Electroless Deposition and Surface-Enhanced Raman Scattering Application of Palladium Thin Films on Glass Substrates

Abstract

In this work, we describe a very simple electroless deposition method to prepare moderate-SERS-active nanostructured Pd films deposited on the glass substrates. To the best of our knowledge, this is the first report on the one-pot electroless method to deposit Pd nanostructures on the glass substrates. This method only requires the incubation of negatively charged glass substrates in ethanol-water mixture solutions of $Pd(NO_3)_2$ and butylamine at elevated temperatures. Pd films are then formed exclusively and evenly on glass substrates. Due to the aggregated structures of Pd, the SERS spectra of benzenethiol and organic isonitrile could be clearly identified using the Pd-coated glass as a SERS substrate. This one-step fabrication method of Pd thin film on glass is cost-effective and suitable for the mass production.

Introduction

Surface-enhanced Raman scattering (SERS) is a phenomenon in which the scattering cross sections of molecules adsorbed on certain metal surfaces are dramatically enhanced.1,2 In recent years, it has been reported that even single-molecule detection is possible by SERS.3,4 SERS has thus been used in many areas of science and technology, including chemical analysis, corrosion, lubrication, catalysis, sensor, and molecular electronics, etc.5,6 One of the weak points of SERS is that only noble metals such as Au, Ag, and Cu can usually provide large enhancement effects,7-9 which severely limits wider applications involving other metallic materials of both fundamental and practical importance. In recent years, even transition metals have been proven to be SERS active when they are subjected to a proper roughening process.10-12 It is still difficult, however, to obtain Raman spectra of molecules adsorbed on transition metals like Pt and Pd, especially in nonelectrochemical environments.

Palladium is an important transition metal with high catalytic activity,13,14 although it is not used as widely as Pt at the present time.15,16 However, a great deal of effort has recently been focused on the preparation of Pd nanostructures and their applications in the fields of catalysis, hydrogen storage, and hydrogen sensors due to their large surface area-to-volume ratio, relatively lower price than Pt, and especially their unique function in absorbing hydrogen.17-19 Unfortunately, Pd has intrinsically non- or very weak-SERS-activity even among the transition metals, such as Pt, Rh, Ru, Fe, Co, and Ni, so that it is difficult to follow catalytic reactions by detecting the SERS spectra of the surface-adsorbed reagents and products on Pd. Nonetheless, many attempts have been made to improve the SERS activity of Pd nanoparticles.20,21 Tian and his colleagues have synthesized the core-shell nanoparticles in order to utilize the strategy of “borrowing SERS activity”.22 Pergolese et al. have also deposited Pd clusters onto Ag nanoparticles in order to combine the SERS enhancement of the noble metal with the catalytic efficiency of Pd.23

Recently, we reported that very stable and optically tunable Ag films can be reproducibly fabricated simply by soaking glass substrates into ethanolic solutions of AgNO3 and butylamine.24 These Ag films were shown to possess a very homogeneous morphology and high-SERS-activity. Similarly, a facile one-step method was applied in this work to produce Pd films on dielectric substrates such as glass. To the best of our knowledge, this is the first report on the very simple electroless-plating method to deposit Pd nanostructures on the glass substrates. This method only requires the incubation of negatively charged slide glasses in ethanolwater (8:2 v/v) mixture solutions of Pd(NO3)2 and butylamine at elevated temperatures. The grain size of Pd nanostructures can be readily controlled by changing the palladium ion-to-butylamine molar ratio. Due to the aggregated structures of Pd, the SERS spectra of benzenethiol and organic isonitrile could be clearly identified using the Pd-coated glass. The enhancement factor (EF) estimated using benzenethiol as a prototype adsorbate reached 1.8 × 103 using a 514.5 nm excitation source. In addition, 2,6-dimethylphenylisocyanide appeared to adsorb on the on-top and the 3-fold hollow sites of Pd nanostructures via the isocyanide group

 

Experimental

Palladium(II) nitrate dihydrate (Pd(NO3)2·2H2O, 99%), butylamine (C4H9NH2, 99.5%), benzenethiol (BT, 99%), and 2,6-dimethylphenylisocyanide (2,6-DMPI, 96%) were pur- chased from Aldrich and used as received. Absolute ethanol (99.9%) was purchased from J. T. Baker. Other chemicals, unless specified, were of reagent grade. Highly pure water (Millipore), of resistivity greater than 18.0 MΩ·cm, was used throughout.

Initially, slide glasses (50 mm × 10 mm × 1 mm, Marienfeld) were soaked in a piranha solution for 30 min and sonicated in distilled water for 10 min, followed by rinsing with ethanol, and finally dried in an oven at 60 ℃ for 1 h. The cleaned slide glasses were dipped in the reaction mixtures and incubated for 12 h at 70 ± 1 ℃ with vigorous shaking. A polypropylene container was used as the reaction vessel. The reaction mixture consisted of 10 mL of 10 mM Pd(NO3)2 in ethanol-water (8:2 v/v) and 40 to 400 μL of butylamine. The Pd-coated glass was finally rinsed with ethanol and air-dried. To record the SERS spectra of BT and 2,6-DMPI, the Pd-coated glasses were soaked in ethanolic solutions of 10 mM BT and 2,6-DMPI, respectively, for 3 h followed by thorough washing with ethanol after evaporation of the solvent.

Ultraviolet-visible (UV-vis) spectra were obtained with an Avantes 3648 spectrometer. Field-emission scanning electron microscopy (FE-SEM) images were obtained using a JSM-6700F field-emission scanning electron microscope operated at 2.0 kV. Energy dispersive X-ray (EDX) characterization was performed with a SUPRA 55VP field-emission scanning electron microscope operating at 15 kV. X-ray diffraction (XRD) was conducted on a Rigaku Model MiniFlex powder diffractometer using Cu Kα radiation. X-ray photoelectron spectroscopy (XPS) measurements were carried out with an AXISH model using Mg Kα X-ray as the light source. Raman spectra were obtained using a Renishaw Raman system Model 2000 spectrometer equipped with an integral microscope (Olympus BH2-UMA). The 514.5 nm line from a 20 mW Ar+ laser (Melles-Griot Model 351MA520) was used as the excitation source. The Raman band of a silicon wafer at 520 cm‒1 was used to calibrate the spectrometer, and the accuracy of the spectral measurement was estimated to be better than 1 cm‒1. Atomic force microscopy (AFM) images were obtained on a Digital Instruments Nanoscope Ⅲa system. Using an 125 μm long etched silicon cantilever with a nominal spring constant of 20-100 N/m (Nanoprobe, Digital Instruments), topographic images were recorded in a tapping mode with a driving frequency of ~300 kHz at a scan rate of 2 Hz.

 

Results and Discussion

As mentioned in the Experimental section, Pd films on glass substrates were prepared by the reaction of Pd(NO3)2 (10 mM, 10 mL of in ethanol-water) with 40, 80, 100, or 400 μL of butylamine. The palladium ion-to-butylamine molar ratios were 1:4, 1:8, 1:10, or 1:40; hereafter, the prepared Pd films will be labeled as Pd(1:4), Pd(1:8), Pd(1:10), or Pd(1:40). Figure 1 shows the FE-SEM images of the Pd films formed on glass substrates. The mean grain sizes are determined to be 85 ± 10, 133 ± 22, and 165 ± 23 nm for the Pd(1:4), Pd(1:8), and Pd(1:10) films, respectively. When excess butylamine was used for the preparation of Pd film, as can be seen in Figure 1(d), the glass substrate was covered fully with Pd nanostructures; the grain size of Pd(1:40) film could thus not be determined explicitly. This means that larger and more aggregated Pd grains are formed by increasing the molar ratio of butylamine. As shown in Figure 2(a), the XRD peaks at 40.1°, 46.6°, and 68.0° for the Pd(1:10) film are assigned to the reflections of (111), (200), and (220) crystalline planes of cubic Pd (JCPDS file 87-0638), respectively.25 The crystalline size of Pd nanostructures calculated from the Scherrer equation, using the half width of the intense (111) reflections was about 12 nm.26 The latter value is 14 times smaller than the apparent size determined from the FE-SEM image in Figure 1(c). This discrepancy can be explained by assuming that the grains of Pd films are actually composed of smaller (~12 nm in diameter) particles. Figure 2(b) shows the XPS spectrum of the as-prepared Pd(1:10) film. Apparently there are two strong peaks with the binding energy (BE) values of 335.7 and 340.8 eV. They can be attributed to arise from Pd3d5/2 and Pd3d3/2 of metal Pd0.27 In fact, those strong peaks are asymmetric in shape with shoulders at higher BE values (337.4 and 342.9 eV, corresponding to Pd3d5/2 and Pd3d3/2, respectively). The origin of the shoulder peaks is not clear; one possible explanation would be that the peaks are originated from the Pd2+ species, assignable to PdO, which often accompanies Pd0 left in an oxygen-containing environment.28

Figure 1.FE-SEM images of Pd films prepared by electroless deposition at different molar ratios of Pd(NO3)2 and butylamine: (a) Pd(1:4), (b) Pd(1:8), (c) Pd(1:10), and (d) Pd(1:40); the scale bar = 1 μm.

Figure 2.(a) XRD and (b) XPS analyses of Pd(1:10) film.

The protocol of electroless deposition of Pd onto the surface of a slide glass and its usage to detect analytes by SERS are schematically drawn in Figure 3. A negatively charged glass substrate, qualitatively sketched in (a), is very effective in the deposition of Pd2+ ions particularly because the hydroxyl groups of a glass surface are partially deprotonated in ethanol-water mixture solution.29 Upon adding Pd2+ ions, the oxygen sites are bound by Pd2+ ions such that the surface-bound Pd2+ ions can subsequently function as seeds for the growth of Pd nanostructures on a glass substrate. As reported in our previous work, very stable and optically tunable Ag films could be reproducibly fabricated simply by soaking glass substrates into ethanolic solutions of AgNO3 and butylamine.24 In the preparation of Pd films on glass substrates, however, pure ethanolic solution of Pd(NO3)2 and butylamine led to the formation of colloidal Pd nanoparticles.30 In order to deposit Pd nanostructures on a glass substrate, a controlled amount of water should be added in the ethanol solution. Without proper amount of water, Pd nanostructures did not form on the glass substrates. The reducing power of pure ethanolic solution with butylamine is strong enough to produce Pd nanoparticles in the solution state. By adding some amount of water to the solution, bulk reaction does not take place and Pd nanostructures start to form on the surface of a slide glass. On the other hand, high concentration of water hampers not only the formation of Pd nanoparticles in the solution but also the development of Pd nanostructures on a glass substrate. Thus, to balance the reducing power of the solution, ethanol-water (8:2 v/v) mixture was used throughout this work. Once Pd nanostructures are formed exclusively on the surface of a slide glass, an analyte solution can be adsorbed on the surface of Pd-coated glass substrate, as sketched in (c) of Figure 3, for chemisorption or physisorption of the analyte molecules that can subsequently be detected by SERS, as sketched in (d).

Figure 3.Electroless deposition of palladium onto the surfaces of a slide glass and its usage to detect chemicals by SERS; (a) formation of surface bound Pd2+ ions to function as seeds for growth of Pd nanostructure, (b) actual formation of Pd nanostructures, (c) chemisorption and/or physisorption of analyte molecules onto Pd nanostructures for SERS analysis, and (d) SERS measurement taken by focusing laser light onto Pd nanostructures formed on the surface of a glass substrate.

Considering the fact that SERS usually occurs with aggregated structures of metal particles in the range of 20-200 nm, the as-prepared Pd-coated glass substrates would be expected to show SERS activity. Prior to evaluating the SERS activity, the optical properties of the prepared Pd films were examined. The UV-vis absorption spectra of the four Pd films are shown in Figure 4. There is no characteristic peak in the region of 300-1000 nm and only a gradual increase in absorption can be identified. We evaluated the performance of our Pd films deposited on glasses as SERS substrates using BT as a model compound. Figures 5(a)-5(d) show four typical SERS spectra of BT adsorbed on Pd(1:4), Pd(1:8), Pd(1:10), and Pd(1:40) films, respectively. The SERS peaks from the Pd(1:10) film are very intense but the peaks from other films are weak (see Figure 5(e)). It is notable that the Pd(1:40) film exhibits the least enhancement, probably associated with its relatively flat surface morphology; this result might highlight the importance of the gap or crevices among the metal nanostructures in SERS measurements.

Figure 4.UV-vis spectra and photographs (insets) of four different Pd films: (a) Pd(1:4), (b) Pd(1:8), (c) Pd(1:10), and (d) Pd(1:40).

We have attempted to estimate the SERS EF by using the following relationship:

where ISERS and INR are the SERS intensity of BT on Pd(1:10) film (Figure 5(c)) and the normal Raman (NR) scattering intensity of BT in bulk, respectively, and NSERS and NNR are the number of BT molecules illuminated by the laser light to obtain the corresponding SERS and NR spectra, respectively.24 ISERS and INR were measured at 1574 ㎝‒1, and NSERS and NNR were calculated on the basis of the estimated concentration of surface BT species, density of bulk BT, and the sampling areas. The equilibrated surface concentration of BT is assumed to be the same as that on Au and Ag, i.e. ~7.1 × 10‒10 ㏖/㎝2.31 Taking the sampling area (ca. 1 μm in diameter) as well as the surface roughness factor (~2.15) obtained from the AFM measurement of Pd(1:10) film into account, NSERS is calculated to be 1.2 × 10‒17 ㏖. When taking the NR spectrum of pure BT, the sampling volume will be the product of the laser spot and the penetration depth (~15 μm) of the focused beam.32 As the density of BT is 1.07 g/㎝3, NNR is calculated to be 1.1 × 10‒13 ㏖. Since the intensity ratio, ISERS/INR, is measured to be ~0.2 for Pd(1:10) film at 514.5 nm excitation, EF can then be as large as 1.8 × 103, which is comparable to previously reported values. Xia and coworkers reported an EF value of 1.3 × 103 for 4-mercaptopyridine adsorbed on Pd nanoboxes.33 For pyridine adsorbed on the electrochemically roughened Pd, Tian and coworkers reported an EF factor of 1.8 × 103.34 It should also be mentioned that as quoted in Figure 5(e), five different spots were randomly selected to take the SERS spectra; the peak intensities at 1574 ㎝‒1 were also normalized with respect to that of a silicon wafer used in the instrument calibration. The fact that the relative standard deviation was less than 10% for all Pf films clearly illustrates the homogenous characteristics of our Pd films.

Figure 5.SERS spectra of benzenethiol adsorbed on four different Pd films: (a) Pd(1:4), (b) Pd(1:8), (c) Pd(1:10), and (d) Pd(1:40). (e) Relative Raman peak intensities of benzenethiol at 1574 ㎝‒1 (ν8a) measured using (a) Pd(1:4), (b) Pd(1:8), (c) Pd(1:10), and (d) Pd(1:40); for each substrate, spectra were measured at five different spots.

We found recently that organic isonitriles adsorb fairly well on metals including gold, silver, and platinum, with their NC stretching frequencies being very susceptible to the kind of metals adsorbed35,36 as well as to the potentials applied to them.37 In order to evaluate an additional SERS performance of our Pd films, we have thus measured the SERS spectra of 2,6-DMPI, and the spectrum obtained from a Pd(1:10) film is shown in Figure 6(b). For reference, the NR spectrum of 2,6-DMPI in neat solid state is shown in Figure 6(a). In the NR spectrum, the NC stretching and the C–NC stretching bands appear at 2121 and 641 ㎝‒1, respectively, while the ring CC stretching (ν8a) and the inplane ring breathing (ν12) bands appear at 1593 and 995 ㎝‒1, respectively.38 In the SERS spectrum, the NC stretching and the C–NC stretching bands are observed at 2147 (1975) and 653 ㎝‒1, respectively. The ring CC stretching n8a band is observed at 1589 ㎝‒1.38 Hence, the NC stretching band at 2147 ㎝‒1 has blue-shifted by as much as 26 ㎝‒1 upon the surface adsorption on Pd, although the ring ν8a band was red-shifted by 4 ㎝‒1. All of these are due to the adsorption of 2,6-DMPI on Pd via the –NC group. In particular, the substantial blue-shift is associated with the antibonding character of the carbon lone pair electrons of the NC group.39

Figure 6.(a) NR spectrum of 2,6-DMPI in neat solid state; the inset shows the molecular structure of 2,6-DMPI. (b) SERS spectrum of 2,6-DMPI adsorbed on a Pd(1:10) film.

The donation of these electrons to Pd should increase the strength of the NC bond. Furthermore, there is another peak, even stronger, at 1975 ㎝‒1 in Figure 6(b). In an earlier investigation of SERS spectrum of 2,6-DMPI on laser-ablated Pt nanoaggregates, three bands appeared at 2166, 2124, and 1997 ㎝‒1 in the NC stretching region, and these three bands were attributed to the adsorption of 2,6-DMPI on the on-top, 2-fold bridge, and 3-fold hollow sites, respectively, of Pt nanoaggregates.40 It is therefore tempting to assign the 2147 and 1975 ㎝‒1 bands in Figure 6(b) to the NC stretching modes of 2,6-DMPI adsorbed on the on-top and the 3-fold hollow sites of Pd nanostructures on glass substrates, respectively. According to the theoretical work of Guy et al.,41 aryl isocyanide can be a good π acceptor. Thus, it may not be unusual to observe multiple NC stretching peaks due to the π back-donation from Pd. The NC stretching frequencies observable from 2,6-DMPI on Pd are even about 19-22 ㎝‒1 lower than their counterparts on Pt, suggesting that the p back-donation capability of Pd must be greater than that of Pt.39

 

Conclusion

In this investigation, we found that very stable, evenly deposited, and moderately SERS-active Pd films can be reproducibly fabricated simply by soaking glass substrates in ethanol-water (8:2 v/v) mixture solutions of Pd(NO3)2 and butylamine. It was revealed that a controlled amount (~20%) of water should be added in the ethanol solution to successfully deposit Pd nanostructures on a glass substrate. The formation and characteristics of Pd thin films were confirmed by FE-SEM, UV-vis, XRD, and XPS analyses. The Pd films deposited onto glass substrates consisted of aggregated granular particles whose mean grain sizes increased as the relative molar ratio of butylamine and Pd(NO3)2 increased. The as-prepared Pd films exhibited very even SERS activity and the enhancement factor estimated using BT as a prototype adsorbate reached 1.8 × 103 using a 514.5 nm excitation source. The close investigation of SERS spectrum 2,6-DMPI on Pd revealed that the 2,6-DMPI molecules may adsorb on the on-top and the 3-fold hollow sites of Pd nanostructures via the isocyanide group. Our method is cost-effective and is suitable for the mass production of homogeneous Pd films on glass substrates. Hence, this method will be useful in the development of Pd-based nanostructures, retaining simultaneously the advantages of high SERS as well as catalytic activities.