Xiaobin
Xie
*,
Marijn A.
van Huis
and
Alfons
van Blaaderen
*
Soft Condensed Matter, Debye Institute for Nanomaterials Science, Utrecht University, Princetonplein 5, 3584 CC Utrecht, The Netherlands. E-mail: xb.xie@hotmail.com; A.vanBlaaderen@uu.nl
First published on 12th January 2021
The surface plasmon resonance of noble metals can be tuned by morphology and composition, offering interesting opportunities for applications in biomedicine, optoelectronics, photocatalysis, photovoltaics, and sensing. Here, we present the results of the symmetrical and asymmetrical overgrowth of metals (Ag, Pd, and Pt) onto triangular Au nanoplates using L-ascorbic acid (AA) and/or salicylic acid (SA) as reductants. By varying the reaction conditions, various types of Au nanotriangle-metal (Au NT-M) hetero-nanostructures were easily prepared. The plasmonic properties of as-synthesized nanoparticles were investigated by a combination of optical absorbance measurements and Finite-Difference Time-Domain (FDTD) simulations. We show that specific use of these reductants enables controlled growth of different metals on Au NTs, yielding different morphologies and allowing manipulation and tuning of the plasmonic properties of bimetallic Au NT-M (Ag, Pd, and Pt) structures.
Great progress was made in the wet chemical synthesis of BNMNPs for realizing applications utilizing the plasmonic properties.8 Most of these reports focused on the synthesis of regular core–shell NPs. Recently, the plasmonic properties of asymmetrically composed Au based bi- and higher-metallic type NPs have been the focus of much attention.8,21 For instance, Liang et al.22 reported that control can be obtained over the symmetry of hybrid Au nanorods (NRs) for achieving tunable optical properties. A ligand-mediated selective deposition of noble metals at Au nanoplates and NRs was developed by the Millston group.23 In another paper on Pt-decorated Au prisms it was shown that the Au plasmon modes are changed by the location of the Pt particle, which influences the energy transfer between Pt and the Au prisms.24 Although the rapid development of this field provides us with a multitude of methods for synthesis and with insights into the relationship between the symmetry and optical property of BNMBPs, more studies are required to achieve a better understanding of the (asymmetric) growth of a second metal.
In many aqueous phase syntheses of BNMNPs, L-ascorbic acid (AA) is used as reductant because of several reasons. First, its moderate reaction activity for various noble metals, in particularly for Au, Ag, Pd, and Pt precursors at room temperature, makes it suitable for the controlled synthesis of various metallic nanomaterials. Moreover, the reaction rate between AA and noble metal precursors can be easily manipulated by changing the reaction temperature, the reactants’ concentration, and the pH value of the growth solution. Finally, it is a low priced and readily available chemical. Nevertheless, its reaction rate for reducing some precursors, like AgNO3, is still relatively fast which makes it less easy to control the reaction. In this case, salicylic acid (SA) also is an easily available chemical like AA, but the reaction rate between SA and noble metal precursors is even slower than AA at similar reaction conditions. Besides, SA has been reported as additional chemical that works with AA together enabling a morphology-controlled synthesis of Au nanorods (Au NRs)25 and Au NRs-Pd bimetallic NPs.26 The use of SA as the only reductant for controlling the morphology of bimetallic NPs has thus far hardly been reported.
In the present work, we systematically investigate the roles that AA and SA play in the synthesis of Au NTs based BNMNPs under similar reaction conditions. As expected, the slower reaction rate between SA and the precursors of the metals used has a strong impact on the final shapes of the products. Second, SA showed a fascinating ability for precisely controlling the atomic layers of Pd and sizes of Pt clusters grown onto the Au NTs. Using AA and/or SA also allowed manipulating the symmetrical and/or asymmetrical growth of noble metals onto Au NTs. The various morphologies that were experimentally obtained for the Au–M heterogeneous nanocrystals are schematically shown in Scheme 1 and will be discussed in detail below. Furthermore, the surface plasmonic properties of all as-prepared Au NTs based BNMNPs were characterized in detail by combining far-field UV-Vis spectroscopy and finite-difference time-domain (FDTD) simulations. In Scheme 1 we schematically describe the differences as observed by use of the two reductants in the bi-metallic overgrowths on Au NTs that were compared in this study and how they differently affected particle morphologies. This scheme further acts as guide for the structure of the paper and acts also as guide how the differences in the morphologies affected experimentally determined and simulated plasmonic responses of the bi-metallic particles made.
Scheme 1 Schematic diagram showing the differences between Au NT@M (Ag, Pd, and Pt) nanostructures synthesized using either AA or SA. |
Fig. 1 Scanning transmission electron microscopy (STEM) images (a, b, & c) of Au nanotriangles (Au NTs); (d) histogram of edge length of Au NTs; (e) UV-Vis spectrum of Au NTs. |
Fig. S1† shows that changing the reaction temperature between 20 and 60 °C hardly had any influence on the final shape in both the AA and the SA case (Fig. S2†). With a decreasing concentration of AgNO3 used for the synthesis, the Ag shell of Au NT-Ag NPs became thinner for those experiments using AA (Fig. S3†), while the Ag shell thickness remained the same when using SA (Fig. S4†), due to the secondary nucleation of Ag at high concentration of AgNO3. The above results indicate that AA is more suitable for the morphology-controlled synthesis of Au NT-Ag bimetallic NPs.
The plasmonic properties of the Au NT-Ag NPs were investigated by measuring extinction spectra of as-synthesized BNMNPs with a UV-Vis spectrophotometer in combination with finite-difference time-domain (FDTD) simulations of the extinction cross-sections of the Au–M NPs. As shown in Fig. 3a, the main LSPR band of both the a-Au NT-Ag NPs (λLSPR = 545 nm) and s-Au NT-Ag NPs (λLSPR = 620 nm) shows a blue-shift compared to the LSPR of Au NTs (λLSPR = 660 nm). To reveal the plasmonic properties of the Au NT-Ag NPs with different types of Ag shells, we measured the LSPR of Au NT-Ag NPs with various thicknesses of Ag shells. Fig. S5a† indicates that the LSPR band of a-Au NT-Ag NPs emerged as a blue shift in the range of 50 to 115 nm when the thickness of the Ag shell was tuned from 10 to 20 nm. To further clarify the relationship between the blueshifts of the LSPR and the thickness of the Ag shell, a series of FDTD simulations were carried out. The FDTD simulations on the Au NT-Ag NPs with an asymmetrical Ag shell with thicknesses varying from 1 nm to 40 nm (Fig. 3b), show good agreement in that a thicker Ag shell leads to a larger blue shift of the main LSPR band. We also performed FDTD simulations on Au NT-Ag NPs with a symmetrical Ag shell with thicknesses from 1 nm to 10 nm. Their LSPR bands, like in the asymmetrical Ag shell cases, show a similar blue shift with increasing thickness of the Ag shell (Fig. 3c). In agreement with this finding, the similar LSPR bands of Au NT-Ag NPs synthesized at different temperatures also confirm that changing the reaction temperature from 60 to 20 °C hardly influenced the morphology of the Au NT-Ag NPs (Fig. S5c & d†).
The thickness of the Pd shell was tuned between 10 nm and 2 nm by changing the concentration of Na2PdCl4 or Au NTs when AA was used for the growth (Fig. S6 & S7†). Similar experiments were conducted using SA replacing AA, upon which the Pd shell became more uneven as the concentration of Na2PdCl4 used for the reaction from 150 μM to 50 μM. This change in morphologies is revealed by the STEM images and STEM-EDS maps shown in Fig. 5. The Pd shell became thinner at the edges, top, and bottom face of the Au NTs, but still had a similar size on part of the vertices.
Fig. 5 Morphology evolution of Au NT-Pd NPs synthesized using SA as reductant when changing the concentration of Na2PdCl4. |
When changing the reaction temperature from 60 to 20 °C, the Pd shells showed no difference in both shape and thickness when AA was used (Fig. S8†). On the contrary, SA was used as reductant, the Pd shell became thinner and more even as the temperature decreased (Fig. S9, S10, & S11†). Besides, as shown in Fig. S12,† polycrystalline Pd was also found at a reaction temperature of 20 °C. The main reason is that the reaction rate between SA and Na2PdCl4 slows down as the temperature decreases.
Fig. 6a shows the extinction spectra of the a-Au NT-Pd NPs and s-Au NT-Pd NPs presented in Fig. 4. Compared to the Au NTs, the main LSPR band of both of the Au NT-Pd NPs became broader and showed a slight blue-shift. To clarify how the Pd thickness influenced the plasmon peaks, we further measured the extinction spectra of Au NT-Pd with a shell thicknesses from 2 nm to 10 nm. As shown in Fig. 6b and Fig. S13a,† the main LSPR band of the Au NT-Pd NPs became broader, and the intensity became weaker when the Pd shells were thicker. The same measurements were performed on the a-Au NT-Pd NPs, and a similar trend was found (Fig. S13b†), which means that the LSPR band of the Au NT-Pd NPs mainly depends on the thickness of Pd shell. Moreover, the FDTD simulations of Au NTs-Pd with a thickness of the Pd shell ranging from 1 nm to 5 nm show good agreement as the plasmon peak decayed as the thickness of Pd shell increases (Fig. 6c). To simulate the extinction spectrum of a-Au NT-Pd NPs, we created a model shown as the inserted model in Fig. 6d. Comparing such a rough particle to particles with the same Pd thickness, the LSPR band of a-Au NT-Pd is wider and shows a prominent tail at the high wavelength side of the peak (Fig. 6d). These results show good agreement with the experimental measurements (Fig. S13b†).
To further elucidate how the reaction conditions influence the shape of the Au NT-Pt NPs, two reaction parameters were analyzed: the concentration of Pt precursor (K2PtCl4) and reaction temperature. In agreement with our expectation, both sizes of the Pt irregular dendritic structures and Pt islands on the Au NT reduced when the concentration of K2PtCl4 for the reaction was decreased from 150 μM to 50 μM (Fig. 8 and Fig. S14†). Next, we performed syntheses by keeping the concentration of solution and reaction time constant but changed the temperature from 60 °C to 20 °C. The size and density of the Pt irregular dendritic structures and Pt islands dramatically decreased upon lowering the reaction temperature from 60 °C to 40 °C and 20 °C in both cases of using AA and SA (Fig. S15, S16, & S17†). There are two possible reasons for this phenomenon. First, the reaction rate slows down as the temperature decreases. Second, the solubility of the [PtCl4]2−-cetyltrimethyl-ammonium ion ([PtCl4]2−CTA+2) complex decreased upon lowering the temperature. It was reported that the formation of a [PtCl4]2−CTA+2 complex impedes the reduction of the Pt precursor,27 which is confirmed as well by our observation that the transparent solution became turbid at 20 °C.
Fig. 8 Morphology evolution of Au NT-Pt NPs synthesized using AA as reductant when changing the concentration of K2PtCl4. |
Fig. 9 shows the extinction spectra of the Au NT-Pt NPs presented in Fig. 7 and the starting Au NTs. The LSPR of the Au NT-Pt NPs broadened and there was a slight blue-shift for the a-Au NT-Pt NPs and a red-shift for the s-Au NT-Pt NPs. The blue-shift was possibly due to a mild etching of the Au NTs during the reaction which is visible in the (S)TEM images in Fig. 7, and the slight red-shift is because of the growth of the Pt islands onto the whole Au NT. The finding of the LSRP bands of the Au NT-Pt NPs with smaller Pt irregular dendritic structures and/or Pt islands supports this explanation. From Fig. 8, it becomes clear that the etching of the Au NTs substrates was almost the same, but the Pt irregular dendritic structures became smaller with decreasing the concentration of K2PtCl4. As shown in Fig. S18a,† the LSPR band position of those particles with larger Pt IDSs was closer to the band position of the original Au NTs. In contrast, growth of smaller Pt islands on Au NTs reduced the redshifts of the LSPR peaks (Fig. S18b†). The damping of Au NT LSPR after Pt deposition is dependent on the Pt density and thickness on Au NTs, where an increase of the Pt density and thickness both lead to broadening of the extinction spectra (Fig. S18c & d†). The damping effect caused by Pt growth is due to the dielectric effects and/or charge transfer between the Au and Pt.
The overall formation process of the nanostructures is governed by three consecutive steps; (1) reactions between reductants and metal precursors, (2) initial deposition of metal atoms and clusters onto the Au surfaces, and (3) kinetic factors determining the final morphology, which we are discussing below in this order.
2Ag+ + C6H8O6 → 2Ag + C6H6O6 + 2H+ |
2Ag+ + C7H6O3 → 2Ag + C7H4O3 + 2H+ |
The total number of hydroxyl groups that serve as the reacting group and take part in the reactions with metal precursors in AA is four, which is larger than that in SA where it is one (Scheme S1†). In addition, due to the ortho-effect of multiple hydroxyl groups in AA and the inactivation of carboxyl in SA, the reactivity of the hydroxyls of AA is higher than the reactivity of the same group in SA. This causes the reaction rates between AA and the metals precursors to be higher than those between SA and the precursors. The concentration of AA and SA used for most of the experiments was 1000 μM and 1800 μM, respectively. Both concentrations were much higher than the concentrations of metals precursors. A set of experiments were conducted using 900 μM SA for synthesising Au NT-M (Ag, Pd, Pt). Compared to the cases of using 1800 μM SA, the morphologies of the resulting nanoparticles showed no differences (Fig. S19†), which implies that the shapes of the Au NT-M (Ag, Pd, Pt) bi-metallic particles would not be affected by changing the concentration of SA in the range between 900 μM to 1800 μM.
Δγn = γSn + γin − γCn |
In the metallic core–shell NPs system, Δγn is the difference in surface energy between n atomic layers of shell and the core metal. γSn, γin, and γCn are the surface free energies of the shell, core–shell interface, and the core metal, respectively. The F–M growth mode is in the regime where: (i) Δγn ≤ 0, (ii) γSn < γCn, and (iii) γin is small. If (ii) and/or (iii) are not fulfilled, then S–K mode directs the growth. Lastly, the V–W mode dominates when Δγn > 0.31
Considering the results of our experiments, those cases where the second metal grew continuous smooth shells on the Au NT cores, like in the case of Ag, can be characterized as F–M growth, while those cases where discontinuous shells, e.g. in the case of Pt, were found, can be considered to have grown via the V–W mode. For the case of Pd, both kinds of shells were observed and have been reported before.16,29,32,33 From the point of view of thermodynamics, the bonding energy between Pd growing onto Pd is lower than that between Au and Pd, which allows continuous Pd shell growth onto the Au core. On the other hand, a discontinuous shell can also form via manipulating kinetic factors, e.g. reductants and/or Pd precursors.
As shown in Fig. 6, the density of Pt irregular dendritic structures was lower than the Pt islands on the Au NTs. The possible reason is that the size of the SA molecule is smaller and the molecular structure is planar, which makes it easier to pass the CTAC double layers and reach the Au NT surface.26 Moreover, the concentration of SA used for the experiments was much higher than the concentration of the metal precursors, and SA played a role as co-surfactant in the metals growth as well. The role of co-surfactant of SA was also reported previously in the synthesis of Au nanorods.25,34 For the larger AA molecule, it is more difficult to pass the CTAC double layers, and thus, the number of growth sites on Au NT surface was smaller. It was also observed in previous reports,23,35,36 that the Pt islands grew into patterns on the Au NTs due to the adsorption of CTAC molecules on the Au NT surface and the geometry of the Au NTs.
Furthermore, temperature plays an important role in controlling the morphology of Au NT-M (Pd, Pt) nanoparticles. For instance, it can affect the reaction rate of AA, SA, and metal precursors. In this respect, SA was affected more than AA and resulted in the shape of both Au NT-Pd and Au NT-Pt being significant changed when the temperature was decreased to 40 °C and 20 °C (Fig. S9–12 and S17†). Second, in the case of Pt growth, temperature could affect the solubility of the [PtCl4]2−CTA+2 complex and then influence the reduction of the Pt precursor, which was found when AA and/or SA were used (Fig. S15 & S17†). Overall, temperature is an essential factor to take into account in the morphology-controlled synthesis of Au NT-M, in particular when using SA as reductant.
Overall, the deposition of Ag, Pd, and Pt onto Au NTs was initially affected by thermodynamical factors yielding different growth modes, in which Ag follows F–M growth, while Pd and Pt grow via V–W mode. Subsequently, kinetical factors influenced the further growth and the final morphology of the Au NT-M nanoparticles. In this regard, the reactivity of AA and SA, which is determined by their molecular structure, influenced the symmetric and/or asymmetric growth of metals onto Au NTs, where the reactivity of SA is lower than that of AA, resulting in the reactions of SA with metals precursors to be more sensitive to the reaction temperature. Although SA showed less control over the shape of Au NT-Ag, by using it, asymmetric Au NT-Pd and Au NT-Pt nanostructures with a nice patterning on the Au NTs surfaces were obtained, which showcases its great potential for synthesizing metals nanocrystals.
More details of experimental section, e.g. Chemicals, Characterization, and FDTD simulations, can be found in ESI.†
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d0nr06789j |
This journal is © The Royal Society of Chemistry 2021 |