Surface Dissociation Effect on Phosphonic Acid Self-Assembled Monolayer Formation on ZnO Nanowires.

Understanding the formation process of self-assembled monolayers (SAMs) of organophosphonic acids on ZnO surfaces is essential to designing their various applications, including solar cells, heterogeneous catalysts, and molecular sensors. Here, we report the significant effect of surface dissociation on SAM formation of organophosphonic acids on single-crystalline ZnO nanowire surfaces using infrared spectroscopy. When employing the most conventional solvent-methanol (relative permittivity εr = 32.6), the production of undesired byproducts (layered zinc compounds) on the surface was identified by infrared spectral data and microscopy. On the other hand, a well-defined SAM structure with a tridentate coordination of phosphonic acids on the surface was confirmed when employing toluene (εr = 2.379) or tert-butyl alcohol (εr = 11.22-11.50). The observation of layered zinc compounds as byproducts highlights that the degree of Zn2+ dissociation from the ZnO solid surface into a solvent significantly affects the surface coordination of phosphonic acids during the SAM formation process. Although the ZnO nanowire surface (m-plane) is hydrophilic, the present results suggest that a weaker solvent polarity is preferred to form well-defined phosphonic acid SAMs on ZnO nanowire surfaces without detrimental surface byproducts.

Introduction

Molecular surface modification on metal oxide nanostructures has shown great promise to tailor their surface functionalities for various applications, including heterogeneous catalysts[1,2] and molecular sensors.[3,4] One of the promising molecular surface modifications is applying self-assembled monolayers (SAMs) on metal oxide surfaces.[5−7] Various SAMs have been successfully introduced onto surfaces of various metal oxide nanostructures, including thin films, nanowires, nanoparticles, and others.[5,8−10] For example, SAMs were applied onto ITO substrates to modulate the work function.[11,12] Among various head groups of SAMs, a phosphonic acid has particularly attracted the significant attention of many researchers[13−18] because phosphonic acids provide a more robust modification layer when compared to those formed from carboxylic acids, especially after annealing to maximize the number of P–O bonds.[19−22] For example, interesting applications using phosphonic acid SAMs on ZnO surfaces have been successfully demonstrated[8−10,13−15,21−23] since ZnO is one of the most popular metal oxide materials.[24−26] Zhang et al. have reported the formation of phosphonic acid SAMs on ZnO surfaces.[14] Lim et al. demonstrated the long-term stability and high sensing selectivity with phosphonic acid SAM modified ZnO nanowire sensors.[27] Despite these successes of phosphonic acid SAMs on ZnO nanostructures, the nature of the formation process is still complex and affected by many experimental parameters, including temperature,[28,29] concentration and modification time,[30] solid surface and material,[18,31,32] and solvent.[20,21] Solvent selection for the SAM formation process is especially important because the solvent must dissolve the SAM molecules.[33] Literature survey for phosphonic acid SAMs on ZnO surfaces reveals that the major solvents for these SAM formation processes are alcohols, including methanol, ethanol, and others.[29,34−37] Although Chen et al. have reported the solvent effect on the phosphonic acid SAM formation process on ITO substrates,[38] such solvent effects on phosphonic acid SAM formation processes on ZnO nanostructures have not been studied. Here, we report the significant impact of surface dissociation on SAM formation of organophosphonic acids on single-crystalline ZnO nanowire surfaces using infrared spectroscopy. We found that the degree of Zn2+ dissociation from the ZnO solid surface into a solvent strongly affects the surface coordination of phosphonic acids during the SAM formation process.

Results and Discussion

Figure a,b shows the time-series data of FT-IR during SAM formation processes of octadecylphosphonic acid (ODPA) on ZnO nanowires in methanol solvent. The concentration of ODPA is 0.1 mM. The details of ZnO nanowire growth processes can be seen in Methods. The analyzed data (the alkyl peak area and the wavenumber of CH2 symmetric stretch vibration-νs(CH2)) are shown in Figure c,d. Clearly, the IR-absorbance of SAMs in Figure a,b tends to be stronger with increasing surface modification time, as also seen in Figure c. As seen in Figure a,d, the wavenumber of νs(CH2) tends to decrease from 2853.7 down to 2850.8 cm–1. This trend is well known as the alkyl chain conformation change from gauche to all-trans geometry with increasing SAM surface density.[39,40] Although the interpretation of absorption bands of alkyl chains (2800–3000 cm–1) is straightforward, the absorption bands for phosphonic acids (900–1250 cm–1) exhibit rather complicated spectra, as seen in Figure b. Previous studies on IR spectra of phosphonic acid SAMs have reported different identifications on their data.[5,10,13,16,41] For example, the peaks around 900–1050 cm–1 were assigned to the P–OH group by comparing the obtained IR peaks of SAMs with the IR peaks of phosphonic acid powder.[5,14,41,42] The peaks around 1040 cm–1 were determined to be the stretching modes of PO32–,[41] which involve P–O and P=O terminations.[41] The peak around 1220 cm–1 was reported to be P=O stretching.[29] Although their identifications are rather different in detail, these identifications based on IR peaks of phosphonic acid powder assume the existence of phosphonic acid SAMs on ZnO surfaces without considering surface side reactions. Figure e shows the scanning electron microscope (SEM) images of ZnO nanowire surfaces when varying the SAM surface modification time. As clearly seen in the SEM images, some foggy nanostructures on the nanowire surfaces appear with increasing SAM modification time. The foggy nanostructure becomes clear in ODPA modified ZnO nanowires with a SAM modification time of 24 h (Figure S1). Since the emergence of such nanostructures during the SAM formation process is distinct, it is important to identify the structure of the foggy nanostructures.

Figure 1

FT-IR spectra of ODPA on ZnO nanowires in the (a) alkyl or (b) phosphonic acid region for an ODPA concentration of 0.1 mM in methanol solvent. ODPA modification time dependence of (c) alkyl peak area and (d) wavenumber of CH2 symmetric stretch vibration (νs(CH2)). Inset figures show enlarged characteristics at an early stage of SAM formation. (e) FESEM images of ODPA modified ZnO nanowires using methanol solvent. The ODPA concentration was 0.1 mM for all samples.

First, we performed X-ray diffraction (XRD) measurements to identify the structure of foggy nanostructures on ZnO nanowire surfaces, as seen in Figure S2. Unfortunately, conventional XRD measurements could not detect any significant signals for the foggy surface nanostructures, presumably due to the randomness of crystal orientation and the small amount. Here, we consider possible side surface reactions during phosphonic acid SAM formation on ZnO surfaces. One of the plausible surface reactions is the formation of layered zinc phosphonate (Zn-ODP) structures, as illustrated in Figure a.[43]Figure b shows the comparison between the present IR spectrum and the previously reported IR spectrum for the layered Zn-ODP structures.[43] Clearly, there is a good correlation between the two spectra. Especially, the six IR peaks (around 940, 980, 1060,1080, 1120, and 1170 cm–1) are well consistent with such IR peaks of the layered Zn-ODP structures. Note that some peaks in the present samples are not consistent with those of the layered Zn-ODP structures. Thus, based on these results, we conclude that the observed surface nanostructures during phosphonic acid SAM formation on ZnO surfaces are layered Zn-ODP. For well-defined phosphonic acid SAMs on ZnO surfaces, this layered Zn-ODP formation is detrimental. It is noted that the present series of experiments were performed using the most conventional solvent-methanol.[34,35] Next, we solve this undesired side surface reaction issue by considering the reaction mechanism.

Figure 2

(a) Schematic diagram of the layered Zn-ODP structure. (b) FT-IR spectrum of ODPA on ZnO nanowires in the phosphonic acid region for an ODPA concentration of 0.1 mM for 60 min in methanol solvent. The reported FT-IR spectrum of layered Zn-ODP[43] is also shown.

To form layered Zn-ODP on ZnO, a dissociation of Zn2+ from ZnO nanowire surfaces must be significant; otherwise, simple phosphonic acid SAM formations should occur. Based on this speculation, we change the solvent from relatively polar methanol (relative permittivity εr = 32.6[44]) to non-polar toluene (εr = 2.379[44]) to suppress the degree of dissociation of Zn2+ from the ZnO nanowire surface. Figure a–e shows the results of phosphonic acid SAM formation on ZnO nanowire surfaces when employing toluene as the solvent. There is a significant difference between the two solvents (methanol and toluene) on the FT-IR spectra and SEM images. First, when comparing between Figures b and 1b on absorption bands for phosphonic acids (900–1250 cm–1), the spectra of toluene-solvent samples exhibit mainly three peaks (970, 1070, and 1140 cm–1), which are also found in Figure b with other peaks from the layered Zn-ODP. Thus, these results highlight that the surface molecular conformation of phosphonic acids is strongly affected by the solvent. As seen in Figure c, the SAM formation speed in toluene solvent is slower than that in methanol solvent in Figure c. Interestingly, the data of Figure d implies that the degree of all-trans geometry is higher for toluene solvent than that for methanol solvent because the wavenumber of νs(CH2) is lower for toluene solvent. More importantly, SEM observations in Figure e do not show any surface nanostructures during the phosphonic acid SAM formation process, which is rather different from the trend in Figure e. Therefore, altering the solvent from polar methanol to non-polar toluene significantly suppresses the emergence of surface side reactions and formation of layered Zn-ODP on the ZnO nanowire surface. The proposed strategy of using a non-polar solvent was confirmed by performing experiments using tert-butyl alcohol solvent (εr = 11.22–11.50[45,46]). Figure S3 shows the FT-IR spectra of ODPA on ZnO nanowires using tert-butyl alcohol solvent, which agrees well with that using toluene solvent (Figure a–d). Thus, the suppression of side surface reactions was achieved in tert-butyl alcohol solvent with weaker polarity than methanol. Finally, we attempt to identify the molecular conformation of phosphonic acids when using toluene by comparing with DFT calculations, as illustrated in Figure f. The calculated wavenumbers for a bidentate coordination with P=O and a tridentate coordination without P=O of phosphonic acids on ZnO surfaces are shown in Figure b. The comparison between experimental spectra and DFT simulations reveals the existence of a tridentate coordination of phosphonic acids on ZnO surfaces when employing toluene solvent. Thus, these results highlight that phosphonic acid SAM formation on ZnO surfaces requires non-polar solvent-toluene rather than conventional polar solvent-methanol to perform well-defined SAM formation without detrimental surface side reactions with byproducts, although the ZnO nanowire surface is hydrophilic.

Figure 3

FT-IR spectra of ODPA on ZnO nanowires in the (a) alkyl or (b) phosphonic acid region for an ODPA concentration of 0.1 mM in toluene solvent. ODPA modification time dependence of (c) alkyl peak area and (d) wavenumber of CH2 symmetric stretch vibration (νs(CH2)). DFT-calculated vibrations of P=O (ν(P=O)) and P–O (ν1(P–O), ν2(P–O)) are also indicated. (e) FESEM images of ODPA modified ZnO nanowires using toluene solvent. The ODPA concentration was 0.1 mM for all samples. (f) Optimized structure of bidentate coordination of phosphonic acid on Zn and tridentate coordination on the hexagonal ZnO (10–10) plane calculated by DFT simulations.

Conclusions

We demonstrate the significant effect of surface dissociation on SAM formation of organophosphonic acids on single-crystalline ZnO nanowire surfaces using infrared spectroscopy. When employing the most conventional solvent-methanol (relative permittivity εr = 32.6), the presence of undesired byproducts (layered zinc compounds) on the surface was identified by infrared spectral data and microscopy. On the other hand, a well-defined SAM structure with a tridentate coordination of phosphonic acids on the surface was confirmed when employing toluene (εr = 2.379) or tert-butyl alcohol (εr = 11.22–11.50). The observation of layered zinc compounds as byproducts highlights that the degree of Zn2+ dissociation from the ZnO solid surface into a solvent significantly affects the surface coordination of phosphonic acids during the SAM formation process. Although the ZnO nanowire surface (m-plane) is hydrophilic, the present results suggest that a weaker solvent polarity is preferred to form well-defined phosphonic acid SAMs on ZnO nanowire surfaces without detrimental surface byproducts.

Methods

ZnO Nanowire Growth

Single-crystalline ZnO nanowires were hydrothermally grown on a ZnO seed layer/SiO2/p-Si substrate. A 5 nm Ti adhesion layer and 100 nm ZnO seed layer were sequentially deposited onto a 100 nm SiO2/p-type Si substrate by radio frequency (RF) sputtering. Solutions for hydrothermal reactions were mixtures composed of 5 mM zinc nitrate hexahydrate, Zn(NO3)2·6H2O (Wako, 99.0%) and 5 mM hexamethylenetetramine (HMTA), (CH2)6N4 (Wako, 99.0%). The ZnO-deposited substrate was immersed in the growth solution and kept at 80 °C for 24 h. A ZnO nanowire array was obtained on the substrate after the reaction. After growth, the samples were rinsed with DI water and IPA. Then the ZnO nanowires were annealed for 1 h at 600 °C in atmospheric air to prevent surface degradation.[47]

Modification of Octadecylphosphonic Acid SAMs on ZnO Nanowires

Modification solutions (0.1 mM) were prepared by dissolving octadecylphosphonic acid (ODPA) in methanol or toluene. The annealed ZnO nanowire array was dipped in the solution (10 mL) at room temperature. Then, the samples were washed with methanol or toluene and tetrahydrofuran. After air flow drying, the ODPA modified ZnO nanowire arrays were annealed for 30 min at 150 °C in atmospheric air.

Characterizations

Scanning electron microscope (SEM) images were acquired using a JEOL JSM7610F instrument. The SEM images (Figures e and 3e) confirm that the grown ZnO nanowires exhibit hexagonal columnar structures (diameter of ∼100 nm), which indicates that the ZnO nanowires have a single wurtzite structure with the prism (10–10) plane as the main face. Structural characterizations of ZnO nanowires were determined by XRD (PHILIPS, X’Pert MRD 45 kV, 40 mA). The FT-IR spectra of the surface molecules on the ZnO nanowires were recorded at room temperature on a Thermo Fisher Scientific Nicolet iS50 FT-IR spectrometer equipped with a mercury-cadmium-telluride (MCT) detector. 300 scans were accumulated to obtain each spectrum. The test room was purged with dry air. The FT-IR spectrum for bare ZnO was used as the background spectrum for the other measurements. For FT-IR experiments, a double-polished float-zone Si substrate was used for the ZnO nanowire array samples. To analyze the change in the relative amount of alkyl chains, the peak area was calculated by integrating the region from 2800 to 3000 cm–1 of the IR absorption spectrum.

Computational Details

We computed the vibrational frequencies of phosphonic acid on ZnO (10–10) surfaces using density functional theory (DFT) to assign the P–O stretching bands. Simplified cluster models with partially fixed coordinations frozen atoms (see Table S1) were employed to consider the adsorbed ZnO (10–10) surface approximately. These models were extracted from a wurtzite ZnO crystal structure with a = b = 3.25 Å and c = 5.2 Å. The DFT calculations were carried out using the Gaussian 16 program suite Revision A03 with the B3LYP hybrid functional.[48] The obtained harmonic vibrational frequencies were shifted using a scale factor of 0.964[49] to incorporate anharmonic effects effectively.