1. Introduction

Nanoclusters draw much attention in materials science because they show quite different properties from their bulk counterparts due to the so-called ‘quantum size effect’. Nanoclusters are also important in industrial fields such as catalysis, sensors, electronic devices, magnetic materials and optics (1–6). The preparation of stable nanoclusters with monometallic and bimetallic compositions colloidally dispersed in solution in the presence of protecting polymers has been reported (4, 7, 8).

Bimetallic alloy systems have been known and exploited for many years in various catalytic reactions such as promising anode catalysts for direct formic acid fuel cells (9). The addition of a second metal in a bimetallic particle provides a way to control the activity and selectivity of the resulting particles for a variety of reactions. By varying the ratio of the two constituents, the distribution of the compounds at the surface may also be altered. In this way it is possible to tune the chemical reactivity at the surface of an alloyed particle (10).

Nanoparticles of palladium and its alloys have been successfully applied to catalyse various chemical reactions. One widely known example is the palladium/tin colloidal solution used as an activator for electroless Cu deposition in printed circuit board manufacture. Electroless Cu deposition (11) has been extensively employed in the plating through hole technique in the printed circuit board industry, Cu interconnection for ultra large-scale integration and circuit fabrication for large-scale liquid crystal display panels.

The preparation of bimetallic systems is not trivial. Often not only alloying but also small particle size and a narrow particle size distribution are required. Various preparation techniques are available, for instance, coimpregnation and coprecipitation. More advanced techniques have also been used, such as impregnation with bimetallic precursors or sequential impregnation in which the second precursor is deposited on the surface of the first precursor. If bimetallic particles are not formed during the impregnation step, one may rely on succeeding steps such as calcination and reduction. The bimetallic nanoparticles formed by these techniques usually exhibit tailed structure and high activity.

Ultrasonic non-destructive evaluation techniques are widely used for the characterisation and analysis of physical and thermal properties of various types of materials. Both theoretical and experimental studies using ultrasonic techniques have been performed in the field of materials science (12, 13).

In this paper we report a novel chemical method for the synthesis of Cu/Pd nanoclusters. The method uses sodium citrate as a complexing agent added to the metal precursors. The obtained Cu/Pd nanoparticles had good stability and were well dispersed with particle sizes in the range 5–15 nm. They were characterised by UV-vis spectroscopy, XRD and TEM. An acoustic particle sizer was also used for particle size distribution analysis, and was found to give accurate results comparable to the XRD and TEM data. The advantage of this technique is that no sample preparation is required for the measurement and the instrument is easy to handle.

2. Experimental

3. Results and Discussion

To confirm whether the metal complex ions were formed after the addition of trisodium citrate, the complexing behaviour was investigated by UV-vis spectroscopy. Figure 1 depicts a strong absorbance band near 250 nm after the addition of trisodium citrate into the copper sulfate solution. Similarly, the UV-vis spectrum of the Pd precursor mixed with trisodium citrate also exhibited an absorption band near 260 nm.This UV-vis absorbance band confirms the formation of metal ion complexes triggered by the complexing agent trisodium citrate. The absence of absorption peaks above 300 nm in all the samples confirmed the reduction of Pd(II) ions (14). This type of behaviour was also found by Yonezawa et al. (15).

Figure 2 shows the typical XRD pattern for these Cu/Pd nanoclusters, indicating the formation of bimetallic nanoparticles. The obtained peaks were indexed using Joint Committee on Powder Diffraction Standards (JCPDS) files (now renamed the International Centre for Diffraction Data (ICDD)) (JCPDS File No. 04-0836 and 05-0681). One broad main peak located between 2θ = 40^° and 43^° was observed in the system (Figure 2). The broadened shape indicates a reduced grain size, as expected for nanoparticles. The location of the peak, between those characteristic of Pd nanoparticles (2θ = 40^°) and Cu nanoparticles (2θ = 43^°), corresponding to (111) planes, represents the formation of a disordered solid solution between Pd and Cu.

Figure 2 reveals small peaks at 2θ = 31.8^° and 45.8^° in the system with complexing agent. Comparison of the peaks at 2θ = 31.8^° and 45.8^° with those found in the JCPDS (No. 46-1211, palladium(II) oxide (PdO): 2θ = 31.7^° (200) and 2θ = 45.6^° (220)) suggests the existence of a cubic PdO structure. The full width at half maximum of this peak is much smaller than that of the characteristic peak of Cu/Pd nanoparticles, indicating PdO structures with an enlarged crystal size. Since the XRD samples were prepared with all particles in solution (including the large precipitated particles initially formed and the precipitates after centrifugation with the addition of acetone), the PdO peak may have arisen from the large particles initially formed, which were not observed in the TEM. Without complexing agent there was a peak at 34^° which was due to the formation of copper(II) hydroxide (Cu(OH)₂). However, Figure 2 reveals no Cu(OH)₂ crystalline structures, implying that the complexing agent may have prevented the formation of Cu(OH)₂ particles, which accounted for the substantially higher stability of Cu/Pd nanoparticles synthesised with the complexing agent. The obtained peaks are well identified by the reference values given in JCPDS (No. 48-1551) (11).The crystallite size was also calculated by Debye Scherrer's formula (16) as 14 nm.

TEM was used to determine the size distribution and morphology of the synthesised nanoparticles and SAED was used to confirm the crystallinity of the samples. The TEM images are shown in Figure 3. The TEM images show that the size distribution of the synthesised bimetallic Cu/Pd nanoparticles is in the range 5–15 nm (Figure 3(a)). It can be observed from this figure that most of the particles are >10 nm, with only a few at ~5 nm. The nanoparticles are well dispersed. The SAED pattern shown in Figure 3(b) corresponds to a crystalline structure – a result consistent with the XRD results.

The acoustic particle sizer was then used to measure the particle size distribution. The results are shown in Figure 4. The lower limit of this technique is 10 nm, which leads to the sharp line seen at this value on the particle size distribution graph. This analysis confirms that the Cu/Pd nanoclusters are in the range 10–15 nm.

APS can be used to perform many repetitive measurements for optimal signal averaging in order to maximise resolution, accuracy and reproducibility. Acoustic attenuation must be measured at multiple spacing for two reasons: (a) high frequency measurements have higher attenuation so they must be made over short paths, whereas at low frequencies, longer path lengths are required due to much lower attenuations; and (b) the attenuation versus frequency curve must be built with as many data points as possible in order to produce reliable particle size distribution data. The attenuation level, as well as the shape of the acoustic attenuation curve, is related to the particle size distribution. The particle size distributions are calculated from the acoustic attenuation data using software based on Epstein and Carhart theory (17).

4. Conclusions

Cu/Pd nanoclusters have been synthesised successfully in aqueous solution under ambient conditions with the addition of a complexing agent, trisodium citrate. These Cu/Pd nanoparticles were stable in suspension. The TEM image and SAED pattern showed a uniform dispersion of crystalline Cu/Pd nanoparticles. Particle size distribution analysis by the acoustic particle sizer was consistent with the TEM analysis and showed particle sizes in the range 10–15 nm. Hence this technique can be considered a very useful and efficient tool for the non-destructive characterisation of bimetallic nanoclusters. It is hoped that this work will prompt future study and characterisation of bimetallic nanoparticles containing platinum group metals for a variety of applications.