Biotemplated Aqueous-Phase Palladium Crystallization in the

Sep 16, 2010 - Decoupling and elucidation of surface-driven processes during inorganic mineralization on virus templates. Oluwamayowa O. Adigun , Glor...
14 downloads 10 Views 401KB Size
pubs.acs.org/NanoLett

Biotemplated Aqueous-Phase Palladium Crystallization in the Absence of External Reducing Agents Jung-Sun Lim,† Seung-Min Kim,‡ Sang-Yup Lee,§ Eric A. Stach,‡ James N. Culver,| and Michael T. Harris*,† †

School of Chemical Engineering and ‡ School of Materials Engineering and Birck Nanotechnology Center, Purdue University, West Lafayette, Indiana 47907, § Department of Chemical and Biomolecular Engineering, Yonsei University, Seoul, Korea 120-749, and | Center for Biosystems Research, University of Maryland Biotechnology Institute, College Park, Maryland 20742 ABSTRACT A new synthetic strategy enabling highly controlled aqueous-phase palladium crystallization on the tobacco mosaic virus (TMV) is demonstrated without the addition of external reducing agents. This low cost, solution processing method yields continuous and uniform coatings of polycrystalline palladium on TMV, creating highly uniform palladium nanowires of tens of nanometers in thickness and hundreds of nanometers in length. Our approach utilizes a palladium chloride precursor to produce metallic Pd coatings on TMV without the need for an external reducing agent. X-ray photoelectron spectroscopy and in situ transmission electron microscopy were used to confirm the reduction of the surface palladium oxide layer on the palladium metal wires during room temperature hydrogenation. This leads to metallic palladium nanowires with surfaces free of residual organics, making these structures suitable for applications in nanoscale electronics. KEYWORDS Mineralization, in situ, nanotube, palladium, reduction, cysteine, tobacco mosaic virus

N

ew synthetic strategies for producing nanostructured organic-inorganic hybrid materials have been inspired by the highly monodisperse nanorod structure of some biomaterials, which can serve as templates for the mineralization of various metals.1-8 Inorganic templates such as porous anodic alumina (PAA) have been utilized to produce metallic or semiconducting nanorods or nanowires with uniform diameter and length. However, these inevitably have irregularities introduced during the synthesis process.9-11 In comparison, biotemplates consist of several nanometer size building units that self-organize to produce highly monodisperse rod structures that are hundreds of nanometers in length.12-14 Also, biotemplates have surface functionalities that can be utilized for selective conjugations or inorganic mineralizations. As a result of these advantages over inorganic templates, metal and semiconductor nanocluster coatings on biotemplatessas well as their patterned assemblyshave been demonstrated as a bottom-up approach to synthesize macroscale organic-inorganic hybrid materials with potential applications in the field of nanoelectronics.4,6,7 Specifically, biotemplate virions have attracted attention as nanoscale templates due to their well-defined structure, monodispersed nature, abundant surface functionalities, and stability in various solvents. Tobacco mosaic virus (TMV) is one exemplary model template that exhibits these benefits.

TMV is a macromolecular assembly of 2130 coat proteins that forms a 18 × 300 nm2 rod-like particle, with a surface composed of repetitive units with a 2.3 × 3.5 nm2 area that is exposed at the fluid interface. In addition, TMV has a uniform density of active functionalities on its external surface. The major functionalities exposed to the external surface of TMV-wild are the hydroxyl, carboxyl, and amine groups.1,15,16 Site-directed mutagenesis enables the insertion of new functionalities on each unit surface, which can improve its capability for metal coating. For example, the insertion of a cysteine on each TMV coat protein has been shown to provide either highly reactive inorganic nucleation sites2,3,5,7 or covalent coupling sites with fluorescent tags.17 Cysteine mutant TMVs demonstrated significant improvements of metal precursor uptake capacity5 and mineralization2 compared to the TMV-wild. As the size of the target coating surface is reduced to the nanometer scale, more precise control of surface mineralization is required to successfully realize organic-inorganic hybrid nanoscale devices prepared by simple solution chemistry techniques. However, controlling the uniformity and efficiency of inorganic mineralization on biotemplates over the entire sample during the solution process has proven difficult. Except for rare cases,7,15 metal salt reduction on biologically driven templates has resulted in inefficient surface nucleation. Dense metal coverage on the TMV virion has been attempted through the use of an excessive amount of the inorganic precursor (palladium chloride × 100 higher than conditions reported by Lee et al.2 and Lim et al.5).

* Corresponding author. E-mail: [email protected]. Telephone: 765-494-0963. Received for review: 04/19/2010 Published on Web: 09/16/2010 © 2010 American Chemical Society

3863

DOI: 10.1021/nl101375f | Nano Lett. 2010, 10, 3863–3867

Despite the fact that the reduction of the metal precursor on the virion improved the surface coverage, it resulted in an uncontrollable and irregular metal layering on the virions.8 Therefore, maintaining a controlled and continuous inorganic coating on a biotemplate without sacrificing uniformity over the entire sample remains a fundamental challenge in biotemplate-directed inorganic mineralizations. These technological limitations are mostly due to the insufficient elucidation of the mechanisms that control inorganic mineralization onto biotemplates. This process has for the most part been explained by the hypothetical speculation1,15,16 that the electrostatically attracted precursors on templates are nucleated and grown by the reduction of the metal ions by chemicals, such as hydrazine and dimethylamine borane (DMAB). Recently, there have been a few experimental characterizations of precursor biosorption on the virion or the correlation of the precursor sorption to the final mineralization after metal ion reduction by the reducing agents.5 An effective alternative coating approach was investigated by immobilizing the TMV on the substrate, using different reducing agents, and controlling the concentration of the reducing agents.18,19 Here, we present a new strategy for creating large quantities of uniform Pd-TMV nanocomposites using a simple and low-cost solution processing approach. As a reminder, this new approach does not use externally supplied reducing agents for the palladium mineralization. Nevertheless, a polycrystalline palladium layer was formed on the TMV surface whose smoothness and density are highly improved, comparing to conventional methods. High-resolution transmission electron microscopy (TEM) images indicate that there is a thin layer of palladium oxide or palladium-oxy-chloride species at the metal-coating (solid)-liquid interface. In this study, a palladium chloride precursor was biosorbed onto two types of TMVs, TMV wild-type and TMV2Cys2,3 (cysteine mutant TMV), in deionized (DI) water for 30 min at 50 °C, resulting in the formation of browncolored precipitates. Biosorptions were accomplished in the absence of any external reducing agent. The precipitates were carefully collected and washed five times with DI water. Thicker palladium layering was achieved by applying two more coating cycles. The Pd-coated TMV produced by this synthesis approach is compared to previous experiments, and a more traditional synthesis method, where an external reducing agent, DMAB, is added to the sample. The palladium stained TMVs were imaged with TEM (FEI 80-300 Titan) and characterized with X-ray photoelectron spectroscopy (XPS) (Kratos Ultra DLD spectrometer). An in situ hydrogenation experiment was performed in an environmental cell TEM (E-TEM) (FEI 80-300 Titan) by flowing H2 gas, and an ex situ hydrogenation was done in a reaction cell of the XPS spectrometer. Figure 1 shows the difference in morphologies of the palladium coating on TMV for samples prepared (a) without adding the external reducing agent and (b) with addition of © 2010 American Chemical Society

FIGURE 1. Palladium biomineralization on the TMV by (a) selfmineralization process and by (b) adding an external reducing agent (DMAB). The scale marker on the insets is 50 nm.

FIGURE 2. Palladium biomineralization on TMV wild-type (a/c) and TMV2Cys (b/d) after first/third palladium mineralization cycle. From images (c) and (d), the thickness of Pd-TMV nanowires after the third mineralization on the TMV wild-type and TMV2Cys was estimated as 25.6 ( 2.4 and 33.4 ( 1.1 nm, respectively. Schematics below thefiguresshowTMVwild-type(left)andTMV2Cys(right),respectively.

the external reducing agent. The detailed experimental procedures for the preparation of samples shown in Figures 1-5 are available in the Supporting Information section. The palladium mineralization on TMV (after three coating cycles and without the external reducing agent), as shown in Figure 1a, produced uniform metal nanowires, whereas samples that were produced by the addition of DMAB (Figure 1b) created incomplete and irregular-shaped coatings. Lim et al.5 proposed that the incomplete coating during the use of the external reducing agent originates from the reduction and mineralization of surface-bound metal ions on the TMV and the unbound metal ions in the medium. The amount of the soluble palladium precursor added to the samples was considerably more than the amount that could be sorbed on 3864

DOI: 10.1021/nl101375f | Nano Lett. 2010, 10, 3863-–3867

supplied reducing agents are not essential to initiate the palladium mineralization on TMV. A mechanism for this selfmineralization process for the reduction of Pd will be discussed below. In order for this new synthesis method to be widely used, it is important to demonstrate not only large-scale uniformity but also consistency of the coating thickness on each TMV nanorod. Figure 2 shows excellent consistency of the thickness of the nanowires by varying the density of surface functionality group (TMV2Cys vs wild-TMV) and the number of mineralization cycles. The TMV2Cys template (insertion of two thiols on each unit of 2.3 × 3.5 nm2, as shown schematically in Figure 2) effectively produced a thicker palladium coating (Figure 2d, 33.4 ( 1.1 nm) than that of the TMV-wild (Figure 2c, 25.6 ( 2.4 nm). The work by Lim et al.5 supports a greater palladium coating thickness on TMV2Cys than on TMV-wild, since enhancement of the surface-bound palladium precursor concentration is critical to improving the overall palladium mineralization on TMV. The extra cysteines on the mutant TMV, in comparison to TMV-wild, resulted in significant improvement of the palladium precursor uptake and mineralization. Attempts to further increase wire thickness by increasing the number of mineralization cycles beyond three did not lead to significant increases in wire thickness. This suggests that the selfmineralization process is retarded as the palladium deposit thickens. Despite this fact, this synthesis method allows effective control of nanowire thickness ranging from 25 to 35 nm. Figure 3a presents a high-resolution TEM (HRTEM) image near the gas-solid interface of the synthesized palladium coating on TMV. There is a thin layer of palladium oxide or palladium-oxy-chloride at the solid-liquid interface of the mineralized coating of polycrystalline palladium on TMV. The inset in Figure 3a is a fast Fourier transformation (FFT) diffractogram from the region boxed in red. A highly crystalline structure is confirmed, and the symmetry in the diffractogram reveals that the nanowire has a face centered cubic (FCC) crystal structure (i.e., the pattern has symmetry and spacings consistent with an FCC 011 zone axis pattern). Figure 3b is an inverse FFT (IFFT) of the inset of Figure 3a, obtained by selecting only the primary reflections in FFT. Here, we measured the 002 and 111 spacings, which are closely matched to those of FCC palladium metal. This confirms that the as-synthesized nanowire is composed of FCC metallic palladium. In addition, electron energy loss spectroscopy (EELS) was obtained to confirm the palladium composition of the nanowire. As shown in Figure 3c, the palladium M4,5 edge at 335 eV loss is clearly identified. The shape of the edge closely resembles that expected for a metallic palladium M4,5 edge, a shape that is clearly different from that of the M4,5 edge from palladium oxide, PdO.20 Thus both HRTEM and EELS confirm that the as-synthesized nanowires on TMV consist primarily of metallic palladium. However, as evident in Figure 3d, there is an amorphous

FIGURE 3. (a) HRTEM image of as-synthesized nanowires deposited on TMV-wild type after three cycles of self-mineralization. Inset in (a) is a FFT from the red box. (b) IFFT image obtained by selecting the primary reflections in the FFT. The IFFT is also from the red box in (a). (c) Background subtracted EELS spectrum showing the palladium M4,5 edge at 335 eV loss. (d) Four-times magnified image from the blue box in (a) showing an amorphous layer on the nanowire surface.

FIGURE 4. The XPS carbon 1s signals extracted from TMV2Cys (a) before and (b) after palladium mineralization. Note the shift of the binding energy from C-O or C-N (286.2) and amide (287.9 eV) for intact TMV2Cys to 286.5 and 288.4 eV for palladium stained TMV2Cys, respectively.

the surface of the TMV. Thus, the addition of the external reducing agent and the presence of a large amount of the unbound palladium precursor in the medium resulted in uncontrollable nanocluster growth in the medium due to the fast reduction reaction. Figure 1a shows that externally © 2010 American Chemical Society

3865

DOI: 10.1021/nl101375f | Nano Lett. 2010, 10, 3863-–3867

layer of 1-8 nm thickness that is always present on the nanowire surface. The significance of this amorphous layer is discussed below. XPS analysis was performed on TMV2Cys before and after palladium mineralization to determine if Pd complexes with the added functional groups on the TMV2Cys surface. The inner TMV organic constituents of Pd-coated TMV could not be detected by XPS. This is due to that the information depth of XPS is limited by ∼10 nm. Therefore, to analyze the chemical nature of TMV-Pd bound, the Pd-coated TMV was fractured by vigorously shaking the sample at temperatures below 2 °C (see the Supporting Information, Figure S1). The fractured Pd-coated TMV was subsequently analyzed with XPS as shown in Figure 4, which demonstrates a clear difference in the C 1s signals of TMV2Cys before and after the palladium biomineralization. Compared to the XPS signals of the unstained TMV2Cys shown in Figure 4a, the Pd-coated TMV2Cys in Figure 4b revealed a shift of the C-O (or C-N) and amide (-NHCO-)21 peaks to higher binding energies, suggesting that surface moieties were likely complexed with the palladium. This also implies that biotemplates or surface functionalities in the biomineralization process attract palladium ions to the TMV surface, resulting in reactive nucleation sites for palladium mineralization. We have also investigated palladium precursor uptake on substrates that are surface-functionalized with hydroxyl, amine, or thiol groups to elucidate the mechanism by which the Pd precursor sorbs onto the TMV surface and review these results briefly here.27 The thiol terminated substrate demonstrated superior uptake of the palladium precursor over amine/hydroxyl functionalities to form reactive nucleation sites under the same physiochemical conditions as those used to coat palladium on TMV. This result is consistent with the observed enhancement of palladium precursor uptake on the TMV2Cys over TMV-wild.5 Similar to the demonstration of external reducer free palladium mineralization on TMV, the palladium precursor stained substrate was exposed to the palladium precursor in DI water at 50 °C to grow the palladium polycrystalline layer without the external reducing agent. One important clue for the elucidation of the mechanism for the external reducer free palladium precursor mineralization is acid formation in the medium. Based on the solution acidification and the UV-vis absorption data that suggests the formation of HClO, PdClx(H2O)y precursors are complexed on the surface functionalities and self-mineralized by oxidizing Cl- to release HClO.27 We also investigated the amorphous layer formed on assynthesized Pd nanowires. Slight variations in processing conditions (i.e., extending washing times) alter the thickness of the amorphous layer from 1-2 to 7-8 nm (see the Supporting Information, Figure S2 for 7 nm thick amorphous layer). Surface composition often plays a significant role in device performance; therefore, the ability to easily remove this amorphous layer is often of paramount importance.22-24 © 2010 American Chemical Society

FIGURE 5. The XPS palladium 3d5/2 signals from palladium nanowires on TMV2Cys (a) before and (b) after the room temperature hydrogenation in a reaction cell of the XPS spectrometer. HRTEM images of palladium nanowires on TMV2Cys (c) before and (d) after flowing hydrogen for 10 min.

Pd-TMV hydrogenation experiment was performed in the XPS sample introduction chamber at room temperature, as shown in Figure 5. Figure 5a shows the Pd 3d5/2 peak at 337.4 eV from as-synthesized palladium nanowires and indicates that the surface amorphous layer mostly consists of palladium oxide.25 Interestingly, after the sample exposure to hydrogen at room temperature, a new Pd 35/2 peak appears at 335.6 eV (Figure 5b), suggesting that the surface palladium oxide or palladium-oxy-chloride layer covering the top of polycrystalline palladium is reduced to metallic palladium during this hydrogenation process. Similar phenomena were also observed through in situ TEM hydrogenation experiments, again conducted at room temperature. Figure 5c shows a HRTEM image of an amorphous layer on an assynthesized palladium nanowire before flowing hydrogen. After flowing hydrogen at 100 mTorr for 10 min, the amorphous layer was completely removed, as evident in Figure 5d. Both XPS and in situ TEM observations clearly indicate that the palladium oxide layer can be easily reduced to metallic palladium by simply flowing hydrogen at room temperature. Palladium is known to be very reactive in hydrogen environments, even at room temperature, and phenomena, such as room temperature Ostwald ripening of palladium nanoparticles and expansion of palladium nanocrystals in hydrogen, have been reported.22,23,26 However, this room temperature reduction of palladium oxide to metallic palladium in minutes is a new observation and is one that might be potentially useful to manipulate the conductivity of palladium nanowires or for producing highly conductive metal inks without thermal annealing. In summary, the formation of palladium nanowires on TMVs has been demonstrated using aqueous-phase biosorp3866

DOI: 10.1021/nl101375f | Nano Lett. 2010, 10, 3863-–3867

tion of palladium chloride on TMV, followed by metal reduction. This technique results in a more efficient and uniform palladium coating of the biotemplate. Finally, we have found that the amorphous palladium oxide surface layer created during synthesis can be effectively reduced by simple exposure to hydrogen gas at room temperatures, creating completely metallic nanowires.

(6) (7) (8) (9) (10) (11)

Acknowledgment. This research is supported by the US Department of Energy, Office of Basic Energy Science, Division of Materials Science and Engineering, Biomolecular Materials Research Program, (DEFG02-02-ER45975 and DEFG02-02-ER45976). The authors would like to thank Dr. Lisa Parsons at UMCP for schematic drawings of TMV wildtype and TMV2Cys with molecular modeling. We also acknowledge the support from Birck Nanotechnology Center at Purdue for the XPS analysis.

(12) (13) (14) (15) (16) (17) (18)

Supporting Information Available. The detailed experimental procedures available for the preparation of samples shown in Figures 1-5. Also, additional TEM images of fractured Pd-TMV and amorphous palladium layer on the TMV wild-type are available (Figures S1-S2). This material is available free of charge via the Internet at http:// pubs.acs.org.

(19) (20) (21) (22)

REFERENCES AND NOTES (1) (2) (3) (4) (5)

(23)

Dujardin, E.; Peet, C.; Stubbs, G.; Culver, J. N.; Mann, S. Nano Lett. 2003, 3, 413–417. Lee, S. Y.; Royston, E.; Culver, J. N.; Harris, M. T. Nanotechnology 2005, 16, S435–S441. Lee, S. Y.; Choi, J. W.; Royston, E.; Janes, D. B.; Culver, J. N.; Harris, M. T. J. Nanosci. Nanotechnol. 2006, 6, 974–981. Lee, Y. J.; Yi, H.; Kim, W. J.; Kang, K.; Yun, D. S.; Strano, M. S.; Ceder, G.; Belcher, A. M. Science 2009, 324, 1051–1055. Lim, J. S.; Kim, S. M.; Lee, S. Y.; Stach, E. A.; Culver, J. N.; Harris, M. T. J. Colloid Interface Sci. 2010, 342, 455–461.

© 2010 American Chemical Society

(24) (25) (26) (27)

3867

Nam, K. T.; Kim, D. W.; Yoo, P. J.; Chiang, C. Y.; Meethong, N.; Hammond, P. T.; Chiang, Y. M.; Belcher, A. M. Science 2006, 312, 885–888. Royston, E.; Ghosh, A.; Kofinas, P.; Harris, M. T.; Culver, J. N. Langmuir 2008, 24, 906–912. Royston, E. S.; Brown, A. D.; Harris, M. T.; Culver, J. N. J. Colloid Interface Sci. 2009, 332, 402–407. Biswas, K. G.; Sands, T. D.; Cola, B. A.; Xu, X. F. Appl. Phys. Lett. 2009, 94, 223116. Takai, A.; Saida, T.; Sugimoto, W.; Wang, L.; Yamauchi, Y.; Kuroda, K. Chem. Mater. 2009, 21, 3414–3423. Wang, W.; Huang, Q. H.; Jia, F. L.; Zhu, J. J. Appl. Phys. 2004, 96, 615–618. Culver, J. N. Annu. Rev. Phytopathol. 2002, 40, 287–310. Klug, A. Philos. Trans. R. Soc., B 1999, 354, 531–535. Lu, B.; Stubbs, G.; Culver, J. N. Virology 1996, 225, 11–20. Knez, M.; Sumser, M.; Bittner, A. M.; Wege, C.; Jeske, H.; Martin, T. P.; Kern, K. Adv. Funct. Mater. 2004, 14, 116–124. Shenton, W.; Douglas, T.; Young, M.; Stubbs, G.; Mann, S. Adv. Mater. 1999, 11, 253–256. Yi, H. M.; Nisar, S.; Lee, S. Y.; Powers, M. A.; Bentley, W. E.; Payne, G. F.; Ghodssi, R.; Rubloff, G. W.; Harris, M. T.; Culver, J. N. Nano Lett. 2005, 5, 1931–1936. Manocchi, A. K.; Horelik, N. E.; Lee, B.; Yi, H. Langmuir 2010, 26 (5), 3670–3677. Yang, C. X.; Manocchi, A. K.; Lee, B.; Yi, H. M. Appl. Catal., B 2010, 93 (3-4), 282–291. Ahn, C. C.; Krivanek, O. L.; Burgner, R. P.; Disko, M. M.; Swann, P. R. EELS Atlas: a reference guide of electron energy loss spectra covering all stable elements; Gatan, Inc.: Warrendale, PA, 1983. Ramanathan, T.; Fisher, F. T.; Ruoff, R. S.; Brinson, L. C. Chem. Mater. 2005, 17, 1290–1295. Favier, F.; Walter, E. C.; Zach, M. P.; Benter, T.; Penner, R. M. Science 2001, 293, 2227–2231. Xu, T.; Zach, M. P.; Xiao, Z. L.; Rosenmann, D.; Welp, U.; Kwok, W. K.; Crabtree, G. W. Appl. Phys. Lett. 2005, 86, 203104. Xiong, Y. J.; Chen, J. Y.; Wiley, B.; Xia, Y. A.; Yin, Y. D.; Li, Z. Y. Nano Lett. 2005, 5, 1237–1242. Hasik, M.; Bernasik, A.; Adamczyk, A.; Malata, G.; Kowalski, K.; Camra, J. Eur. Polym. J. 2003, 39, 1669–1678. Di Vece, M.; Grandjean, D.; Van Bael, M. J.; Romero, C. P.; Wang, X.; Decoster, S.; Vantomme, A.; Lievens, P. Phys. Rev. Lett. 2008, 100, 236105. Lim, J. S.; Kim, S. M.; Lee, S. Y.; Stach, E. A.; Culver, J. N.; Harris, M. T. J. Colloid Interface Sci. 2010, submitted.

DOI: 10.1021/nl101375f | Nano Lett. 2010, 10, 3863-–3867