Letter pubs.acs.org/NanoLett
Control of Electron Beam-Induced Au Nanocrystal Growth Kinetics through Solution Chemistry Jeung Hun Park,†,‡ Nicholas M. Schneider,§ Joseph M. Grogan,§ Mark C. Reuter,‡ Haim H. Bau,§ Suneel Kodambaka,*,† and Frances M. Ross*,‡ †
Department of Materials Science and Engineering, University of California Los Angeles, 410 Westwood Plaza, Los Angeles, California 90095, United States ‡ IBM T. J. Watson Research Center, 1101 Kitchawan Road, Yorktown Heights, New York 10598, United States § Department of Mechanical Engineering and Applied Mechanics, University of Pennsylvania, 220 South 33rd Street, Philadelphia, Pennsylvania 19147, United States S Supporting Information *
ABSTRACT: Measurements of solution-phase crystal growth provide mechanistic information that is helpful in designing and synthesizing nanostructures. Here, we examine the model system of individual Au nanocrystal formation within a defined liquid geometry during electron beam irradiation of gold chloride solution, where radiolytically formed hydrated electrons reduce Au ions to solid Au. By selecting conditions that favor the growth of well-faceted Au nanoprisms, we measure growth rates of individual crystals. The volume of each crystal increases linearly with irradiation time at a rate unaffected by its shape or proximity to neighboring crystals, implying a growth process that is controlled by the arrival of atoms from solution. Furthermore, growth requires a threshold dose rate, suggesting competition between reduction and oxidation processes in the solution. Above this threshold, the growth rate follows a power law with dose rate. To explain the observed dose rate dependence, we demonstrate that a reaction-diffusion model is required that explicitly accounts for the species H+ and Cl−. The model highlights the necessity of considering all species present when interpreting kinetic data obtained from beam-induced processes, and suggest conditions under which growth rates can be controlled with higher precision. KEYWORDS: liquid cell TEM, Au nanocrystal growth, radiolysis, electron beam induced growth, nanoprisms, reaction-diffusion model
D
connection between the nucleation of a twinned nanocrystal and the subsequent evolution into a nanoprism has been established,13,14 but the subsequent growth kinetics of the plates has not been quantified at an individual level. To obtain quantitative information on nanoprism growth, we use liquid cell transmission electron microscopy (TEM), a technique with unique combination of temporal and spatial resolution.15,16 The electron beam precipitates metal particles from a solution of the metal ions via radiolysis,17,18 analogous to nanocrystal growth that is stimulated with other energetic beams such as γ-radiation.19 A good understanding of the interactions between the electrons and the irradiated medium is
uring liquid phase growth of nanocrystals, size and morphology control are essential to tailor the physical and chemical properties of the resulting structures.1,2 Nanoprisms are a morphology of particular interest1,3,4 due to their applications in areas5 including catalysis, electronics, encryption strategies, molecular diagnostics, and gene therapy. Plate-like crystals such as prisms have been synthesized in the solution phase.4,6,7 For zero- (0D) and one-dimensional (1D) nanomaterials (quantum dots and nanorods), solution-based synthesis is usually explained and rationally designed by growth mechanisms8 such as solution-liquid−solid growth,9 screw dislocation-driven growth,10 or anisotropy-induced crystal growth.11 However, solution-grown nanoprisms or other flat “quasi-two-dimensional” (2D) crystals often grow as part of an ensemble of different shape and size particles.12,13 This makes it harder to quantify their growth rates and mechanisms. The © XXXX American Chemical Society
Received: April 28, 2015 Revised: July 18, 2015
A
DOI: 10.1021/acs.nanolett.5b01677 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 1. (A) Representative bright field TEM images of Au crystal growth in 20 mM HAuCl4, acquired at times t indicated after irradiation began at t0. Areal dose rate 160 e−/Å2·s (beam current 8.0 nA spread over 2 μm diameter). All the images are obtained at the same magnification so the scale bar applies to all images. Movie M1 shows the complete data set. (B) Measured area S vs (t − t0) for the crystal in (A). (C) Image intensity vs (t − t0) showing that the intensities transmitted through the crystal and the liquid do not change during growth. (D) Atomic force micrograph of a different crystal, obtained from an air-dried sample taken out of a Hummingbird cell after growth at 42 e−/Å2·s (beam current 2.1 nA spread over 2 μm diameter).
necessary to interpret the processes occurring and to extract mechanistic information. The irradiating electrons change the solution chemistry by creating transient molecular and radical products, including e−h (the hydrated electron), H•, OH•, H2O2 and H3O+.20 The hydrated electron is a strong reducer of metal ions present in the solution, and enables metal nanocrystals to grow in solution or on the silicon nitride windows of the liquid cell.20 In this Letter, we track the growth kinetics of individual Au nanoprisms during irradiation of gold chloride solution in a well-defined thin liquid geometry. By selecting conditions that favor the growth of faceted Au nanoprisms, we obtain growth rates for crystals as a function of size, location, and irradiation conditions. We discuss these kinetics in the context of a process limited by formation of Au atoms in solution. We then extend
an existing reaction-diffusion radiolysis simulation, adapting it to the case of Cl− ions and examining explicitly the dose rate dependence of the radiolytic species for known parameters and crystal and liquid geometry, so that we can interpret the measurements more quantitatively. We find that the dependence of growth rate on dose rate can be explained only by accounting for the species in solution, specifically Cl− and H+. In addition to providing a more detailed understanding of the growth kinetics, these results also illustrate how the chemistry of a solution changes during radiolysis on the electron beam parameters and the initial solution composition. We suggest that growth at lower pH, for example, provides simpler kinetics that may allow a higher level of prediction and control of morphology. B
DOI: 10.1021/acs.nanolett.5b01677 Nano Lett. XXXX, XXX, XXX−XXX
Letter
Nano Letters
Figure 2. (A) Composite bright field TEM image showing three neighboring Au crystals. The red outlined crystal is located 0.5 μm from the blue outlined crystal. (B) Area S vs time (t − t0) for the three crystals in (A). Growth took place at 42 e−/Å2·s (2.1 nA beam current spread over 2 μm diameter). (C) Plots of S vs (t − t0) data for Au crystals grown at several areal dose rates, obtained by varying the beam current at fixed illumination area. t0 is the time at which the crystals first became visible within the field of view. Solid lines in (B) and (C) are least-squares linear fits to the S vs (t − t0) data. (D) Plot of growth rate dS/dt vs areal dose rate with each data point corresponding to the average growth rate of 2−6 individual Au crystals at a given e-beam current. The error bars show the standard deviation of the dS/dt measurements of all crystals growing at each dose rate. The beam current i varied from 0.1 nA (2 e−/Å2·s) to 8.0 nA (160 e−/Å2·s) over an illuminated diameter of 2 μm. The solid line is the least-squares fit using the equation shown in the plot and yields a growth exponent β ≈ 0.76 and a threshold dose rate of 30 e−/Å2·s (≈ 1.38 nA spread over 2 μm diameter).
The growth experiments were performed in two different liquid cell platforms, the Nanoaquarium21,22 mounted in a custom holder in a Hitachi H9000 TEM, and a Hummingbird flow system23 in an FEI CM30 TEM (see Supporting Information for details). Both microscopes were operated at 300 kV. We used TEM mode, fixing the beam size at typically 2 μm in diameter to irradiate a cylinder of liquid. Images were recorded at 30 frames per second. The solutions contained 20 mM HAuCl4 and were not deaerated before use. In the absence of the electron beam, Au nucleation does not take place, but upon irradiation, faceted Au crystals became visible on the liquid cell windows within 1−2 s (Figure 1). The morphology of these crystals depends on the thickness of the liquid layer, and examples are shown in Supporting Information Figure S1. In regions where the liquid completely fills the cell and is several hundred nanometers thick, we observe growth of spiky almost spherical structures composed of multiple crystals. These are similar to the deposits that have been described elsewhere.13,24−26 However, we focus here on the nanocrystals that form in thinner (
