Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Angew Chem Int Ed Engl. Author manuscript; available in PMC 2010 August 17.
Published in final edited form as:
PMCID: PMC2922693

Etching and Growth: An Intertwined Pathway to Ag Nanocrystals with Exotic Shapes

Claire M. Cobley and Matt Rycenga
Department of Biomedical Engineering, Washington University St. Louis, MO 63130 (USA)
Fei Zhou and Zhi-Yuan Li
Institute of Physics, Chinese Academy of Sciences Beijing 100080 (China)

Shape-controlled synthesis has proven to be a powerful means for controlling the properties of metal nanocrystals and optimizing them for applications in catalysis,[1] electronics,[2] sensing,[3] biomedical imaging,[4] and surface-enhanced Raman scattering (SERS).[5] Through careful control of reaction conditions (e.g., reaction temperature, surface capping, and concentrations of reagents and ionic species), nanocrystals with a wide variety of shapes have been synthesized.[6] However, the majority of nanocrystals that have been achieved so far are highly symmetric, as confined by the face-centered cubic lattice taken by most metals. Of the anisotropic shapes that have been observed, the majority (e.g., bars, rods, and wires) are a result of preferential growth along a single direction.[2,7,8] To further increase the diversity of nanocrystal shapes, we need to find new routes to break the cubic symmetry and thus force the growth process into other anisotropic modes.

Recently, seeded overgrowth has been demonstrated as a versatile route to the formation of nanocrystals with both simple and complex shapes and compositions, including bimetallic samples.[9, 10] A typical example is the transformation of Ag nanocubes into their geometric dual, octahedrons, through preferential overgrowth at all {100} facets.[10] In this Communication, we present a new etching-induced growth mechanism by which Ag nanocubes are transformed into nanocrystals with an exotic, previously unattained shape: anisotropically-truncated octahedrons. In this case, the overgrowth occurs preferentially on three adjacent faces of the six available, all of which surround a corner slightly truncated due to oxidative etching, resulting in a non-centrosymmetric shape despite the single crystal structure.

The synthesis started with a typical sulfide-mediated polyol process for Ag nanocubes.[11] At the end of this process, a second aliquot of AgNO3 solution was added and, to our surprise, the cubic nanocrystals were found to evolve into anisotropically-truncated octahedrons, a shape of lower symmetry relative to a cube or octahedron. Figure 1 shows electron micrographs of the sample before and after the second aliquot of AgNO3 solution was introduced. As shown in Figure 1a, the sulfide-mediated synthesis gave a uniform sample of Ag nanocubes 46 nm in edge length. Ten minutes after the addition of the second aliquot of AgNO3 solution, essentially all the Ag cubes had been transformed into anisotropically-truncated octahedrons (Figure 1b) of 68 nm in size as measured along the longest edge. Different from a regular octahedron, three adjacent corners of this new nanocrystal are snipped significantly, as illustrated in Figure 1c.

Figure 1
When a second aliquot of AgNO3 was introduced at the end of a sulfide-mediated polyol synthesis, the Ag nanocubes evolved into a new anisotropic structure rapidly: a, b) SEM images with TEM insets of the product (a) before and (b) 10 min after introduction ...

It has been previously shown that gradually adding AgNO3 and PVP solutions to the product of a conventional polyol synthesis over the course of 2 h could facilitate the transformation of Ag cubes of 80 nm in edge length into octahedrons of 300 nm in size.[10] This shape transformation could be attributed to faster addition of Ag atoms to the {100} faces of the cube than the {111}-capped corners. In the present work, we propose that the same principle of more rapid growth on {100} faces is still valid, but in a much less symmetrical pattern. Instead of being added to all six faces of the cube evenly, the Ag atoms were added to three adjacent {100} faces more rapidly than the other three {100} faces. Figure 1c shows a schematic of this new growth mechanism, where white and grey signify the {111} and {100} facets, respectively. The three fast-growing {100} faces are determined by a slightly truncated corner, which is believed to be the source of the highly anisotropic growth (see below). As a result, half of the cube grows into an octahedron while the other half retains a truncated cubic morphology. When sitting on a substrate, such an exotic nanocrystal typically takes on one of two orientations. Either it sits on the large {111} facet on the “octahedron side” of the crystal, or it sits on one of the three square {100} facets.Figure 1d shows both of these orientations, which match well with what was observed under SEM and TEM (Figure 1b).

Both the formation of Ag nanocubes that serve as the seeds and the subsequent growth into anisotropically-truncated octahedrons are rapid processes. The presence of sulfide species, in this case HS, results in accelerated growth of the cubes in the first step due to the generation of Ag2S, a known catalyst for Ag reduction.[13] The reaction involved in the second step is also rapid; the injection of AgNO3 takes under one minute and the final product is harvested 10 min later. This rapid rate makes it possible for the final shape to be kinetically determined instead of being the thermodynamically favored shape at this size, a cubo-octahedron (or a truncated octahedron). To confirm the importance of this rapid growth, a similar experiment was performed except the second aliquot of AgNO3 was added 15 times more slowly, at a rate of 0.05 mL/min rather than 0.75 mL/min. Instead of growing anisotropically, the Ag atoms were added uniformly to all six {100} faces, retaining the cubic morphology during overgrowth (Figure S1a).

Oxidative etching is also an important factor in determining the final shape of nanocrystals obtained in a solution phase synthesis and has been proposed as a basis for the activation of specific face(s) of a nanocrystal for further growth.[6,12] In our previous work, both a rapid reduction rate and localized oxidative etching were shown to be critical to break the cubic symmetry and promote anisotropic growth of Ag and Pd cubes into rods or bars. Adding a capping agent that prevents oxidative etching was shown to shorten Pd nanobars significantly, resulting in a cubic shape.[12] Etching may also play a role in the synthesis of Ag nanobars as increased concentrations of the etchant Br were necessary for their growth.[7] A similar mechanism appears to be responsible for the anisotropic growth observed in the present study as well, with a significant difference: instead of a single face being activated for further growth, the etching of one corner of the cube promotes growth on all three adjacent faces connected by this corner. Mild etching is known to occur in the late stages of a sulfide-mediated Ag nanocube synthesis. The corners of the cubes have been shown to round irregularly if the reaction was left unquenched after all the AgNO3 had been consumed.[11] As indicated by arrows in the inset of Figure 1a, a number of particles could be seen in the TEM images of the seed cubes with uneven corner truncation. A close examination of multiple micrographs of the nanocubes in Figure 1a indicates that ~20% of them appear to be truncated at one corner. It is not expected to see truncation on every cube since the contrast difference can be difficult to see if the truncation is not significant enough. This uneven etching could activate one corner of the cube, and the Ag that is dissolved from this region is likely to be re-deposited in a nearby area, activating the adjacent three faces for further growth once additional AgNO3 is introduced. The stirring in this reaction is mild and thus should allow for local forces to play a significant role. To confirm the importance of oxidative etching in our mechanism, the same reaction was performed with an Ar-saturated solution. In this case, no shape transformation was observed and the final product was simply Ag nanocubes (Figure S1b) as no face was preferentially activated.

To confirm our assignment of this unusual shape, extensive electron microscopy analysis was performed and all results were consistent with an anisotropically-truncated octahedron. Other geometries were also investigated, including bipyramids, truncated tetrahedrons, and unevenly truncated cubes, but none of them could explain all the data presented here. Figure 2a shows a high-magnification SEM image, allowing the faceted nanocrystals to be better resolved. This image clearly shows multiple examples of the two most common profiles of the nanocrystals: “houses” and triangles. A typical “house” orientation is shown in Figure 2b at three different tilting angles, in comparison to a model. The two sets of images match closely, though the actual nanocrystals are slightly more rounded when compared with the idealized model due to corner truncation. Further tilting of this model also suggests that the seemingly non-uniform appearance of the sample is most likely caused by its highly anisotropic nature (Figure S2).

Figure 2
a) A high-magnification SEM image of the anisotropically-truncated octahedrons of Ag with well-developed facets. b) TEM images taken from a single anisotropically-truncated octahedron at three different tilting angles and images of a model which has been ...

High-resolution TEM analysis in Figure 3 validates the single crystal structure of the product, confirms the presence of both {100} and {111} facets, and verifies the fringe spacings expected from the model at different orientations. Figure 3, a and b, shows the analysis of a nanocrystal sitting on the large triangular face bound by a {111} plane. The fringe spacing of 1.4 Å can be indexed to the {220} reflection of Ag. The inset in Figure 3b shows a fast Fourier transform (FFT) of the high-resolution image, where the spots have a 6-fold symmetry and can be indexed to the {220} reflection, indicating that the nanocrystal was sitting on a {111} face. Figure 3, c and d, shows the analysis of a nanocrystal sitting on a square {100} face. The fringe spacing of 2.0 Å corresponds to the {200} reflection of Ag. The FFT pattern in the inset shows a square symmetry and spots for both the {200} and {220} reflections, indicating that the nanocrystal is sitting on a {100} face.

Figure 3
High-resolution TEM analysis of the two orientations typically observed, with the anisotropically-truncated octahedron sitting on a (a, b) triangular, {111} face and (c, d) square, {100} face. Insets for (a) and (c) are models of the anisotropically-truncated ...

Figure 4a shows UV-vis spectra recorded from solutions of the two different stages depicted in Figure 1. After the second AgNO3 aliquot was added, the primary peak red-shifted by 25 nm, as would be expected from a size increase, and a new peak developed at 380 nm between the two peaks seen in the spectrum for nanocubes. Previously, discrete-dipole approximation (DDA) calculations for 40-nm Ag cubes and octahedrons showed that the main peak for octahedrons is located between the two peaks of a similarly sized cube as we see here, supporting our claim that our structure is a hybrid between a cube and an octahedron.[14] Additionally, we performed DDA calculation for an anisotropically-truncated octahedron that is depicted Figure S3. We used 3424 dipoles and the three sharp points of the “octahedron side” were snipped by 11.7 nm to reflect the slightly truncated nature of the particle and to match the experimental spectra more closely. As shown in Figure 4b, the same overall shape can be seen, with a clear shoulder at 380 nm.

Figure 4
a) Normalized UV-vis spectra of aqueous suspensions of the Ag nanocubes (dashed) and the corresponding product of anisotropically-truncated octahedrons (solid). b) DDA calculations of the extinction (solid), absorption (dashed), and scattering (dotted) ...

The unusual nanocrystals were further investigated for SERS applications. Silver is an ideal substrate for SERS due to its high polarizability, and can provide enhancement factors an order of magnitude larger than other metals such as gold.[15] Furthermore, nanocrystals with sharp tips, such as the points of an octahedron, can concentrate the field into small volumes and thus create regions with higher enhancement.[16] However, it is still not clear if the sharp corners need to be positioned in a specific configuration in order to generate a strong, localized electric field. We performed some preliminary measurements with 1,4-benzenedithiol (1,4-BDT) and a 514 nm laser to test the SERS capabilities of the new nanocrystals. Well-resolved spectra could be easily obtained and a typical example of the solution-phase spectrum is shown in Figure 4c, from which an enhancement factor of 7.5×103 was obtained for the 9a ring breathing vibration at 1183 cm−1. Single particle SERS spectra could also be obtained and a typical example is shown in Figure 4d. Interestingly, despite the sharp corners, the overall enhancement factor was slightly lower than what would be predicted based on studies of Ag nanocubes with a similar size.[16] This is likely due to the fact that though the nanocrystal has sharp corners, all of them are opposite a flat face instead of another corner, leading to a weaker dipole polarization. In this regard, the availability of Ag nanocrystals with sharp corners and different symmetries would provide a set of ideal substrates for better understanding the SERS phenomenon.

In summary, we have demonstrated anisotropic overgrowth that proceeds more quickly from three adjacent faces sharing a common, single corner of a nanocube activated due to oxidative etching. Rapid reduction was also found to be a key factor in this unique shape transformation. The final products are anisotropically-truncated octahedrons, which show a novel, non-centrosymmetric shape with interesting features for fundamental studies of SERS.


Briefly, 6 mL of ethylene glycol (EG) was preheated for 1 hour in a 24 mL vial at 152 °C. At this point, 70 μL of a 3 mM sodium hydrosulfide solution in EG was injected, followed by a poly(vinyl pyrrolidone) solution (30 mg in 1.5 mL EG) and a AgNO3 solution (24 mg in 0.5 mL EG) 8 minutes later. After 10 min, an aliquot was taken with a glass pipette through a small hole that had been drilled in the cap. Next, an additional aliquot of AgNO3 solution (24 mg in 0.5 mL EG) was added using a syringe pump at a rate of 0.75 mL/min (0.05 mL/min for the control experiment), also through the hole in the cap. After 10 min, the reaction was quenched in an ice bath and the product was washed with acetone and water. For argon-protected syntheses, the preheated EG and all solutions were bubbled with argon for 1 h and argon flow was maintained over the surface during the reaction. Additional details are available in supporting information.

Supplementary Material

supp mat


[**] This work was supported in part by the NSF (DMR-0804088), an NIH Director’s Pioneer Award (DP1 OD000798), and startup funds from Washington University in St. Louis. Z.Y.L. would like to acknowledge financial support from the National Natural Science Foundation of China (10525419, 60736041, and 10874238).


[1]a) Aiken JD, III, Finke RG. J. Mol Catal. A.: Chem. 1999;145:1. b) Lewis LN. Chem. Rev. 1993;93:2693.
[2] Xia Y, Yang P, Sun Y, Wu Y, Mayers B, Gates B. Adv. Mater. 2003;15:353.
[3]a) Rosi NL, Mirkin CA. Chem. Rev. 2005;105:1547. [PubMed] b) Daniel MC, Astruc D. Chem. Rev. 2004;104:293. [PubMed]
[4]a) Hu M, Chen J, Li ZY, Au L, Hartland GV, Li X, Marquez M, Xia Y. Chem. Soc. Rev. 2006;35:1084. [PubMed] b) West JL, Halas NJ. Annu. Rev. Biomed. Eng. 2003;5:285. [PubMed]
[5]a) Nie S, Emory S. Science. 1997;275:1102. [PubMed] b) Haynes CL, Van Duyne RP. J. Phys. Chem. B. 2001;105:5599.
[6] Xia Y, Xiong Y, Lim B, Skrabalak SE. Angew. Chem. Int. Ed. 2008;48:60. [PMC free article] [PubMed]
[7] Wiley BJ, Chen Y, McLellan JM, Xiong Y, Li Z-Y, Ginger D, Xia Y. Nano Lett. 2007;7:1032. [PubMed]
[8] Murphy CJ, Sau TK, Gole AM, Orendorff CJ, Gao J, Gou L, Hunyadi SE, Li T. J. Phys.Chem. B. 2005;109:13857. [PubMed]
[9]a) Lim B, Lu X, Jiang M, Camargo PHC, Cho EC, Lee EP, Xia Y. Nano Lett. 2008;8:4043. [PubMed] b) Habas SE, Lee H, Radmilovic V, Somorjai GA, Yang P. Nat. Mater. 2007;6:692. [PubMed] c) Fan FR, Liu DY, Wu YF, Duan S, Xie ZX, Jiang ZY, Tian ZQ. J. Am. Chem. Soc. 2008;130:6949. [PubMed] d) Burt J, Elechiguerra J, Reyesgasga J, Martinmontejanocarrizales J, Joseyacaman M. J. Cryst. Growth. 2005;285:681. e) Seo D, Yoo CI, Jung J, Song H. J. Am. Chem. Soc. 2008;130:2940. [PubMed] f) Skrabalak SE, Xia Y. ACS Nano. 2009;3:10. [PubMed]
[10]a) Tao A, Sinsermsuksakul P, Yang P. Angew. Chem. Int. Ed. 2006;45:4597. [PubMed] b) Tao AR, Ceperley DP, Sinsermsuksakul P, Neureuther AR, Yang P. Nano Lett. 2008;8:4033. [PubMed] c) Seo D, Yoo CI, Park JC, Park SM, Ryu S, Song H. Angew. Chem. Int. Ed. 2008;47:763. [PubMed]
[11] Siekkinen A, McLellan JM, Chen J, Xia Y. Chem. Phys. Lett. 2006;432:491. [PMC free article] [PubMed]
[12] Xiong Y, Cai H, Wiley BJ, Wang J, Kim MJ, Xia Y. J. Am. Chem. Soc. 2007;129:3665. [PubMed]
[13] Kryukov A, Stroyuk A, Zinchuk N, Korzhak A, Kuchmii S. J. Mol. Catal. A: Chem. 2004;221:209.
[14] Wiley BJ, Im SH, Li Z-Y, McLellan JM, Siekkinenen AR, Xia Y. J. Phys. Chem. B. 2006;110:15666. [PubMed]
[15] Quinten M. Appl. Phys. B. 2001;73:245.
[16] McLellan JM, Li Z-Y, Siekkinen AR, Xia Y. Nano Lett. 2007;7:1013. [PubMed]