Fast One-Step Synthesis of Anisotropic Silver Nanoparticles

: The shape control of metal nanoparticles, along with the size, is critical for most of their applications as they control their optical properties. Anisotropic metal nanoparticles show superior performance in a number of applications compared to spherical ones. Shape control is usually achieved by a two-step process, where the ﬁrst involves the formation of spherical nanoparticles and the second is about the actual shape transformation. In this paper, we report on a fast and facile synthesis of silver nanoplates in a single step, involving laser ablation of a silver target in a liquid medium while this is exposed to light irradiation and hydrogen peroxide ﬂow. We obtained anisotropic particles with a mixture of shapes, of 70–80 nm in size and 10–20 nm in thickness, which showed a plasmon sensitivity greater than 200 nm/RIU.


Introduction
Metal nanoparticles (NPs) have generated a great deal of interest in a range of fields including sensing, photonics, biology, and catalysis, owing to the unique interaction of their conduction electrons with electric fields [1][2][3][4]. Such an interaction is commonly known as Surface plasmon resonance (SPR) and is strongly related to the size and shape of the NPs, as well as the surrounding medium, which makes them particularly suitable for sensing applications [5]. While spherical NPs are easily obtained by chemical or physical methods, significant effort has been devoted to the synthesis of anisotropic metal NPs, with gold nanorods and silver nanoplates (NPTs) being among the most popular [6,7], due to their ability to better confine light below the diffraction limit, which enhances their sensitivity. Silver NPTs in particular show the highest sensitivity to the refractive index of the surrounding medium among metal NPs [7][8][9][10][11].
Ag NPT synthesis often occurs in two steps, since the first step is required to produce spherical NPs, while the second step transforms them into flat triangular nanoplates. This is necessary in order to separate the nucleation of new particles from their growth and shape transformation. The most widely used process so far is the so-called seed-mediated growth, where spherical particles are first produced by chemical reduction of Ag ions by NaBH 4 , then transformed chemically by hydrazine and citrate [7,12,13]. Other methods were also proposed in place of the first or second step of the seed-mediated growth or both. For example, chemically produced NPs can be transformed into NPTs through the use of light irradiation or H 2 O 2 or a combination of both [14][15][16][17][18][19][20][21][22][23]. More recently, Ag NPTs were produced from laser-ablated Ag NPs using light irradiation and hydrogen peroxide (H 2 O 2 ), fully avoiding the chemistry involved in the seed-mediated growth [10,11,[23][24][25]. A few examples of a single-step synthesis of anisotropic Ag NPs have been reported [26][27][28][29][30]. However, most of the processes reported are very time consuming (from 12 h to 90 h). In this paper, we report on the synthesis of Ag NPTs in a single step, combining the laser ablation process with light irradiation and H 2 O 2 , a process that is completed within less less than one hour. To our knowledge, this is the first time Ag NPTs have been obtained by a single laser ablation step. We also investigated the plasmon sensitivity of such nanoparticles under a varying refractive index.

Materials and Methods
Ag NPT synthesis took place according to the scheme shown in Figure 1. A quartz optically transparent reaction vessel was filled with a 1 mM solution of trisodium citrate (TSC). This gave a pH in the region of 8.5-9. A silver target was immersed in the vessel. The beam of a Nd:YAG pulsed laser (Quanta System Spa, Varese, Italy) at 1064 nm (fluence 0.6-1 J/cm 2 , 5 ns pulse duration, 10 Hz repetition rate) was focused by a lens onto the silver target and produces Ag NPs by laser ablation. During the laser ablation process, a white light LED (Ekoo IP66, 100 W; see the emission spectrum in Figure S1) illuminated the reaction vessel, and H2O2 flowed into the solution as regulated by a peristaltic pump at a flow of 69 μL/min. The overall setup is schematized in Figure 1. Small aliquots of the solution were taken at regular intervals to monitor the evolution of the reaction by UV-Vis spectroscopy (Agilent Cary 60 spectrometer, Santa Clara, CA, USA). The morphology of the obtained Ag NPTs was characterized by scanning electron microscopy (SEM) using a Zeiss SUPRA 55-VP (Carl Zeiss Microscoy, Oberkochen, Germany) system and atomic force microscopy (AFM) using a Witec Alpha 300 RS system (WITec, Ulm, Germany). For the SEM and AFM analysis, the Ag NPTs were deposited onto a silane-functionalized Si substrate immediately after synthesis. SEM image analysis was performed using the software ImageJ (Author: Wayne Rasband, National Institute of Mental Health, Bethesda, MD, USA). Simulations were performed using the commercial COMSOL Multiphysics package (from COMSOL Inc., Stockholm, Sweden) in the frequency domain. The simulated spectra were obtained using radiation perpendicular to the flat/larger side of the NPT, with polarization parallel to the same side.
For plasmon sensitivity measurements, a solution obtained by the described process (15 mL) was initially centrifuged for 20 min at 10 K rpm. After the supernatant was removed, the deposit was redissolved in 1 mL of water and homogenized by very short ultrasonication (1 min). Then, 100 μL of the solution was added to 3 mL of water for a refractive index of 1.333 and to 3 mL of sucrose solution (22%, 40%, and 50% for a 1.367, 1.399, and 1.420 refractive index, respectively [31]), and the absorption spectrum was then measured. Figure 2 shows the UV-Vis absorption spectra of a solution that was exposed to our process for 30 min at 69 μL/min H2O2 flow, as well as a solution that was produced by laser ablation only. The latter showed a single sharp feature at 395 nm arising from the surface plasmon resonance of spherical Ag NPs. On the other hand, the spectrum from The morphology of the obtained Ag NPTs was characterized by scanning electron microscopy (SEM) using a Zeiss SUPRA 55-VP (Carl Zeiss Microscoy, Oberkochen, Germany) system and atomic force microscopy (AFM) using a Witec Alpha 300 RS system (WITec, Ulm, Germany). For the SEM and AFM analysis, the Ag NPTs were deposited onto a silane-functionalized Si substrate immediately after synthesis. SEM image analysis was performed using the software ImageJ (Author: Wayne Rasband, National Institute of Mental Health, Bethesda, MD, USA). Simulations were performed using the commercial COMSOL Multiphysics package (from COMSOL Inc., Stockholm, Sweden) in the frequency domain. The simulated spectra were obtained using radiation perpendicular to the flat/larger side of the NPT, with polarization parallel to the same side.

Results and Discussion
For plasmon sensitivity measurements, a solution obtained by the described process (15 mL) was initially centrifuged for 20 min at 10 K rpm. After the supernatant was removed, the deposit was redissolved in 1 mL of water and homogenized by very short ultrasonication (1 min). Then, 100 µL of the solution was added to 3 mL of water for a refractive index of 1.333 and to 3 mL of sucrose solution (22%, 40%, and 50% for a 1.367, 1.399, and 1.420 refractive index, respectively [31]), and the absorption spectrum was then measured. Figure 2 shows the UV-Vis absorption spectra of a solution that was exposed to our process for 30 min at 69 µL/min H 2 O 2 flow, as well as a solution that was produced by laser ablation only. The latter showed a single sharp feature at 395 nm arising from the surface plasmon resonance of spherical Ag NPs. On the other hand, the spectrum from the solution exposed to irradiation and H 2 O 2 flow during the laser ablation process appeared significantly different, as the main feature was red-shifted from that of spherical NPs, and additional features arose, including a low intensity one at 335 nm. Indeed, this is strong evidence that the Ag NPs in solution in this case were far from spherical and suggests that the nanoplates were formed during the process. The spectrum did not show significant variations by changing the H 2 O 2 flow rate between 23 µL/min and 69 µL/min, while increasing the flow rate above 100 µL/min caused complete oxidation of the formed material, and the corresponding absorption spectrum would appear flat. Similarly, decreasing the TSC concentration to 0.1 mM did not change the absorption spectrum significantly, while increasing it up to 10 mM caused complete oxidation of the formed material (flat absorption spectrum). The main feature around 500 nm can be attributed to the in-plane dipole mode, while the peak at 335 nm can be attributed to the out-of-plane quadrupole mode, which is typical of Ag anisotropic structures [14,32]. The other feature around 400 nm may be attributed to larger spherical NPs and/or to the in-plane quadrupole and out-of-plane dipole modes for NPTs or both. We also performed control experiments using either light irradiation or H 2 O 2 flow only. When only irradiation was used, without introducing any H 2 O 2 , the process yielded only spherical NPs, with an absorption spectrum close to that shown in Figure 2. When only H 2 O 2 was introduced in the solution, without light irradiation, the transformation happened only to some extent (see Figure S2). We also performed a further control experiment, using no citrate in the solution. In this case, we observed an absorption spectrum with a single broad and asymmetrical feature ( Figure S3), suggesting that only a broad distribution of spherical particles was formed. Such results indicate that citrate, light irradiation, and H 2 O 2 are all essential ingredients that cooperate in a successful process, as already shown for a similar two-step process [11].

Results and Discussion
Appl. Sci. 2021, 11, x FOR PEER REVIEW 3 of 9 the solution exposed to irradiation and H2O2 flow during the laser ablation process appeared significantly different, as the main feature was red-shifted from that of spherical NPs, and additional features arose, including a low intensity one at 335 nm. Indeed, this is strong evidence that the Ag NPs in solution in this case were far from spherical and suggests that the nanoplates were formed during the process. The spectrum did not show significant variations by changing the H2O2 flow rate between 23 μL/min and 69 μL/min, while increasing the flow rate above 100 μL/min caused complete oxidation of the formed material, and the corresponding absorption spectrum would appear flat. Similarly, decreasing the TSC concentration to 0.1 mM did not change the absorption spectrum significantly, while increasing it up to 10 mM caused complete oxidation of the formed material (flat absorption spectrum). The main feature around 500 nm can be attributed to the inplane dipole mode, while the peak at 335 nm can be attributed to the out-of-plane quadrupole mode, which is typical of Ag anisotropic structures [14,32]. The other feature around 400 nm may be attributed to larger spherical NPs and/or to the in-plane quadrupole and out-of-plane dipole modes for NPTs or both. We also performed control experiments using either light irradiation or H2O2 flow only. When only irradiation was used, without introducing any H2O2, the process yielded only spherical NPs, with an absorption spectrum close to that shown in Figure 2. When only H2O2 was introduced in the solution, without light irradiation, the transformation happened only to some extent (see Figure  S2). We also performed a further control experiment, using no citrate in the solution. In this case, we observed an absorption spectrum with a single broad and asymmetrical feature ( Figure S3), suggesting that only a broad distribution of spherical particles was formed. Such results indicate that citrate, light irradiation, and H2O2 are all essential ingredients that cooperate in a successful process, as already shown for a similar two-step process [11].  Figure 3a,b shows representative SEM images of Ag NPTs produced by our process. We can observe Ag NPTs showing a mixture of shapes, including triangles, hexagons, circles, and other irregular ones. The size distribution shows that most NPTs were 70-90 nm in size (Figure 3c). Figure 4 shows an AFM image and the profiles obtained from the same samples analyzed by SEM. Figure 4a also shows NPTs with mixed shapes, and Figure 4b shows the thickness ranging between 10 nm and 25 nm.  We can observe Ag NPTs showing a mixture of shapes, including triangles, hexagons, circles, and other irregular ones. The size distribution shows that most NPTs were 70-90 nm in size (Figure 3c). Figure 4 shows an AFM image and the profiles obtained from the same samples analyzed by SEM. Figure 4a also shows NPTs with mixed shapes, and Figure 4b shows the thickness ranging between 10 nm and 25 nm.
The growth mechanism in this process can be elucidated within the framework of the reactions taking place among citrate, Ag, and H 2 O 2 in the presence of light. It is known that, in the presence of light, citrate is able to reduce Ag + to Ag 0 [33]: Ag + + e − → Ag 0 At the same time, H 2 O 2 can both oxidize and reduce silver [17,18]: Appl. Sci. 2021, 11, x FOR PEER REVIEW 4 of 9 The growth mechanism in this process can be elucidated within the framework of the reactions taking place among citrate, Ag, and H2O2 in the presence of light. It is known that, in the presence of light, citrate is able to reduce Ag + to Ag 0 [33]: At the same time, H2O2 can both oxidize and reduce silver [17,18]: H2O2 + 2Ag + + 2OH − → 2Ag + 2H2O + O2 (4)  Besides reducing Ag in the presence of light, citrate also has the function of being a capping agent for Ag NPs, as it easily binds to the (111) facets of silver [34,35]. This allows both NP stabilization in solution and their preferential growth along the Ag (100) direction, which is less preferred by citrate. Hence, Ag NPs initially produced by the laser ablation process were first oxidized by H 2 O 2 , being partially dissolved in the liquid medium, and then reduced back by citrate, supported by H 2 O 2 , adding preferentially to the uncapped (100) Ag facet. It is interesting to note that the absorption spectrum shown in Figure 2 did not dramatically change its shape after about 10 min of this process, while the intensity increased and then stabilized over time ( Figure S4). This suggests that an equilibrium was reached within the solution between the production of NPs by laser ablation and their conversion into NPT and that after reaching this equilibrium, the process only increased the NPT concentration in the solution.
Besides reducing Ag in the presence of light, citrate also has the function of being a capping agent for Ag NPs, as it easily binds to the (111) facets of silver [34,35]. This allows both NP stabilization in solution and their preferential growth along the Ag (100) direction, which is less preferred by citrate. Hence, Ag NPs initially produced by the laser ablation process were first oxidized by H2O2, being partially dissolved in the liquid medium, and then reduced back by citrate, supported by H2O2, adding preferentially to the uncapped (100) Ag facet. It is interesting to note that the absorption spectrum shown in Figure 2 did not dramatically change its shape after about 10 min of this process, while the intensity increased and then stabilized over time ( Figure S4). This suggests that an equilibrium was reached within the solution between the production of NPs by laser ablation and their conversion into NPT and that after reaching this equilibrium, the process only increased the NPT concentration in the solution. Simulations of the extinction spectra were performed to support and complement the experimental results. Figure 5 shows simulated extinction spectra for NPTs of a round and hexagonal shape, with a size of 70 nm and 82.5 nm and a thickness of 10 and 20 nm, representative of the structures observed by SEM and AFM. Figure 6 shows the experimental absorption spectrum, already shown in Figure 2, with the simulated ones. It can be observed that the simulated spectra were mainly in agreement with the longer wavelength part of the experimental spectrum. This partial disagreement might be due to the fact that the simulated spectra were obtained for isolated NPTs, thus not taking into account the interaction among the particles. However, this aspect is still unclear and will be subject to further investigation.
The obtained NPTs were finally tested for plasmon sensitivity (S), i.e., the variation of the plasmon resonance peak position due to a change in the refractive index (S = Δλ/Δn), which is typically reported in terms of nm/RIU (refractive index unit). To change the refractive index, we used sucrose solutions at set concentrations, since the relation between sucrose concentration in water and the refractive index is well established [31]. Figure 7 shows the plasmon resonance peak position of a Ag NPT solution produced by our process by varying the refractive index, along with the data simulated for disks of 70 nm in size and 20 nm in thickness. Fitting the data yielded a plasmon sensitivity S of 216 nm/RIU, close to that obtained by the simulations. This value is comparable to the others, typically below 250 RIU, reported in the literature for Ag NPTs with a similar plasmon resonance [7][8][9][10][11], with the advantage of obtaining the nanostructures in a fast and simple one-step process, despite the poor SPR tunability. If we compare the experimental data with those obtained from the simulation of discs and hexagons (70 nm × 20 nm), we observe that the simulated data showed a higher plasmon sensitivity. This was probably due to a higher SPR wavelength in the simulated spectra, which was expected to exhibit higher plasmon sensitivity, in agreement with our previous work [36]. Simulations of the extinction spectra were performed to support and complement the experimental results. Figure 5 shows simulated extinction spectra for NPTs of a round and hexagonal shape, with a size of 70 nm and 82.5 nm and a thickness of 10 and 20 nm, representative of the structures observed by SEM and AFM. Figure 6 shows the experimental absorption spectrum, already shown in Figure 2, with the simulated ones. It can be observed that the simulated spectra were mainly in agreement with the longer wavelength part of the experimental spectrum. This partial disagreement might be due to the fact that the simulated spectra were obtained for isolated NPTs, thus not taking into account the interaction among the particles. However, this aspect is still unclear and will be subject to further investigation.
The obtained NPTs were finally tested for plasmon sensitivity (S), i.e., the variation of the plasmon resonance peak position due to a change in the refractive index (S = ∆λ/ ∆n), which is typically reported in terms of nm/RIU (refractive index unit). To change the refractive index, we used sucrose solutions at set concentrations, since the relation between sucrose concentration in water and the refractive index is well established [31]. Figure 7 shows the plasmon resonance peak position of a Ag NPT solution produced by our process by varying the refractive index, along with the data simulated for disks of 70 nm in size and 20 nm in thickness. Fitting the data yielded a plasmon sensitivity S of 216 nm/RIU, close to that obtained by the simulations. This value is comparable to the others, typically below 250 RIU, reported in the literature for Ag NPTs with a similar plasmon resonance [7][8][9][10][11], with the advantage of obtaining the nanostructures in a fast and simple one-step process, despite the poor SPR tunability. If we compare the experimental data with those obtained from the simulation of discs and hexagons (70 nm × 20 nm), we observe that the simulated data showed a higher plasmon sensitivity. This was probably due to a higher SPR wavelength in the simulated spectra, which was expected to exhibit higher plasmon sensitivity, in agreement with our previous work [36]. Appl. Sci. 2021, 11, x FOR PEER REVIEW 6 of 9

Conclusions
We demonstrated the fast and facile synthesis of Ag NPTs in a single step by combining laser ablation in liquids with light irradiation in the presence of H2O2. We obtained flat NPTs of mixed shapes with sizes between 60 and 80 nm and thicknesses of 10-20 nm. The SPR resonance of these nanostructures fell around 500 nm. While at this stage, the process seems to be efficient in a narrow range of parameters, with poor SPR tuning capability, we foresee that exploring a wider range of conditions, such as laser ablation parameters or the chemical composition of the solution, might provide some SPR tuning capability to some extent. Changing the refractive index of the surrounding medium yielded a plasmon sensitivity of 216 nm/RIU.

Supplementary Materials:
The following are available online at www.mdpi.com/xxx/s1, Figure S1: Emission spectrum of the LED used for irradiation during the laser ablation process. Figure S2: Absorption spectrum from a solution obtained when irradiation was taken out of the described process, Figure S3: Absorption spectrum from a solution when citrate was taken out of the described process, Figure S4: Absorption spectra taken at different times during the described process.  Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data presented in this paper are available upon request from the corresponding author.