The synthesis and characterisation of gold nanorods have been carried out by reduction of the gold salt HAuCl4. This has been done using a single reducing agent, acetylacetone, rather than the two reducing agents, sodium borohydride and ascorbic acid, normally required by standard wet chemistry methods of gold nanorod formation. Using this novel method, the nanorods were synthesised at several different pH values which were found to greatly affect both the rate at which the nanorods form and their physical dimensions. The concentrations of acetylacetone and silver nitrate used relative to the gold salt were found to alter the aspect ratio of the nanorods formed. Rods with an average length of 42 nm and an aspect ratio of 4.6 can be easily and reproducibly formed at pH 10 using this method. Nanorods formed under optimum conditions were investigated using TEM.
The focus that anisotropic metallic nanoparticles and in particular nanorods have received from the scientific community in the past decade continues to grow. The expansion of interest within this field is driven primarily by the unusual optical, catalytic, magnetic and biological properties of these rod-shaped objects. There are three principle methods by which nanorods can be synthesised: template (Foss et al. 1992), electrochemical (Yu et al. 1997) and wet chemical methods (Jana et al. 2001; Nikoobakht and El-Sayed 2003). Of these three, synthesis by wet chemistry has emerged as the method of choice for obtaining nanorods due to the relative simplicity and lower cost of this technique as compared to the other two methods. Such development of the methods by which nanorods or nanoparticles in general are produced has undoubtedly made these objects more accessible to the wider scientific community and has led to break through applications of scattering (Huang et al. 2006) and two photon luminescent chromophores (Imura et al. 2005; Wang et al. 2005) for biological imaging, biological sensors (Li et al. 2005; Liao and Hafner 2005; Sudeep et al. 2005), carriers for drug delivery (Takahashi et al. 2005) and therapeutic agents for photothermal cancer treatment (Huang et al. 2006).
The purpose of this article is to outline a novel method for forming gold nanorods by wet chemistry. Although there are several methods in the literature for forming spherical nanoparticles, there are far fewer techniques for forming nanorods by wet chemistry. To date, all high-yield wet chemistry methods are restricted to the use of sodium borohydride as a means to producing nucleation of the gold salt and require ascorbic acid in order to grow nanorods on these seeds as the predominant nanoobject in the sample (Jana 2005). In this new one-step method, the reducing agent used is the β-diketone, acetylacetone (acac). The use of β-diketones has been reported previously as a means by which nucleation of gold salts can occur leading to the formation of spherical nanoparticles (Kundu et al. 2005), but it is believed that this is the first report of such chemicals being used in the formation of nanorods.
In a typical experiment, a sodium carbonate buffer (0.1 M) of pH 10 is used in conjunction with the surfactant cetyl trimethylammonium bromide (10 mL, 0.2 M) to form the aqueous solution in which nanorod formation takes place. To this solution was added HAuCl4 (0.1 mL, 0.1 M) resulting in an orange precipitate which slowly dissolved to form an orange-coloured solution. It is known that this precipitate results from the substitution of chlorine atoms on the gold salt by bromine atoms from the CTAB surfactant forming AuBr4−. This was followed by a small quantity of silver nitrate (0.018 mL, 0.1 M) and finally by the acac (0.1 mL, 0.35 M). On addition of acac, the solution rapidly changes from orange to colourless indicating reduction of the Au(III) salt. This is followed by the gradual appearance of a deep purple/red colour which corresponds to the appearance of peaks at 509 and 879 nm in the UV spectrum. These two peaks are due to the surface plasmon resonances of the nanorods, the peak at 509 nm being the axial surface plasmon resonance (ASPR) and the peak at the longer wavelength of 879 nm being the longitudinal surface plasmon resonance (LSPR). It is noted that the formation of nanorods in the presence of AgNO3 reported here does not follow the same reaction kinetics found when using the standard reductants NaBH4 and ascorbic acid in the presence of AgNO3. When these latter two reductants are used, the wavelength of the LSPR initially increases and then decreases just before the nanorods cease formation (Sau and Murphy 2004). For the method reported here, the wavelength of the LSPR of the nanorods continues to increase right up until formation stops (see Fig. 1).
In order to gain a greater understanding of how the nanorods develop under these novel reaction conditions, the pH, acac concentration and the effect of varying the AgNO3 concentration were studied. It is noted that the pKa value for acac (pKa = 9.0) is considerably higher than that of ascorbic acid (pKa = 4.2); therefore, it was expected that the optimum pH for nanorod formation would be very different for the two reactions, and so we started by investigating this parameter. It was found that the rate at which the nanorods form increases from 2 days at pH 6.5 to within half an hour at pH 12.5. The longest wavelength corresponding to the LSPR was obtained at pH 10. However, it is noted that there was only a very slight increase in the wavelength of the LSPR when the pH was increased from 8 to 10 followed by a very sharp decrease in the wavelength of the LSPR from pH 10 to 12.5 (Fig. 2a). The use of only an aqueous surfactant solution without any buffer present did not produce any nanoparticles, although addition of the acac to such a bufferless solution did produce a colourless solution indicating that reduction of the gold species had occurred. It is known that when using ascorbic acid as the reductant the acid itself is converted to its oxidised form via an enolate intermediate, while the gold salt is reduced and HBr is formed (see Scheme 1; Chen and Liu 2006). It is therefore assumed that when using acac it also reduces the gold salt via an enolate intermediate (see Scheme 1). This would mean that at low pH the acac reducing agent is predominantly in the keto-enol form and as the pH is raised the enolate ion becomes more stable. As the enolate anion becomes more available, it will begin to reduce the gold salt at a faster rate by abstracting the bromide ions and thereby forming 3-bromo-2,4-pentadione (see Scheme 1; Kundu et al. 2005). At the highest pH of 12.5, it is assumed that the rate of gold reduction occurs too fast which does not allow the CTAB surfactant to effectively stabilise the rod shape and thereby limit further growth of the nanorods. This would result in the larger number of spherical nanoparticles being formed relative to nanorods as is demonstrated in the UV spectrum taken at pH 12.5 in Fig. 2a. It has been found that the reaction conditions wherein nanorods are formed as the predominant shape within the sample occur between pH 8 and 10 which is centred on the pKa of acac. A similar phenomenon has been discussed previously for the formation of nanorods using ascorbic acid where the optimum pH for formation is centred on the pKa of ascorbic acid (Busbee et al. 2003).
Variation of the acac concentration demonstrated an initial increase in the wavelength of the LSPR with increasing concentration followed by a decrease. The maximum wavelength for the LSPR was obtained with a molar ratio of HAuCl4 to acac of 1:4 (see Fig. 2b). It was found that increasing the silver nitrate concentration gave a longer wavelength for the LSPR, but after a concentration of 0.18 × 10−3 M it was found that the peak in the UV corresponding to the LSPR began to broaden and reduce in intensity relative to the peak corresponding to the ASPR indicating that the polydispersity of the nanorods was increasing and also that the abundance of nanorods in the sample relative to spherical nanoparticles was being affected (see Fig. 2c).
TEM analysis of two samples of nanorods formed at pH 7.5 and 10 (see Fig. 3) was carried out. It was found that in the sample formed at pH 10 the nanorods were better formed with straight cylindrical lengths and curved ends whereas in the sample at pH 7.5 distorted versions of this typical nanorod shape were commonly found. In both samples a range of aspect ratios are found for the nanorods present and the mean values are given in Table 1. It was found that the nanorods formed at pH 10 had a larger aspect ratio to those formed at pH 7.5, and this is also demonstrated by the positions of the longitudinal plasmon resonance in the UV spectra (see Fig. 2a). Interestingly, the initial yield of nanorods to other nanoobjects was found to be slightly better when formed at pH 7.5 (66%) than at pH 10 (62%), and yet after one centrifuge cycle the higher pH nanorods gave a greatly improved ratio of 87% of nanorods to other objects as compared to those formed at lower pH (75%). This is due to the nanorods formed at higher pH having larger dimensions overall, and therefore, they are more efficiently filtered by the centrifuge process. The improvement in polydispersity of the nanorods in both samples after centrifuge, as demonstrated by a decrease in the relative standard deviation (RSD), is attributed to the loss of smaller nanorod-shaped particles in addition to filtering out other small nanoobjects. Previously, reported RSDs for nanorod samples are often between 25% and 5% after one centrifuge cycle (Jana 2005; Busbee et al. 2003) which compares favourably with the 16% reported in this article. More recent methods of nanorod synthesis using ascorbic acid and sodium borohydride are capable of forming near monodisperse nanorods in yields of 90–95%. (Jana 2005) However at the early stages of developing such methods, polydisperse nanorod samples were commonly reported with yields as low as 4% (Jana et al. 2001). As with all other wet chemistry methods used to form nanorods, small spherical nanoparticles were present in the crude solution of this method. Energy-dispersive X-ray spectroscopy showed that the centrifuged TEM samples consisted of only gold with minor amounts of silver, chromium and cobalt present (see Supporting Information).
Yield of nanorods relative to other nanoobjects and the average dimensions of the nanorods for reactions carried out at pH 7.50 and 10.0
Aspect ratio, RSD (%)
26 ± 15
7.3 ± 4.2
3.6 ± 1.1
35 ± 10
9.4 ± 2.3
3.7 ± 0.84
36 ± 14
8.5 ± 2.3
4.2 ± 1.2
42 ± 9.1
9.1 ± 1.2
4.6 ± 0.75
The error values refer to 1 SD
In conclusion we have investigated acac as a novel chemical for the reduction of HAuCl4 as a means of producing well-formed gold nanorods. This method involves only one reducing agent as opposed to two required for other methods and does not require the formation of seed nanoparticles prior to nanorod formation thereby simplifying the way in which nanorods are synthesised. We have systematically varied the parameters of pH, AgNO3 and acac concentration relative to the gold salt concentration in order to find the optimum conditions for nanorod growth. It has been found that pH has a dramatic effect on the rate at which the nanorods form, and this has been attributed to the presence of an enolate anion intermediate which is important for the reduction of the gold salt. Nanorods with an average length of 42 nm and an aspect ratio of 4.6 can be easily and reproducibly formed at pH 10 using this method.
The authors thank the Gobierno Vasco and the Diputación Foral de Gipuzkoa for financial support through the i-NANOGUNE Etortek project and the Spanish Ministry of Science and Innovation: Project HOPE CSD2007-0007 (Consolider-Ingenio 2010).