Supporting Information: Manipulation of the geometry and modulation of the optical response of surfactant-free gold nanostars: a systematic bottom-up synthesis Agampodi S. De Silva Indrasekara 1,2, Sean F Johnson 1, Ren A Odion 1,2, Tuan Vo-Dinh 1,2,3* 1 Department of Biomedical Engineering, Duke University, Durham, NC 2 Fitzpatrick Institute of Photonics, Duke University, Durham, NC 3 Department of Chemistry, Duke University, Durham, NC Figure S1: The outline for the (a) core diameter and (b) tip-to-tip distance calculation for GNS used the automated MatLab algorithm developed in our lab. Note that the estimated core diameter is an overestimation in order to avoid any personal biases from manual calculations. Description of Matlab Code: The script was written to load in specific TEM images of specific resolution and file type (e.g. 2000 x 2000, tif format). The original image was split to two images, one that only contains the size scale and the other that contains only the TEM image of the nanostars. The scalebar image is quickly processed to obtain the pixel length of the scale bar. A calibration is created by dividing this pixel length by the known nanometer length of the scalebar. The nanostar image is processed separately with the watershed algorithm described above. The tip to tip length was calculated by a function that calculates the largest Euclidian distance value between any two points on the perimeter of the nanostar while the core diameter was roughly estimated by a creating a circular mask centered on the nanostar while maintaining the same area. The matlab functions used for this process are: watershed, regionprops, and pdist2. The compiled distances are visualized as a histogram using the hist function on matlab with automatic binning. The effect of orders of reactant addition on reproducible GNS synthesis: A set of control experiments was conducted in which the order of seed addition and growth solution ph was changed, and the resultant GNS was characterized by UV-Vis and TEM to evaluate the reproducibility. The three conditions compared are; Condition A: sequential addition of Au seeds, silver nitrate and ascorbic acid to an acidic mixture of AuCl 4 - with 5 seconds delay between each addition. (The protocol given in the manuscript) 1
Condition B: sequential addition of silver nitrate, Au seeds, and ascorbic acid to an acidic mixture of AuCl 4 - with 5 seconds delay between each addition. Condition C: sequential addition of Au seeds, silver nitrate and ascorbic acid to aqueous AuCl 4 - (No acid added) with 5 seconds delay between each addition. All three conditions resulted GNS, but it was evident that the reproducibility of the LSPR peak position is much greater under the condition A (the protocol given in this manuscript). It was also noticeable that condition C, where no acid was present, relatively smaller GNS was resulted, which could be attributed to the relatively lower reduction rate of Au salt and or low equilibrium concentration of reactive Au 3+ available under less acidic growth solution ph in comparison to condition A and B (ph 3). Therefore, based on the current experimental evidences available, we find that reproducibility of GNS using the current protocol is not significantly affected by partial oxidation of Au seeds (if it does take place). Figure S2: The addition of ascorbic acid followed by silver nitrate to an acidic mixture (ph 3) of gold salt and gold seeds yield most reproducible GNS formation. (a) LSPR peak position and (b) representative TEM micrographs of GNS prepared under three conditions of reactant addition. A: The addition of ascorbic acid followed by silver nitrate to an acidic mixture (ph 3) of gold salt and gold seeds, B: The addition of gold seeds followed by ascorbic acid to an acidic mixture (ph 3) of gold salt and silver nitrate, C: the addition of ascorbic acid followed by silver nitrate to a mixture (ph 6) of gold salt and gold seeds. Scale bars: 100 nm 2
Figure S3: EDX map and quantification of atomic Ag and Au content in the representative single GNS shown in Figure 1. Figure S4: TEM micrographs of GNS and the MatLab program based size analysis show yield of GNS with a higher homogeneity in particle size distribution. (a) TEM micrographs (c) core diameter and (d) tip-to-tip diameter of GNS obtained by changing the amount of d=5nm gold seeds to Au 3+ in the growth solution. 3
Figure S5: TEM micrographs of GNS and the MatLab program based size analysis show yield of GNS with a higher homogeneity in particle size distribution. (a) TEM micrographs (c) core diameter and (d) tip-to-tip diameter of GNS obtained by changing the amount of d=12 nm gold seeds to Au 3+ in the growth solution. Figure S6: The formation of GNS takes place even in the absence of halides. (a) TEM micrograph and (b) absorption spectrum of GNS synthesized at ph 3 (ph adjusted by adding HNO 3 ) but in the absence of any additive Cl -. 4
Figure S7: Representative absorption spectra of GNS synthesized at varying concentrations of (a) Cl - at ph 3, (b) Cl - at ph 6 and Br - at ph 3. Figure S8: GNS shape evolution is drastically affected by the presence of I - in the growth solution. TEM micrographs showing the nanoparticles formed in the presence of (a) 10 μm (b) 1 μm (c) 100 nm, and (d) 10 nm I - in the GNS growth solution. (d) Normalized absorption spectra of nanoparticles in the presence of varying amount of I -. 10 μm (red), 1 μm (blue),100 nm (green) and 10 nm (black). Scale bars: 100 nm 5
Figure S9: Characterization of the Au seeds used for the GNS synthesis. Absorption spectra (a,c), TEM micrographs (b,d) of d= 5 nm and d= 12 nm gold nanoparticles. Scale bars: 20 nm 6
Figure S10: The morphological and spectral stability of GNS is improved by surface passivation of GNS by PEG and storage at 4 0 C. (a) The LSPR peak position of bare and PEGylated GNS stored at room temperature (RT) and 4 0 C over 14 days. TEM micrographs showing the morphology of (a) just prepared bare GNS at RT, t=0 days, (c) bare GNS and (d) PEGylated GNS stored at room temperature and 4 0 C after t= 4 days. Scale bars: 100 nm 7