Abstract

Using soft acids as reducing agents in synthesizing metallic nanoparticles have constituted a clear framework for achieving selected morphological properties with minimal toxicity. The system's complexity and the many variables involved represent a challenge for experimental studies desiring to design reproducible synthesis protocols. In this work, we explore the exclusion of any stabilizing agent to synthesize, in an aqueous solution, non-spherical gold nanoparticles (AuNPs) via Microwave-assisted synthesis, and instead, we employed pH control over reducing agent L-Ascorbic acid (AH₂). The use of AH₂ presents a direct approach that allows an understanding of the role of soft acids in synthesizing metallic nanoparticles. The results indicate that AuNPs synthesized at pH ≥ 10 exhibit relatively different morphologies than those obtained at higher pH values. The AuNPs were characterized via Transmission electron Microscopy and UV-Vis Spectroscopy. Our simulations reveal the plasmon distribution according to particle shape. The experimental analysis suggests that the pH variation mechanism over the reduction agent correlates with AuNPs geometry. These results indicate that pH is an applicable parameter for controlling the nanoparticle's geometry and extend the possibility of exploring computational studies on the impact of acids adsorbed on gold colloidal surfaces.

1. Introduction

Gold nanoparticles (AuNPs) have been widely studied in experimental and theoretical approaches because of their attractive electric and optical properties ¹. Furthermore, the AuNP material is a conductor that, under normal conditions, crystalizes in a face-centered cubic cell (fcc) ². Therefore, AuNPs' electric and optical properties can be used in wide-ranging applications ³, such as photovoltaic devices ⁴, OLEDs ⁵, perovskite solar cells ⁶, or photocatalysis ⁷.

Several methods have been described for synthesizing AuNPs; recently, some authors ⁸ extended the understanding of the Turkevich method ⁹, which is considered a milestone in AuNPs preparation. Additionally, a tuning of the AuNP shape was proposed by Murphy. et al. ¹⁰ by a seed-mediated growth procedure, and most recent research has tuned particle shape using two different capping agents (surfactants/polymers) ¹¹.

Usually, shape tuning is promoted by using strong reducing and stabilizing agents, such as Ascorbic acid (AH₂) and CTAB, respectively ¹². Numerous applications and synthesis approaches have aimed to control AuNPs' shape since their optical properties, such as the Localized Surface Plasmon Resonance (LSPR) effect ¹³, strongly depend on the AuNPs' shape ¹⁴. One of the methods used has been micro-wave-assisted synthesis ¹⁵; for example, Kwok Wei Shah et al. achieved hexagonal AuNPs reduced by organosilane ¹⁶. The microwave method avoids side reactions due to selective heating ^{17, 18}, in addition to increasing the solution temperature based on direct heating of the sample without using the volumetric flask as an intermediate ¹⁹, and offers short synthesis times compared to synthesis routes based on conventional heating ^{20, 21}.

Early studies have synthesized spherical AuNPs using ascorbic acid at room temperature, e.g., Himanshu Tyagi et al. obtained a narrow size distribution (31±5, 36±6, and 40±5 nm) of spheroidal AuNPs readily stabilized by adjusting the reaction solutions' initial pH conditions ²². Later, Shohifah Annur et al. developed spherical AuNPs of different sizes range (20-40 nm) at room temperature by adjusting the pH of ascorbic acid at a range of 2.0 to 10 ²³. However, these methods have obtained only spherical gold nanoparticles, limiting the LSPR effect exclusively to a dipolar field distribution ²⁴. Moreover, these methods reduced the possibility of getting quadrupolar, hexapolar, and octupolar plasmon field distributions ²⁵. Therefore, the particle applications range ²⁶ is limited in areas such as bifunctional NO2 light sensing ²⁷, light-sensing devices ²⁸, or photocatalysis process ^{29, 30}.

Therefore, the preparation of anisotropic colloidal gold nanoparticles represents an experimental challenge to control the size and morphology to benefit from their excellent optical properties provided by the LSPR effect in various applications.

Consequently, in this work, we have explored an experimental approach to synthesize, to some extent, anisotropic gold colloidal nanoparticles using microwave-assisted synthesis by controlling the pH value of ascorbic acid (AH₂) using it exclusively as both a reducing and stabilizing agent.

2. Materials and Methods

All chemicals were of analytical grade and were used without further purification. Gold (III) chloride acid trihydrate (HAuCl₄⋅3H₂O) (99.9%) and L-Ascorbic acid(C₆H₈O₆) (99%) (AH₂) were purchased from Sigma-Aldrich. Sodium hydroxide (NaOH, pastilles, from Molar) was used to adjust the pH. The water used in all experiments was prepared using a Milli-Q Integral water purification system (EMD Millipore Direct -Q® 3UV-R) and had a resistivity of 18.2 MΩ cm. Samples were prepared using a CEM Focused Microwave™ Synthesis System, with a continuous microwave power delivery system with operator-selectable power output from 0- 300 watts. Optical absorption studies were carried out using an ultraviolet-visible–near-infrared (UV-vis-NIR) spectrophotometer (JASCO V-670). Particle size and morphologies were studied using a transmission electron microscope (TEM, JEOL JEM-2010). Interplanar spacing in the AuNPs was studied using high-resolution (HR)-TEM (Tecnai F20 microscope with FEI, acceleration voltage 200kV).

Unifactorial experiments with two replicates optimized synthetic conditions. First, different molar ratios of HAuCl₄, AH₂, and pH were tested. Briefly, 50 ml of 100 mM AH₂ solution was prepared and separated into vials of 4.5 ml. Next, different Molar ratios R = [NaOH]/ [AH₂] were tested to obtain a pH variation on each sample (R from 0 to 2, steps of 0.2 each). The NaOH/AH₂ mixtures were left in closed vials under constant stirring for three hours at room temperature under ambient conditions to reach pH equilibrium. Then, at room temperature, 45 µl of each pH-adjusted AH₂ solution was added to 4.5 ml of a 0.5 mM fresh Gold (III) chloride acid trihydrate (HAuCl₄⋅3H₂O) solution. The mixtures were immediately mixed and radiated at 200°C, 15 bar, and 300 W for 30 sec using a microwave reactor system. The samples were left to cool down at room temperature without an external cooling source. Subsequently, the samples were centrifuged at 1000 rpm, and the supernatant was extracted as the final product.

The ImageJ software, with standard plugins, was implemented to determine particle distribution and population, with a minimum of 56 particles per micrograph. The polydispersity (P) was estimated employing the standard Equation 1:

(1)

Where σ is the standard deviation, σ = FWHM/2, FWHM is the full width at half maximum of the total distribution, and s is the mean particle diameter (in nm) ³¹. Gatan Digital-Micrograph software was used to analyze the (HR)-TEM micrographs. In addition, the crystallographic toolbox (CrysTBox) software was used to study diffraction patterns obtained from (HR)-TEM micrograph ³².

We used the boundary element method (BEM) ³³ to calculate the electric field distribution and the local photonic density of states (LDOS) ³⁴ on the gold nanoparticles. We implemented the methodology described in the MNPBEM (Metallic Nano Particles BEM) MATLAB toolbox ^{35, 36}. The underlying idea of the BEM method is to solve Maxwell's equations for a dielectric system. The system is modeled as elements with homogeneous and isotropic dielectric functions separated by sharp interfaces ^{37, 38}. Hence, the method employs a geometrical mesh to discretize the interfaces aiming at solving Maxwell's equations using the proper boundary conditions ^{39, 40, 41}. Input parameters were the absorption wavelength, particle size, and particle shape obtained from the experimental results of steady-state absorption studies and TEM measurements. In addition, the optical constant for bulk gold was obtained from reference data ⁴².

3. Results

The reduction of gold was achieved by employing only AH₂ in a pH-controlled solution. The proposed method produced AuNPs in red color solutions with increasing intensity concerning pH value (see Figure S2- Supplementary Material). At molar ratios (R) greater than 1.2 (pH ≥ 10), the more intense color was obtained and remained unchanged at higher R. As will be discussed later, this result was directly related to the morphology of AuNPs. In this regard, the AuNPs synthesized at pH = 10 were chosen for further analysis.

TEM image (Figure 1) shows AuNPs (at 100 nm resolution) obtained by reacting gold (III) (HAuCl₄⋅3H₂O) with AH₂ at pH 10. This image shows that the pH reaction conditions yield non-spherical particles with increasingly amorphous shapes. The average particle size was determined by analyzing twenty TEM micrographs and a minimum of 56 particles per micrograph (Supplementary Material Figure S1 shows a gallery of a portion of TEM micrographs used to calculate the particle size distribution pH 10) and standard imaging protocols with the ImageJ software package. The results from the particle size study are given in the histograms in Figure 1(b). A Gaussian fit to the size distribution shows that the pH 10 reaction conditions yield 25 ± 13 nm diameter particles. The polydispersity for the sample set was calculated as 52%.

Figure 2(a) shows a TEM image of 20 nm resolution, suggesting different particle shapes. The electron diffraction pattern from the TEM measurements is shown in Figure 2(b). The primary diffraction peaks correspond to (111), (200), (220), and (113) reflections, which are characteristic of the gold fcc structure ². The surface structure of the particles was further analyzed using high-resolution TEM, as shown in Figure 2(c). Two well-resolved (111) and (200)-type crystallographic planes were detected (measured D-spacing of about 2.32 Å and 2.03 Å, which run perpendicular to the sidewalls of the structure. Figure 2(d) shows the corresponding electron diffraction profile from Figure 2(b) for the nanoparticles set at pH 10. Again, the diffraction peaks' relative intensities vary, suggesting a difference in the particle shape, as Figure 2(a) shows.

The AuNPs' optical absorption spectrum was studied between 400 to 800 nm and was fitted as a sum of Gaussian functions ⁴³. The synthesized sample exhibit a typical AuNPs absorption peak over ~525 nm (Figure 3(a)), which corresponds to absorbed green light and agrees with Localized Surface Plasmon Resonance (LSPR) frequency (ω) from previous AuNPs reports ⁴⁴. Figure 3(b) shows the scattering cross-section (σ_Sct) result computed numerically for a 25 nm diameter AuNP and an incoming plane wave excitation polarized along x, between 400 to 800 nm. The 25 nm diameter AuNPs' light scattering was calculated using the MNPBEM toolbox ^{35, 36}, and the size obtained in the histogram analysis (Figure 1(b)) was used as the input parameter. As a result, a value of approximately 26 nm² was obtained for σ_Sct. Figure 3(a-b) shows that the scattering and absorption spectrum's peaks consist primarily of yellow and green light ⁴⁵. Therefore, the non-scattered or absorbed light compounds the red transmitted light by the sample (image color sample Figure 3(a)).

A qualitative study of electric field distribution (EFD) and the electron energy loss probability of the synthesized AuNPs were computed using the simulation toolbox MNPBEM for plasmonic nanoparticles ^{35, 36}. First, the EFD has computed considering four radiation conditions described by the combination of indices uⁱ_α; i with i=1,2,3 indicating the polarization direction (1= x, 2 = y, and 3 = z) ⁴⁶ and (x, y, z) the direction of the incident field (525 nm obtained from the Absorption spectrum, Figure 3). Then, based on the TEM results (Figure 2(a)), the EFD was calculated for a nanosphere, nanotriangle, and nanohexagon, as shown in Figure 4(a). The results show that although the incident field is the same (525 nm), the EFD around the particles changes as the incident field polarization changes, which is associated with the particle shape influence over the plasmon distribution of metallic nanoparticles ^{35, 36}.

However, more precise information on plasmonic states in metallic nanoparticles is required for many applications, such as photonic or optoelectronic devices ^{47, 48, 49}. A technique such as Electron energy loss spectroscopy (EELS) is an effective tool for investigating plasmon modes in single and coupled metallic nanoparticles ^{50, 51, 52}. Thus, the EELS probability map for the same single nanotriangle and nanohexagon was additionally computed. In a 50nm square grid, the nanoparticles were localized at 10 nm from the plane, as shown in Figure 4(b). The particles were impacted with 2.74, 2.36, 2.29, and 2.21 eV, equivalent to the local wavelength values in the maximum absorption point (Figure 3). The results show that once the particles are excited with photons of 2.74 eV ~500nm, above the resonance energy (2.36 eV~ 525nm), the plasmon distribution for both particles is localized in the particle surface, as shown in Figure 5b - 2.74eV case. When the particles are excited with the resonance energy, 2.36 eV~ 525nm, the plasmon field adopts a different distribution for both particles and is localized on the edge of the particles (Figure 4b - 2.36eV case). A similar distribution is localized for the photoexcitation cases 2.29eV~540nm and 2.21 eV~560nm.

4. Discussion

TEM results (Figure 2) showed that, on average, particles associated with pH 10 were approximately 70% bigger than the particles obtained by different synthesis reaction pH approaches ⁵³. Earlier works had shown that when AH₂ was used as a reducing agent to produce AuNPs, the size and shape did not depend on the reactant concentrations ^{54, 55}. Previous results showed that pH control over AH₂ induces size particle change ^{22, 23}. However, our tuning-pH approach synthesis demonstrates not only size particle change but also induces particle shape, as shown in the TEM results from Figure 2(a and d) and Figure S2-supplementary material. Earlies microwave-assisted synthesis has obtained self-supported superstructures based on 1-dodecanethiol ^{56, 57}, known for its excellent properties as a sulfur source, capping ligand, and size controller ⁵⁸. This synthesis approach's success is based on stabilizing agents such as 1-dodecanethiol, poly(vinyl pyrrolidinone) (PVP) ¹⁵, or CTAB ⁵⁹. In comparison, the results obtained in this investigation show the possibility of conjugating non-spherical gold nanoparticles with the exclusive use of AH₂ as a reducing agent and a microwave source (Figure 2(a and d) and Figure S2-supplementary material).

In general, below pH 10, AuNPs were formed as conglomerated spherical particles. However, at pH ≥ 10, a more discrete size distribution (25 ± 13 nm) was obtained. An explanation of the pH effect on AuNP morphology is shown in Figure 5. This mechanism was based on the results of a recent depth analysis of the oxidation mechanisms of ascorbic acid ⁶⁰.

Multiple AH₂ molecules can be coordinated to Au³⁺ atoms since AH₂ have multiple positions with high electronic density (hydroxyl groups). As increasing pH, AH₂ is deprotonated to form ascorbate AH-. An electron from each coordinated AH- is transferred to Au³⁺ to reduce it to Au⁰, producing a radical AH∙. As the pH increased, the second pka was reached, and the second deprotonation of ascorbic acid took place to form the radical anion A.-. This process started occurring at pH ≥ 10. Later, the second release of an electron from each A.- equivalent reduced Au³⁺ to Au⁰, and dehydroascorbate A was finally formed. Note that at increasing pH with every deprotonation of ascorbic acid, anionic species were formed, then the negative charge of the corresponding ascorbates was available to coordinate to Au⁰ and stabilize the AuNPs. In other words, increasing pH not only improved the reductive effect over gold but also decreased the possibility of multiple coordination positions of ascorbic acid to the Au⁰ atoms, stabilizing the structure and forming more regular shapes of AuNPs.

Furthermore, as F. J. García De Abajo et al. ³⁹ had suggested, the EELS probability results could provide direct information on the local photonic density of states (LDOS) ⁶¹ which plays a crucial role in plasmonic nanoparticles' optical properties ^{34, 62}.

5. Conclusions

Using soft acids to synthesize AuNPs combines functionality with suitable stability. Many experimental degrees of freedom and synthesis complexity require merging controlled experimental conditions and simplified yet relevant computational models. We used Ascorbic acid, AH₂, to synthesize non-spherical AuNPs in an aqueous solution with controlled pH via microwave-assisted synthesis and without any reducing agent. We show that, to some extent, the resulting AuNP shape depends on pH-microwave radiation. The combination of Transmission electron microscopy and UV-Vis spectroscopy of the adsorption of ascorbic acid on gold nanoparticle surface reveals a shape distribution and standard adsorption.

Furthermore, the plasmon modes are exposed by employing a qualitative simulation of electric field distribution and the electron energy loss probability map of the synthesized AuNPs. Ascorbic acid mainly binds to gold surfaces via its negative charge of the ascorbates group. Thus, pH 10 exhibits more robust binding to gold surfaces than lower pH values. The underlying mechanism in the Ascorbic acid-Au interaction deserves further examination, with reducing agents such as ascorbic acid and any other reducing-stabilizing agent whose structure is modified as pH changes.

Acknowledgments

R.BC. acknowledges the financial support from the National Research and Development Agency (ANID-Chile) National Ph.D. scholarship from ANID /National Doctorate (N° 2016–21160562). Additionally, R.BC thanks Dr. Charusheela Ramanan for her suggestions and discussions in this research and extended thanks to Max Planck Institute for Polymer Research (MPIP) for using the microwave reactor to carry out the synthesis, and thanked Katrin Kirchhoff from MPIP for helping us with the TEM and HR-TEM measurements. Likewise, R.BC. Thanks to Dr. Shirly Espinoza for providing the synthesis material. E.BM. acknowledges the financial support from PAIP-UNAM 5000-9192.

References

Supplementary Information