Oligonucleoside assisted one pot synthesis and self-assembly of gold nanoparticles

Deoxycytidine (C₉H₁₃N₃O₄ ⩾ 90%) and deoxyadenosine (C₁₀H₁₃N₅O₃ ⩾ 99%) were purchased from Sigma. Choloroauric acid (HAuCl₄ ⩾ 97%) was purchased from Loba Chemie. Sodium hydroxide pellets (NaOH ⩾ 98%), hydrochloric acid (HCl) and nitric acid (HNO₃) were purchased from Merck. All the glass wares utilized in the experiments were extensively washed with aquaregia in order to remove any pervious adsorbed metal particles and with double distilled water several times, and finally rinsed with milli-Q water (18.2 mΩ). Milli-Q water was used throughout the experiments. All the chemicals were used without any further purification.

A novel and facile one-pot route for synthesis of water-soluble AuNPs using single deoxynucleoside has been reported. Briefly, HAuCl₄ a stock solution of specific concentration was prepared and was isolated from light, in order to prevent any photo-induced reduction. The preserved metal salt solution was put on a temperature-controlled heater and was kept undisturbed at 150 °C. Within seconds, refrigerated dC solution (1 mM) was gently added, so that the concentration of HAuCl₄ in aqueous solution became 0.12 mM, and homogenized by gentle shaking. This mixture was left for heating, say about 15 min at 150 °C. During the reaction process the solution colour was transformed from very light yellow to deeper yellow in the case of reaction with dC; on the other hand, the solution colour changed from light yellow to orange while dA was used as the reaction agent. The as-synthesized AuNPs, obtained using dC and dA, will be referred as dC–Au and dA–Au throughout the report.

The as-synthesized AuNPs aqueous dispersions were kept in a dark environment. Interestingly, the AuNPs synthesized using dC (dC–Au nanoparticles) showed a spontaneous change in the solution colour as yellow → greenish yellow → green without any precipitation, indicating the self-assembly of AuNPs to bigger clusters. The self-assembled clusters of dC–AuNPs will be referred as dC-AuC. On the other hand, the AuNPs synthesized using dA (dA–AuNPs) were found to be stable for months without any change in the solution colour.

The surface characterization of AuNPs was carried out using Fourier transform infrared (FTIR) spectrophotometer, Perkin Elmer (USA) with resolution ±1.0 cm⁻¹. Dried powder of AuNPs and KBr were mixed and pelletized for FTIR measurements. The zeta potential measurements for AuNPs dispersions were carried out using Zetasizer Nano ZSP, Malvern (UK). The measurements were in regular time intervals to monitor the change in zeta potential of AuNPs dispersions. The UV–visible spectrophotometer, Shimadazu UV 2470, with a quartz cuvette of 1 cm path length was employed for optical absorbance measurements. The dynamic light scattering (DLS) measurements were performed using a home-made setup [31, 32] at 90° scattering angle and data were analysed using the CONTIN algorithm [32] to determine the hydrodynamic diameter (D_H) of AuNPs. Direct images at different stages of self-assembling of dC–AuNPs were obtained using a high-resolution transmission electron microscope (HRTEM, FEI Technai G230). X-ray diffraction (XRD) patterns of dried powders of dC–AuNPs, dA–AuNPs and dC–AuC were obtained using the Bruker D8 Advance.

Figure 1(a) shows the FTIR spectra of dA (filled squares) and dA–AuNPs (open squares) and figure 1(b) shows those of dC (filled circles) and dC–AuC (open circles). Figure 1(a) shows prominent peaks within 1500–1800 cm⁻¹ representing the exocyclic amino region of dA (see scheme 1(a)). In the case of dA–AuNPs, reduction of the intensity of the peak at 1654 cm⁻¹, which is assigned to NH₂ scissoring, might indicate that the respective functional group is responsible for the reduction of the metal salt. The absence of 1604 cm⁻¹ peak, which is the combination of NH₂ and ${\rm C}-\!\!\!-{\rm N}$ scissoring and stretching vibrations [33], in dA–AuNPs spectrum indicates that dA might adsorb onto the AuNP surface through ${\rm C}-\!\!\!-{\rm NH}_2$ as reported [34]. Similarly, dC–AuNPs were surface functionalized with dC (figure 1(b)), which has further instigated the self-assembling of dC–AuNPs to bigger clusters, i.e. dC–AuC, in aqueous medium at pH ∼ 6.5. The absence (or a very weak presence) of the absorption peak at 1616 cm⁻¹, which corresponds to the bending vibrations of ${\rm H}-\!\!\!-{\rm N}-\!\!\!-{\rm H}$, indicates that NH₂ played an important role in the synthesis of dC–AuNPs. A shift of the peak at 1292 cm⁻¹, which corresponds to the pyrimidine ring of dC (see scheme 1(b)), towards longer wavenumber region indicates that it might adsorbed on to the AuNP surface, as reported [35].

Zoom In Zoom Out Reset image size

Figure 2 shows the surface plasmon resonance (SPR) absorption peaks of as-synthesized dC–AuNPs and dA–AuNPs dispersions around 534 and 538 nm, respectively. Insets of figure 2 show the photos of as-synthesized dispersions.

Zoom In Zoom Out Reset image size

The SPR peak position λ_SPR of dC–AuNPs was found to shift with time towards longer wavelength regime up to around 628 nm (figures 3(a)–(c)), indicating the self-assembling process and the formation of stable aggregates of cauliflower type structures in the dispersion. As a consequence, the colour of dC–AuNPs dispersion was seen to vary spontaneously from pale yellow to deep green (insets of figure 3(c)). The final dispersion was stable for a long period (see figure S1 of supplementary information, available from stacks.iop.org/ANSN/4/045014/mmedia). On the other hand, dA capped AuNPs (dA-AuNPs) dispersion did not show any change in the SPR absorption peak position over a period of one month or more, indicating no aggregation (figure S2 of supplementary information) took place in the as-synthesized dispersion. It has also been observed that dA–AuNPs were spherical in morphology of size around 23 nm and were sterically capped by dA, which has acted as both a reducing and a stabilizing agent (see figure S4(a) of supplementary information).

Zoom In Zoom Out Reset image size

As discussed earlier, each dC has one oxygen (O) atom in the benzene ring (scheme 1(b)) which is capable of developing a hydrogen bond with the $-\!\!\!-{\rm OH}$ group of the neighbouring dC–AuNPs, and hence, forming a cascade structure (dC–AuC) as shown by a cartoon in figure 3(d). Hydrogen bonds are weak, and therefore, the self-assembled cascade structure can be easily broken.

Using DLS we have investigated change in the particle size (D_H) with time, as a consequence of aggregation, in both dC–AuNPs and dA–AuNPs dispersions. Figure 4(a) shows the increase of D_H of dC–AuNPs with time from 31 to around 260 nm, indicating the self-assembly of as-synthesized nanoparticles with time to a stable structure. The kinetics of self-assembly was sluggish. The D_H of dA–AuNPs was also measured using DLS and found almost unchanged with ageing time (figure S2(b) of supplementary information, available from stacks.iop.org/ANSN/4/045014/mmedia).

Zoom In Zoom Out Reset image size

Using TEM we have investigated the size of dC–AuNPs and dC–AuC. Figure 4(b) shows the image of as-synthesized dC–AuNPs and figure 4(c) shows the histogram obtained from TEM image of as-synthesized dC–AuNPs. The average size was found to be around 10 nm which further self-assembled to the stable cauliflower type structure (dC–AuC) of size ∼235 nm in around 25 days (figure 4(c)). This is in consistency with the DLS result. The intermediate stages of assemblies were also monitored using TEM and was found systematic (see figure S3 of supplementary information). XRD patterns for dried dC–AuNPs and dC–AuC were also taken (see figure S5 of supplementary information) which showed the presence of Au (111) peak.

We have measured the zeta potential of each dispersion to investigate the stability of the dispersion and given the result in table 1. The zeta potential of the dC–AuNPs dispersion was very low (∼ 0.75 mV), indicating that the as-synthesized dC–AuNPs were unstable and formed self-assembled bigger clusters (dC–AuC) of size ∼235 nm. The zeta potential of the aggregates was high (∼ 32 mV) indicating the stabilization of the dispersion. On the other hand, the zeta potential of the as-synthesized dA–AuNPs dispersion was high (∼ 29 mV) revealing that the dispersion was stable, as we have discussed in earlier sections.

It was observed that optical property of the dC–AuC dispersion could be varied by changing the pH. Change of the self-assembled structure by changing the dispersion pH was studied by observing the shift in SPR peak position of dC–AuNPs dispersion, as well as by measuring the change of D_H in the corresponding DLS measurements. Interestingly, the clustering was found to be reversible on both sides of the neutral pH (∼ 6.5) of dC–AuC dispersion.

The SPR position shifts towards lower wavelength side, around 530 nm at both acidic and basic pH and again returns back to 628 nm when the pH of the solution was brought back to neutral. This indicates that a slight variation in the solution pH (figures 5(a) and (b)) could break the hydrogen bonds between $-\!\!\!-{\rm OH}$ and $=\!\!\!={\rm O}$ of neighbouring dC–AuNPs (as discussed in earlier section). This kind of pH dependent self-assembling of nanoparticles can be utilized for advanced delivery application in cancer therapy [12].

Zoom In Zoom Out Reset image size

It is to be noted that there are very feeble signatures of other plasmon modes at different solution pH corresponding to few intermediate structures of unknown size of dissociating (or re-associating) clusters. The decrease of cluster size of dC–AuC with the solution pH was also investigated using DLS. Figure 5(c) shows the variation of D_H of the cluster with variation of pH from 6 to 12.

We have also observed that upon heating above 35 °C, the green dC–AuC dispersion changed to the yellow dispersion which remained stable for a long period of time, indicating that clusters were permanently dissociated, and no further clustering was taking place.

We have generated nanoparticle–nanoparticle interaction plots for dC–AuNPs dispersion at different ageing times using the following Derjaguin, Landau, Verwey and Overbeek (DLVO) expression [36]

$$\begin{equation} V_{{{\rm DLVO}}} (x) = V_{{{\rm elec}}} (x) + V_{{{\rm vdw}}} (x), \end{equation} \tag{ 1 }$$

here x is the distance between the NPs. Considering the case of κa > 1, where κ is the inverse of the Debye length, and a is the particle radius, the electrostatic repulsion can be expressed as

$$\begin{equation} V_{{\rm elec}} = \frac{{\varepsilon a\phi_o^2 }}{2}{\rm exp}\,( - 2\kappa ax), \end{equation} \tag{ 2 }$$

where ε is the permittivity of the solvent and ϕ_o is the surface potential of the particles and the Debye parameter $\kappa = \sqrt {\frac{{8\pi ce^2 z^2 }}{{\varepsilon k_{\rm{B}} T}}}$, where c is the concentration of the z valent ionic species and e is the charge of proton or electron. The simplified van der Waals attractive potential can be written as

$$\begin{equation} V_{{{\rm vdw}}} = - \frac{A}{{12 x^2 }} . \end{equation} \tag{ 3 }$$

In the calculation, z for dC–AuNPs has been taken as 1; ϕ_o has been taken as the experimental values of ζ-potential for as-synthesized and 12 days aged dC–Au dispersions. The Hamakar constant A has been taken as k_BT (in joules) [36]. The interaction energy versus particle–particle distance for dC–AuNPs before and after ageing for 12 days (obtained using equation (1)) is plotted in figure 6. It is clear that the interaction potential was attractive (or weakly repulsive) in as-synthesized dC–AuNPs dispersion and caused self-assembly, as mentioned in previous sections. The self-assembly was automatically controlled by the interaction sites of dC molecules on nanoparticles surfaces. The interaction potential subsequently became more repulsive as ageing time for dC–Au dispersion increased, and appeared constant after forming stable cauliflower-type clusters.

Zoom In Zoom Out Reset image size