NANO LETTERS

Metal Nanoshell Assembly on a Virus Bioscaffold

Corey Radloff and Richard A. Vaia*
Air Force Research Laboratory, Materials and Manufacturing Directorate, Wright-Patterson, Ohio

2005 Vol. 5, No. 6 1187-1191

Jason Brunton, Gregory T. Bouwer, and Vernon K. Ward

Department of Microbiology and Immunology, School of Medicine, UniVersity of Otago, Dunedin, New Zealand
Received April 8, 2005; Revised Manuscript Received April 25, 2005

ABSTRACT
Chilo iridescent virus is demonstrated as a useful core substrate in the fabrication of metallodielectric, plasmonic nanostructures. A gold shell is assembled around the wild-type viral core by attaching small, 25-nm gold nanoparticles to the virus surface by means of the chemical functionality found inherently on the surface of the proteinaceous viral capsid. The density of these nucleation sites was maximized by reducing the repulsive forces between the gold particles through electrolyte addition. These gold nanoparticles then act as nucleation sites for the electroless deposition of gold ions from solution around the biotemplate. The optical extinction spectra of the metalloviral complex is in quantitative agreement with Mie scattering theory. Overall, the utilization of a native virus and the inherent chemical functionality of the capsid afford the ability to grow and harvest biotemplates for metallodielectric nanoshells in large quantities, potentially providing cores with a narrower size distribution and smaller diameters (below 80 nm) than for currently used silica.

The emergent field of plasmonics seeks to engineer, on the nanoscale, the electrooptical interaction between metal nanostructures and their surroundings.1-5 Applications range from surface-enhanced Raman spectroscopy,6-8 surface plasmon resonance sensing,2,9 and cancer treatment10,11 to the enhancement of the nonlinear optical response of organic materials.12,13 For these specific applications, metallodielectric nanoshells have been shown to provide a suite of beneficial characteristics. Facile and systematic tunability of the plasmon resonance can be achieved through systematic changes of the size ratio of the dielectric core and surrounding gold or silver metallic shell.4,9,14 For example, the tuning of the plasmon resonance from 650 nm to 6.0 m of silica-gold, core-shell particles has been demonstrated.4,15,16 Although numerous routes to fabricate these nanostructured materials have been articulated (the vast majority utilize conventional colloidal templates), increasing the fidelity and flexibility of the fabrication is still desired. In general, biological templates offer much promise but unfortunately have not been extensively applied to metallodielectric nanoshells. One such template, viruses, exhibits the characteristics of an ideal nanobuilding block: prescribed composition, monodispersity, site-specific heterogeneous surface chemistry, accessible interior, and extensive chemical tai* Corresponding author. E-mail: richard.vaia@wpafb.af.mil. Current address: Materials Laboratory, 3M Corporate Research Laboratory, 3M Center, Building 0201-01-W-28, St. Paul, Minnesota 55144. 10.1021/nl050658g CCC: $30.25 Published on Web 05/04/2005 2005 American Chemical Society

lorability. Viruses are generically composed of a proteinaceous shell (capsid) surrounding genomic material. They offer a suite of beneficial characteristics for the synthesis of metallodielectric nanoshells, including a number of morphologies (spheres, rods, and tubules) that are absolutely monodisperse and both external and internal surfaces that are chemically addressable by a broad range of organic and inorganic chemistries.17-19 For example, amino acids on the surface of the capsid, such as cystine, glutamic acid, and aspartic acid, present amine, carboxylate, and thiol groups that are amenable to complexation with metal nanoparticles. A capsid can also be easily modified to present additional thiol groups on the particle surface.17,19-22 The use of a virus core and adaptation of the aforementioned chemistries to the growth of metal nanoshells are discussed herein. Utilization of a native virus and the inherent chemical functionality of the capsid affords the ability to grow and harvest biotemplates in large quantities. These natural templates simplify the preparation of the dielectric core particle and provide a narrower size distribution and accessible core sizes below 80 nm. Furthermore, upon removal of the core genetic material, the resultant virus-like particle provides a hollow shell, enabling facile incorporation of various electrooptic molecules within the center of the nanoshell and thus accessing the local field enhancements within as well as at the surface of the nanoshell. Toward

these objectives, the first metallic nanoshells based on bioscaffolds are produced using Chilo iridescent virus as the dielectric core. Chilo iridescent virus (CIV) is a member of the genus IridoVirus in the family IridoViridae.23 From detailed studies of Yan et al.,24 CIV exhibits a layered structure, consisting of a dsDNA-protein core that is surrounded by a lipid bilayer. This is encapsulated by an inner and outer capsid shell composed of 51.4 kDa major capsid proteins. Fibers greater than 35 nm and centered over a hexavalent capsomer subunit protrude from the outer surface. Further investigations in support of this bioscaffold effort indicated that the fibrils are not easily removed without the complete disassembly of the virion particle. Furthermore, the composition of the fibril was not unambiguously determined; however, one putative candidate is an approximately 129 kDa protein that is strongly detected in western blots by surface-reactive antibodies to CIV. This protein can be coupled to biotin via a succinimde linkage to free amines (lysine residues), and these biotin molecules can be detected on the surface of viral particles by immunogold labeling with antibiotin monoclonal antibodies. In addition, the virus genome includes an open reading frame encoding a 129 kDa protein (CIV orf 261R, GenBank accession number NC_003038) with a repetitive -sheet structure analogous to that of amyloid fibrils and silk fibroins. Confirmation of the composition of the surface fibrils of IridoVirus particles would be an important step in the further use of these viruses as bioscaffolds. The production and collection of the Chilo iridescent virus is described in detail elsewhere.25-27 Collected particles are monodisperse with a diameter of 140 nm, Figure 1a. CIV was stored in concentrated, buffered solutions (10 mg/mL, 0.02% sodium azide). Particles were centrifuged and redispersed in ultrapure water to a concentration of nominally 1 mg/mL prior to use as cores for gold nanoshell deposition. The virus particle concentration was determined with respect to the optical density of the CIV solution at 260 nm (OD260).28,29 Concentration was determined by diluting a stock virus solution (2 mL, 1 mg/mL) with ultrapure water (Millipore, Billerica, MA) at different ratios from 0.5 to 0.9 and subsequently measuring the absorbance at each dilution. The absorbance range for the virus solutions was 0.16-1.38 corresponding to a virus concentration of 7.0 109-6.0 1010.28 This particle concentration is comparable with typical preparations for silica nanoparticles used as nanoshell cores. Gold nanoshells were fabricated using a modification of the method used for silica-gold nanoshells wherein the dielectric core is decorated with a high surface density of ultrasmall (2-5 nm) gold nanoparticles to act as nucleation sites for the electroless deposition of gold ions from solution.4 To seed the CIV surface, 0.1-0.5 mL of the stock CIV solution is added to 20-30 mL of a gold nanoparticle solution prepared using a previously reported method.30 This ensures an excess ratio of gold nanoparticles to CIV particles (40 000:1). To achieve high-density attachment of the gold nanoparticles, though, the ionic strength of the gold-CIV nanoparticle solution had to be increased using NaCl (1 M) electrolyte solution. Figure 1b represents a CIV core seeded
1188

Figure 1. Chilo iridescent virus (CIV) particles: (a) wild-type virus (scale bar 500 nm), (b) dilute gold seed coverage (0.060 M ionic strength), (c) dense gold seed coverage (0.102 M ionic strength). The scale bar is normalized for the images in b and c (Philips CM200, 200 kV).

from the as-prepared gold nanoparticle solution with an ionic strength of 0.060 M. The CIV particle can be seen as the light-gray sphere in the center of the image. The dark spots found sparsely on and around the core are the small gold nanoparticles. The few particles spatially removed from the capsid surface suggest that the fibrils also scavenge the gold nanoparticles. Upon increasing the ionic strength of the solution to 0.102 M, a significant increase in the gold nanoparticle coverage is observed (Figure 1c). Nanoparticle distribution does not appear completely random, though, but concentrated in specific regions and even lines. This would suggest that the increased salt concentration is not only reducing the repulsive forces between the gold nanoparticles but also collapsing the gold-decorated fibrils onto the surface of the virus capsid. The optimal electrolyte content (ionic strength between 0.102 and 0.337 M) and pH (5) were determined spectroscopically (UV/vis) and verified with TEM. The as-prepared gold nanoparticle solution exhibits a weak plasmon resonance at 515 nm.31 Maximum seed coverage occurred directly before the appearance of a strong resonance peak around 515 nm, which signified the presence of an increased concentration of aggregates of the gold particles. The increase in electrolyte concentration reduces the repulsive forces between the gold nanoparticles.32 At ionic strengths slightly less than required for aggregation in bulk solution, the repulsive forces between the CIV surface and gold nanoparticles are less than between gold nanoparticles in solution, and thus deposition on the CIV surface occurs. As observed from TEM, if the ionic strength was increased beyond this range, then gold aggregation began to decrease
Nano Lett., Vol. 5, No. 6, 2005

Figure 3. TEM images of gold nanoshell growth on CIV core particles. Shell thickness increases from a to c. The scale is normalized for all images.

Figure 2. Results of zeta-potential measurements on CIV particle solutions as a function of (a) pH and (b) ionic strength. Golddecorated CIV particles were prepared at pH 5 and an ionic strength of 0.20 M (approximately -7 mV). Increasing ionic strength decreased the overall surface charge of the virus particles.

surface density as islands of enlarged gold aggregates formed on the CIV surface. To more completely understand if these processing conditions alter virus-virus interactions or disrupt viral stability, Figure 2 summarizes the zeta potential (ZetaPALS, Brookhaven Instruments Corp.) for CIV solutions for a pH (2-9) and electrolyte (0.1-0.6M at pH 5) range bracketing fabrication conditions. From pH 5 to 9, the CIV exhibit a negative surface charge (-15 mV). Dynamic light scattering throughout the pH range indicated that the particle size is 145 ( 15 nm, in agreement with TEM and verifying that the CIV is dispersible with no signs of agglomerates for these solution conditions. With increased ionic strength, the CIV particles maintained their negative surface charge; however, it is reduced by almost half (-7.5 mV) for the ionic strength used for the preparation of the nanoshells (0.1-0.3 M). For the ionic strength range from 0.06 to 0.60 M, dynamic light scattering verified CIV dispersion. These results suggest that the increase in ionic strength has a greater effect on gold nanoparticle interactions than CIV interactions. The reduction of the surface charge surrounding the gold nanoparticles in conjunction with the decreased surface charge on the virus allows the gold and CIV to interact more closely, increasing the overall gold nanoparticle coverage. Following attachment of the gold nanoparticle seeds, gold shells of different thickness are formed by adding different volume aliquots of a gold-seeded CIV core solution to a stock gold ion solution (0.4 mM HAuCl4 and 1.8 mM potassium
Nano Lett., Vol. 5, No. 6, 2005

carbonate (K2CO3)). The ratio of gold-decorated CIV to stock gold solution ranged from 0.013 to 0.225. The addition of a small amount of hydroxylamine hydrochloride (20-30 L, 18 mM) acted as the reducing agent for electroless gold deposition. The progression of gold shell coalescence around the CIV is shown in Figure 3. The TEM image in Figure 3a shows that the gold colloid has increased in size and is beginning to coalesce on the virus surface. In Figure 3b, the shell has just begun to coalesce, and a complete shell can be seen in Figure 3c. The final structure of the gold shell is rough with an irregular surface. Extensive TEM during the growth process leads us to believe this is a direct result of less than ideal gold seed coverage achieved. Although the density of nucleation sites is satisfactory for shell growth, the heterogeneity and less than ideal distribution necessitate excessive gold reduction to complete the metallic shell over regions that were devoid of nuclei. The scavenging of the gold nanoparticles and partial collapse of the fibrils on the capsid surface likely exasperate the effect of the localized seed coverage. Figure 4 summarizes the evolution of the plasmon resonance as the shell thickness is increased on the CIV core. The bottom two spectra represent the enlarged gold nuclei on the capsid surface where the shoulder observed in the second spectrum is the coalescence of nearest-neighbor particles into islands and platelets. A strong dipole resonance is observed at 800 nm in the next spectrum corresponding to the just coalesced shell seen in Figure 3b. Increased absorption between 900 and 1000 nm, which arises from low-order nanoshell aggregates formed during the processing of the nanoparticles after electroless gold deposition, obscures this resonance in the spectra at later stages of shell growth. As shell thickness increases, a second peak evolves corresponding to a quadrupole resonance. The evolution of this peak begins in the third spectrum and can be seen as a shoulder at 550 nm. The quadrupolar mode of the plasmon resonance is due to phase retardation effects where the size
1189

Figure 4. Absorbance spectra for gold nanoshells on CIV particles. (a) Spectral series of gold nanoshell growth with increasing shell thickness from bottom to top (curves normalized and offset for clarity, Hewlitt Packard 8453 UV/vis spectrophotometer). (b) Comparison of observed (black) and calculated (shaded) nanoshell extinction spectra. The difference in the spectra can be attributed to increased inhomogeneous broadening effects due to irregularities in the gold shell and low-order nanoshell aggregates in solution.

and uneven shell structure discussed above. Irregularities in the shell result in an increase in the effects of inhomogeneous broadening on the nanoshell plasmon.37 The excess near-IR absorption at longer wavelengths (>900 nm) is due to loworder nanoshell aggregates in solution.38 Nevertheless, there is good agreement in the plasmon position and the overall structure of the plasmon response. The peak at 800 nm is attributed to a strong dipole resonance, and the shoulder observed at 600 nm is due to a developing quadrupole resonance. In conclusion, gold shell growth around Chilo iridescent virus cores is demonstrated using modifications to previous established fabrication routes. By decreasing the surface charge of the capsid through controlled introduction of excess electrolyte, maximization of the nonselective attachment of gold nanoparticles to the capsid surface enabled the growth of a complete gold shell around CIV. The resulting metallodielectric nanoshell exhibited a dipole resonance in the near-IR region and a strong quadrupole resonance at 550 nm. The extinction spectra of the virus-gold nanoshells were in quantitative agreement with model predictions of Mie scattering theory. The inherent surface chemistries of the wild-type viral capsid of CIV enabled a facile fabrication route for metallodielectric assembly. This is in contrast to virus-like particles where genetic engineering of the capsid and subsequent protein expression and assembly limit scaffold supply and significantly increase preparation and processing concerns. More extensive understanding of the amino acid distribution on the viral surface of the chosen capsid, rather than site-specific mutation, will improve the ability to form smooth shells through use of specific rather than nonspecific seed-capsid interaction. Acknowledgment. We gratefully acknowledge E. Thomas (MIT) and B. Farmer (AFRL) for valuable discussions and Richard Easingwood (South Campus Electron Microcope Unit, University of Otago) for microscopy of the wild-type virus. C.R. also acknowledges support from the National Research Council. The Air Force Office of Scientific Research, Asian Office of Aerospace Research and Development, Air Force Research Laboratory, and Materials and Manufacturing Directorate provided funding. References
(1) Hale, G. D.; Jackson, J. B.; Shmakova, O. E.; Lee, T. R.; Halas, N. J. Appl. Phys. Lett. 2001, 78, 1502. (2) Haes, A. J.; Van Duyne, R. P. J. Am. Chem. Soc. 2002, 124, 10596. (3) Maier, S. A.; Brongersma, M. L.; Kik, P. G.; Meltzer, S.; Requicha, A. A. G.; Atwater, H. A. AdV. Mater. 2001, 13, 1501. (4) Oldenburg, S. J.; Averitt, R. D.; Westcott, S.; Halas, N. J. Chem. Phys. Lett. 1998, 288, 243. (5) Sershen, S. R.; Westcott, S. L.; Halas, N. J.; West, J. L. J. Biomed. Mater. Res. 2000, 51, 293. (6) Haynes, C. L.; Van Duyne, R. P. J. Phys. Chem. B 2003, 107, 7426. (7) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. ReV. 1999, 99, 2957. (8) Oldenburg, S. J.; Westcott, S. L.; Averitt, R. D.; Halas, N. J. J. Chem. Phys. 1999, 111, 4729. (9) Sun, Y.; Xia, Y. Anal. Chem. 2002, 74, 5297. (10) Hirsch, L. R.; Stafford, R. J.; Bankson, J. A.; Sershen, S. R.; Rivera, B.; Price, R. E.; Hazle, J. D.; Halas, N. J.; West, J. L. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13549. (11) ONeal, D. P.; Hirsch, L. R.; Halas, N. J.; Payne, J. D.; West, J. L. Cancer Lett. 2004, 209, 171.

of the illuminated nanoparticles is on the order of the incident light waves. This results in a spatially varying electromagnetic field across the nanoshell, exciting higher-order multipolar resonances. The image in Figure 3c correlates well with the top spectrum in Figure 4a, showing a strong plasmon resonance composed of a dipole mode at 800 nm and a quadrupole mode at 550 nm. This is characteristic of a complete, thick gold shell. In general, this sequence of spectra during gold shell growth on CIV is comparable to growth on a silica core.4 The plasmon response of the CIV-gold nanoshells agrees well with existing Mie scattering theory for coated spheres.33,34 As an example, Figure 4b depicts the agreement between the experimental spectra for nanoshells (Figure 4a) with a calculated far-field extinction spectrum of a nanoshell with a CIV core that is 139.5 nm in diameter and a gold shell that is 22 nm thick using a scattering parameter of A ) 2. These structural parameters are in good agreement with TEM and light scattering analysis. A core dielectric constant of 1.85 was determined from experimental data on the effective refractive index of viral nanoparticles.35,36 The plasmon resonance calculated for the virus core is narrower than the measured nanoshell plasmon. This is attributed to the rough
1190

Nano Lett., Vol. 5, No. 6, 2005

(12) Yelin, D.; Oron, D.; Thiberge, S.; Moses, E.; Silberberg, Y. Opt. Express 2003, 11, 1385. (13) Wenseleers, W.; Stellacci, F.; Meyer-Friedrichsen, T.; Mangel, T.; Bauer, C. A.; Pond, S. J. K.; Marder, S. R.; Perry, J. W. J. Phys. Chem. B 2002, 106, 6853. (14) Graf, C.; van Blaaderen, A. Langmuir 2002, 18, 524. (15) Oldenburg, S. J.; Jackson, J. B.; Westcott, S. L.; Halas, N. J. Appl. Phys. Lett. 1999, 75, 2897. (16) Prodan, E.; Radloff, C.; Halas, N. J.; Nordlander, P. Science 2003, 302, 419. (17) Blum, A. S.; Soto, C. M.; Wilson, C. D.; Cole, J. D.; Kim, M.; Gnade, B.; Chatterji, A.; Ochoa, W. F.; Lin, T.; Johnson, J. E.; Ratna, B. R. Nano Lett. 2004, 4, 867. (18) Douglas, T.; Young, M. Nature 1998, 393, 152. (19) Gillitzer, E.; Willits, D.; Young, M.; Douglas, T. Chem. Commun. 2002, 2390. (20) Behrens, S.; Rahn, K.; Habicht, W.; Bohm, K.-J.; Rosner, H.; Dinjus, E.; Unger, E. AdV. Mater. 2002, 14, 1621. (21) Dujardin, E.; Peet, C.; Stubbs, G.; Culver, J. N.; Mann, S. Nano Lett. 2003, 3, 413. (22) Douglas, T.; Strable, E.; Willits, D.; Aitouchen, A.; Libera, M.; Young, M. AdV. Mater. 2002, 14, 415. (23) Ward, V.; Kalmakoff, J. Invertebrate Iridioviridae. In Viruses of InVertebrates; Kurstak, E., Ed.; Marcel Dekker: New York, 1991; p 197. (24) Yan, X.; Olson, N. H.; Van Etten, J. L.; Bergoin, M.; Rossman, M. G.; Baker, T. S. Nat. Struct. Biol. 2000, 7, 101.

(25) Cerutti, M.; Devauchelle, G. Virology 1985, 145, 123. (26) Cerutti, M.; Devauchelle, G. In Molecular Biology of IridoViruses; Darai, G., Ed.; Kluwer Academic Publishers: Boston, 1985; p 81. (27) Webby, R. J.; Kalmakoff, J. Virus Res. 1999, 59, 179. (28) Constantino, M.; Christian, P.; Marina, C. F.; Williams, T. J. Virol. Methods 2001, 98, 109. (29) Kalmakoff, J.; Tremaine, J. H. J. Virol. 1968, 2, 738. (30) Duff, D. G.; Baiker, A.; Edwards, P. P. Langmuir 1993, 9, 2301. (31) Radloff, C.; Vaia, R. A.; Brunton, J.; Ward, V.; Kalmakoff, J.; Dokland, T. Bioscaffolds for Metal Nanostructures; SPIE - The International Society for Optical Engineering: Denver, CO, 2004. (32) Kooij, E. S.; Brouwer, E. A. M.; Wormeester, H.; Poelsema, B. Langmuir 2002, 18, 7677. (33) Mie, G. Ann. Phys. 1908, 24, 377. (34) Sarkar, D.; Halas, N. J. Phys. ReV. E 1997, 56, 1102. (35) Cole, T.; Kathman, A.; Koszelak, S.; McPherson, A. Anal. Biochem. 1995, 231, 92. (36) Camerini-Otero, R. D.; Franklin, R. M.; Day, L. A. Biochemistry 1974, 13, 3763. (37) Westcott, S. L.; Jackson, J. B.; Radloff, C.; Halas, N. J. Phys. ReV. B 2002, 66, 155431. (38) Hirsch, L. R.; Jackson, J. B.; Halas, N. J.; West, J. L. Anal. Chem. 2003, 75, 2377.

NL050658G

Nano Lett., Vol. 5, No. 6, 2005

1191