allara, D. L.; Tour, J. M. Direct Covalent Grafting of Conjugated Molecules onto Si, GaAs, and Pd Surfaces from Aryldiazonium Salts. J. Am. Chem. Soc. 2004, 126, 370−378. (9) Pandey, D.; Zemlyanov, D. Y.; Bevan, K.; Reifenberger, R. G.; Dirk, S. M.; Howell, S. W.; Wheeler, D. R. UHV STM I(V) and XPS Studies of Aryl Diazonium Molecules Assembled on Si(111). Langmuir 2007, 23, 4700−4708. (10) Bernard, M.-C.; Chausse, A.; Cabet-Deliry, E.; Chehimi, M. M.; Pinson, J.; Podvorica, F.; Vautrin-Ul, C. Organic Layers Bonded to Industrial, Coinage, and Noble Metals through Electrochemical Reduction of Aryldiazonium Salts. Chem. Mater. 2003, 15, 3450−3462. (11) Combellas, C.; Delamar, M.; Kanoufi, F.; Pinson, J.; Podvorica, F. I. Spontaneous Grafting of Iron Surfaces by Reduction of Aryldiazonium Salts in Acidic or Neutral Aqueous Solution. Application to the Protection of Iron against Corrosion. Chem. Mater. 2005, 17, 3968−3975. (12) Chausse, A.; Chehimi, M. M.; Karsi, N.; Pinson, J.; Podvorica, F.; Vautrin-Ul, C. The Electrochemical Reduction of Diazonium Salts on Iron Electrodes. The Formation of Covalently Bonded Organic ayers and Their Effect on Corrosion. Chem. Mater. 2002, 14, 392−400. (13) Adenier, A.; Barre, N.; Cabet-Deliry, E.; Chausse, A.; Griveau, S.; Mercier, F.; Pinson, J.; Vautrin-Ul, C. Study of the Spontaneous Formation of Organic Layers on Carbon and Metal Surfaces from Diazonium Salts. Surf. Sci. 2006, 600, 4801−4812. (14) Polsky, R.; Harper, J. C.; Wheeler, D. R.; Arango, D. R.; Brozik, S. M. Electrically Addressable Cell Immobilization Using Phenylboronic Acid Diazonium Salts. Angew. Chem., Int. Ed. 2008, 47, 1−5. (15) Kullapere, M.; Matisen, L.; Saar, A.; Sammelselg, V.; Tammeveski, K. Electrochemical Behavior of Nickel Electrodes G

dx.doi.org/10.1021/la503630f | Langmuir XXXX, XXX, XXX−XXX