Mimicking nature's strategies for the design of nanocatalysts

Abstract

Recent developments in bionanotechnology have produced a knowledge pool that enables the fabrication, functionalization, and activation of inorganic nanostructures. Continued progress in this field has led to advances in inorganic nanomaterial control, providing for the generation of catalysts that operate under biologically influenced conditions of temperature, pressure, and solvent. Outlined in this Perspective are a selection of catalysts active for a variety of reactions including C–C coupling, chemical reduction, electrocatalysis, and bond cleavage reactions, where a combination of both the inorganic core and biological surface work in concert to achieve the final functionality. By fully understanding the total structure/function relationship of these bio-inspired nanomaterials, new catalytic structures could be designed using biological principles that are energy neutral, eco-friendly, and selective, all of which represent grand challenges in light of the current global condition.

---

1. Introduction

Advancing catalytic technologies towards sustainable approaches for synthetic applications represents a grand challenge that is necessitated by the global condition. These changes are critically important due to the predicted exhaustion of fossil fuels, specifically oil, within ∼40 years.¹ This effect will cause a significant increase in the energy costs associated with catalysis, thus transforming catalytic technologies from highly energy consumptive to energy neutral processes will significantly reduce the costs associated with chemical and materials generation. Furthermore, this reduction in fossil fuel reserves also dramatically impacts the feed chemicals presently used as the building blocks of modern day organic syntheses. Such materials are extremely important for the continuity of molecular synthesis, thus designing new pathways to form these chemicals is required. Taken together, the impending exhaustion of fossil fuels can negatively impact the chemical and materials industries where new sustainable catalytic methods could rapidly prove to be useful. While recent advances in catalysis have striven to achieve energy neutrality, pathways to fabricate highly efficient catalysts for broad reactivity to reach these goals remain challenging.

Nanoparticles represent emerging systems that have demonstrated significant catalytic activity.² This activity arises from the high surface-to-volume ratio that is achieved at the nanoscale. This affords more efficient exposure of the metallic materials, rather than being buried away from reagents in solution within the metallic particle core, thus enhancing the efficiency of the materials on a per catalyst atom basis. The full activity of the composite structure is likely tempered by the ligands bound to the particle surface that imparts both stability and functionality.^2,3 These ligands must be designed to efficiently expose the inorganic catalytic materials while maintaining particle stability against undesired aggregation. Many ligands have been designed and used to achieve these goals including alkyl thiols,⁴ dendrimers,^2b,5 proteins,⁶ peptides,^3,7 and other complex organic species;⁸ however, a complete understanding of nanoparticle surface interfacial effects remains understudied, especially for catalytic technologies. It is possible that the functionality of an individual catalytic nanosystem could be fine tuned by the ligands to increase functionality, impart catalytic selectivity, and enhance recyclability for long-term use.

Over the past 10–15 years, many researchers have turned to nature for inspiration in the design and fabrication of inorganic materials that are functional under ambient conditions.⁹ Biology has used millennia of evolution to master the fabrication of inorganic structures for applications ranging from structural support to bioremediation.¹⁰ A hallmark of these pathways includes peptide and/or protein mediated syntheses where biomacromolecules typically interact with the growing materials at the nanoscale to direct both the synthesis and application.^9,10 Unfortunately, the inorganic materials typically observed in nature, while complex and of interesting morphologies, do not possess desirable function outside of their evolved application. Recently, researchers across numerous disciplines ranging from chemistry and biology to materials science and engineering have begun to mimic and exploit these approaches for the production of non-natural materials employing biological methods.^9,11 Using biocombinatorial approaches such as phage display, evolution can be rapidly processed to isolate short peptides with the ability to recognize and bind inorganic materials.^7b,9,12 These septamer or dodecamer peptides possess high binding affinities for their target materials and have been used to generate technologically important inorganic nanomaterials compositions including noble metals (Au, Ag, and Pd),^7b,12a,13 metal sulphides (CdS and ZnS),^12b,14 and metal oxides (TiO₂, SiO₂, and BaTiO₃).¹⁵ Furthermore, such peptide-mediated methods have been shown to form high temperature crystal structures of materials at low temperatures,^11,16 be highly precise in forming nearly monodisperse nanoparticles,^7b,13 and could potentially be used on an unlimited composition of materials. These unique bio-inspired structures have been employed as energy storage materials,¹⁷ biosensors,¹⁸ catalysts,^3a,7,19 and as motifs for complex materials assembly,²⁰ all of which occur under the ambient conditions of biology. These effects are in stark contrast to the conditions usually employed using standard materials that typically require extremes of temperature, pH, and pressure. This is especially true for catalysis that can require toxic organic solvents, high reaction temperatures, and can consume precious metal catalytic materials. By adapting biological-based materials for catalytic applications, new structures may be developed to address both the energy and environmental concerns of catalysis.

In this perspective, we discuss recent efforts of adapting bio-inspired nanotechnology towards catalytic applications, which is a main interest of our research group. In this discussion, we specifically focus on peptide-based nanoparticle systems that possess catalytic activities and operate under unprecedented reaction conditions of low temperatures in water for reactions ranging from C–C coupling to chemical reduction. Such approaches could potentially represent a new paradigm in catalytic materials technology to pave the road towards next generation efficient catalysts for sustainable organic and materials synthesis.

2. Using peptides to generate monometallic nanoparticle catalysts

Biological molecules have been used in directing the fabrication of inorganic materials such that they typically cap the surface to prevent aggregation.^9,21 While the peptides provide a stabilizing effect to the nanoparticles, they also can be selectively designed to impart functionality into the composite system. For instance, peptides can be programmed to control the shape of the final nanoparticle²² and optimize catalytic efficiency.^3a,23 These effects are likely due to the binding of the peptide at the particle surface; however, complete structural resolution of this factor remains difficult to fully determine due to limited analytical characterization methods at this small size scale. Nevertheless, peptide-capped nanoparticles have demonstrated catalytic functionality for a variety of reactions, all of which typically occur under biological conditions. For instance, Pd-based coupling reactions and chemical reductions have been well studied and represent model systems to translate bionano-materials to catalytic technologies.^{3,7,19,21,22b,23}

C–C coupling reactions such as Stille, Suzuki, and Heck^2a have been studied using peptide-capped Pd nanoparticles.^3,7a,b,19,23 Pd materials are not typically generated in biological systems; therefore, to fabricate such compositions, phage display was employed to isolate a dodecamer peptide sequence that specifically recognizes and binds to Pd metal.^7b In this approach, a bacteriophage library that possesses 10⁹ viruses that display random peptide sequences on the pIII minor coat protein are introduced to a target surface for inorganic selectivity. After multiple elution steps using increasingly stringent washing solutions, the remaining phage were heat ruptured, DNA-amplified, and sequenced to identify the peptides responsible for Pd binding. From this, multiple peptides were isolated from which a dodecamer peptide sequence (TSNAVHPTLRHL), termed Pd4, was chemically synthesized for use in the production of Pd nanoparticles.^7b The Pd4 peptide contains two histidine residues at the six and eleven positions, which computational modeling has suggested possess the highest affinity for Pd surfaces.²⁴ Secondary binding residues including arginine and asparagine are also present in the peptide. Based upon the modeling studies, it is suggested that the peptide forms a chelate-type interaction with the Pd surface through the two histidines to form a pinched structure along the nanoparticle surface,^7b,21 as shown in Fig. 1. From this structural motif, the Pd surface is more exposed to solution from which substrate could readily interact with the catalytic surface to drive selected chemical reactions.

Fig. 1 Scheme of the proposed binding of the Pd4 peptide to a Pd nanoparticle surface. Histidines are shown in red (Reprinted with the permission from M. Sethi, D. B. Pacardo and M. R. Knecht, Langmuir, 2010, 26, 15121–15134. Copyright 2010 American Chemical Society.)

Employing the Pd4 sequence, peptide-capped Pd nanoparticle fabrication is possible via standard reduction-based methods, shown in Fig. 2.^7b When the Pd4 peptide is introduced to K₂PdCl₄, a Pd²⁺ salt, complexation is observed between the peptide and metal ions, resulting in the formation of a ligand to metal charge transfer band (LMCT) that is monitorable via UV-vis spectroscopy. The LMCT arises from binding of the Pd salt to the amines of the peptide, which is noted at 215 nm in the UV-vis spectrum of the complex prior to reduction. Upon reduction with a significant excess of NaBH₄, stable nanoparticles are formed with a diameter of ∼2 nm as observed by transmission electron microscopy (TEM). Surprisingly, as the Pd:peptide ratio is varied during nanoparticle synthesis to values <5, stable nanoparticles of ∼2 nm are consistently noted.^7a This event is unique as particles of increasing size should be generated at higher metal : ligand ratios, as previously observed.²⁵ Current results suggest that the peptide binds to the particle surface at a certain point during particle growth due to highly specific biological recognition events to facilitate the formation of nanoparticles of identical sizes,^7a which is discussed further below. Such a small size is well positioned for catalytic applications due to the maximization of the surface-to-volume ratio to efficiently expose the catalytic Pd. Furthermore, the peptide-capping agent imparts a strong hydrophilic character to the materials, thus they are highly soluble in water for potential aqueous-based reactivity to avoid the use of large volumes of toxic organic solvents.

Fig. 2 Approach employed for the peptide-based synthesis of Pd nanoparticles and their use as a catalyst for Stille coupling. (Reprinted with the permission from D. B. Pacardo, M. Sethi, S. E. Jones, R. R. Naik and M. R. Knecht, ACS Nano, 2009, 3, 1288–1296. Copyright 2009 American Chemical Society.)

Once fabricated, the peptide-capped materials were found to be highly catalytically active for the C–C coupling Stille reaction (Fig. 2) in water at room temperature using a model reaction of coupling 4-iodobenzoic acid with PhSnCl₃ to generate biphenylcarboxylic acid.^7b Remarkably, the materials were able to generate quantitative product yields for this reaction, as well as reactions employing other aryl halide reagents that possessed both electron donating and withdrawing groups, down to a Pd loading of 0.005 mol% over a 24 h reaction timeframe. Using these materials at a Pd loading of 0.05 mol%, a turnover frequency (TOF) value of ∼2400 mol product (mol Pd × h)⁻¹ is observed, which is comparable to other systems prepared using non-biological surface ligands.^2a Furthermore, when using Pd loadings of <0.05 mol%, statistically equivalent TOFs are observed; however, at higher Pd concentrations, a decrease in the TOF value is evident.¹⁹ This decrease is attributed to a leaching-based catalytic mechanism where a Pd atom is abstracted from the surface of a particle during oxidative addition.^2a,19,26 At low Pd concentrations, the amount of bare Pd atoms in solution is low, thus resulting in great distances between atoms and preventing bulk aggregation. As this concentration is increased, more Pd atoms are leached into solution to facilitate bulk formation, resulting in the decrease in the TOF value.

While the Pd nanoparticles are highly reactive, the reaction substrates must penetrate the biological ligand layer to interact at the metal surface. Recent research suggests that the Pd4 peptide recognizes the Pd crystal structure to facilitate specific binding, which is likely imbedded during the phage-display isolation process.^7a As indicated above, while varying the Pd : peptide ratio used during material synthesis results in the fabrication of roughly the same sized nanoparticles, other characterization techniques indicate that the peptide surface structure on all of the materials is identical across the differently prepared samples.^7a Here, circular dichroism (CD) studies suggest that the Pd4 peptide adapts to a weakly structured coil conformation on the particle surface, which is the same for all Pd4-capped Pd nanoparticles, regardless of the reaction conditions employed. Furthermore, ζ-potential analysis also suggests that the particles possess the same surface charge when prepared at different Pd : peptide ratios; however, X-ray absorbance fine structure (XAFS) analysis of the materials indicates that while the same particles are generated, the complex prior to reduction plays an important role in the actual reduction step.^7a From this analysis, the degree of Pd²⁺ reduction was demonstrated to be proportional to the Pd : peptide ratio used during the materials synthesis such that as this ratio increased, a larger fraction of Pd atoms are reduced. As such, more zerovalent Pd⁰ atoms are present in the higher ratio samples for catalytic reactivity. This effect is likely due to the coordination environment of the Pd²⁺ prior to reduction in which the larger amine coordination at lower Pd : peptide ratios shifts the Pd reduction potential, lowering its ability to be reduced using standard NaBH₄-based methods. The degree of Pd²⁺ reduction during materials synthesis mediates the catalytic functionality for the model Stille system, suggesting that all steps from materials fabrication to catalytic application are important in determining the materials capabilities.

While it is likely that the Pd4 sequence recognizes Pd crystallographic features for materials fabrication and activity, the arrangement of the peptide on the nanoparticle surface is critically important to the catalytic functionality.^3a While the Pd4 peptide is likely to coordinate via the histidine residues, as computationally suggested (Fig. 3a),²⁴ alterations to this binding motif was demonstrated to substantially affect the catalytic abilities.^3a Changing of the histidine residues to alanines at the 6, 11, and 6 and 11 positions (Fig. 3) was employed to generate the A6, A11 and A6,11 peptide sequences, respectively.^3a This was used to directly probe peptide binding effects, shown in Fig. 3a–c, where residue swapping for alanine will remove binding at these positions. Nanoparticle synthesis with the modified peptides yielded similarly sized materials compared to the Pd4-capped structures (1.9 ± 0.4 nm). For instance, the A6 peptide generated Pd nanoparticles with an average size of 2.2 ± 0.4 nm, while the A11- and A6,11-capped materials possessed average diameters of 2.4 ± 0.5 nm and 3.7 ± 0.9 nm, respectively.^3a The secondary peptide structures bound at the materials surface were demonstrated to be significantly different, again as a function of the modifications, as observed by CD analysis. These changes in secondary structure indicate that different peptide/metal interfaces are present on the particle surface, which could dictate surface availability and reactivity.

Fig. 3 Comparison of the proposed binding of peptides (a) Pd4, (b) A6, and (c) A11 to a spherical Pd particle (histidines shown in red), while part (d) presents the Stille coupling TOF comparison for nanoparticles prepared with the selected peptide analogues at a Pd loading of 0.05 mol%. (Image d reproduced with permission from ref. 3a, copyright 2010 Wiley-VCH Verlag GmbH & Co. KGaA.)

Catalytic Stille analysis of the materials fabricated with the modified peptides employing the model reaction demonstrated a significant difference in TOF values as compared to the Pd4-capped Pd nanoparticles.^3a Using the parent Pd4-based structures, a TOF value of 2234 ± 99 mol product (mol Pd × h)⁻¹ is typically observed. The nanoparticles capped with the A6 peptide revealed a TOF of 5224 ± 381 mol product (mol Pd × h)⁻¹, which is a substantial enhancement in reactivity; however, the A11-prepared materials showed a decreased TOF of 1298 ± 107 mol product (mol Pd × h)⁻¹ and the A6,11-passivated materials demonstrated the largest decrease in activity with a TOF value of 361 ± 21 mol product (mol Pd × h)⁻¹ (Fig. 3d).^3a These results strongly suggest that the peptide along the particle surface can not only control the stability of the materials, but also the catalytic reactivity of the structures. This functional ability is likely achieved through controlling the access to the nanoparticle surface via the steric effects of the peptide binding and secondary structure along the nanoparticle surface. This is important knowledge in the a priori design of materials directing peptides, especially for catalytic-based materials, to increase their functionality and potentially their selectivity.

Beyond C–C coupling reactions, Huang and Li have synthesized peptide-capped Pt nanocrystals with electrocatalytic activity.^22b Phage display using Pt wires as the target was employed to isolate a peptide sequence with affinity for Pt metal. The final septamer sequence, Ac-TLHVSSY-CONH₂ (termed BP7A) was isolated after the third biopanning cycle.^22b Using this sequence, Pt crystals were synthesized at room temperature with a particle morphology showing a dependence on both the K₂PtCl₄ precursor and the peptide. Employing this peptide results in non-spherical materials termed multipods, where linear controlled growth from a core Pt particle along specific facets results in shapes that resemble pods in a single, bi, tri or tetra-pod orientation, as shown in Fig. 4.^22b This relationship between passivant and substrate elicits changes in the size and shape based upon growth kinetics. With a high quantity of peptide present, a larger fraction of the particle surface is passivated to inhibit growth and produce spherical structures; however, for the opposite case where low concentrations of peptide are present, gaps in the peptide-bound surface leave room for growth from more energetically favorable facets to create multipods. A study of the reversible potential of these structures with cyclic voltametry showed a two-fold increase in the electrochemical surface area (ECSA) of the Pt nanocrystals with larger numbers of pod growth structures. As a result, this increased surface area with respect to material mass increases the electrocatalytic activity.^22b Increasing the peptide concentration, the structures possessed fewer pods and became more spherical in shape that decreased the ECSA toward values typical of Pt black standards.

Fig. 4 HR-TEM images of the Pt multipod nanoparticles fabricated using the BP7A peptide. (Reproduced with permission from ref. 22b, copyright 2010 Wiley-VCH Verlag GmbH & Co. KGaA.)

While multipod-based Pt nanostructures were fabricated using the initial peptide sequence, Huang and coworkers also employed phage display to isolate facet selective peptides to generate Pt nanoparticles of varied shapes.^22a Such studies are highly important for catalytic applications as non-spherical nanomaterials, such as cubes and tetrahedra, have demonstrated enhanced reactivities due to the incorporation of higher numbers of coordinatively unsaturated metal atoms that are highly reactive.²⁷ Facet selective peptides were isolated via an M13 phage display library where Pt cubes (bound by six {100} facets) and Pt octahedra (bound by eight {111} facets) deposited on silicon were employed as the target surface.^22a The surface was sterilized in 99% ethanol for 30 min, followed by washing with water and tris-buffered saline (TBS). The phage library was then exposed to the targets in the presence of TBS followed by washing and elution steps that eliminated weakly bound phage. The phage bound to the surface were amplified and used for the next biopanning round. After three rounds, the phage were eluted and sequenced. Following this procedure, two septamer peptides named T7 (TLTTLTN) and S7 (SSFPEPD) were identified that bound the Pt {100} and {111} facets, respectively.^22a

Using the T7 and S7 sequences, non-spherical Pt nanocrystals were synthesized using a dual reductant-based approach.^22a,c Here, K₂PtCl₄ and the peptides were mixed with ascorbic acid, followed by the addition of NaBH₄.^22a In this situation, the peptide again acts as the capping agent, while the NaBH₄ rapidly reduces only a fraction of the Pt²⁺ ions to generate single crystal seed particles in solution that are capped with the peptide. Once the BH₄⁻ is consumed, ascorbic acid gradually reduces the residual Pt²⁺ ions to facilitate the growth of non-spherical Pt nanoparticles. Using the T7 sequence that recognizes the Pt {100} surface resulted in the formation of nanocubes enclosed by six {100} facets; however, when the same reaction conditions were employed in the presence of the Pt {111} binding S7 peptide, tetrahedra enclosed by Pt {111} facets were generated.^22a For the T7 stabilized Pt nanocubes, an average edge length of 7.6 ± 1.0 nm was observed after a reaction time of 30.0 min, while the S7 functionalized Pt tetrahedra possessed an average dimension of 6.0 ± 0.9 nm over a growth time period of 1.00 h.^22a

The extent of particle morphology development was directly dependent on the concentration of the facet specific peptide employed in the reaction.^22a Here an optimal peptide concentration was determined that led to the formation of more fully formed cubes and tetrahedra. Furthermore, when free S7 peptide was introduced to a solution of T7 nanocubes, the particle morphology quickly changed to tetrahedral geometries.^22a This suggests that the binding of the S7 peptide to Pt surfaces is stronger, as compared to the T7, which could be employed to control the morphology of the materials and could have dramatic impacts on their catalytic functionality.

Beyond nanocubes and tetrahedra, Huang and colleagues have been able to fabricate right bipyramids and {111}-bipyramids using the S7 and T7 peptides, respectively.^22c To alter the shape of the materials, a singly twinned Pt seed crystal was employed over the single crystalline seeds used above.^22a To generate the twinned seed materials, the BP7A peptide used to make the electrocatalytically active Pd multipods was employed.^22b In this sense, the BP7A generates the single-twinned particles, where Pt crystal multipod growth was prohibited under the synthetic condition. These seed crystals could then be used to grow the right bipyramids and {111}-bipyramids based upon the effects of the seed structure and peptide facet selectivity.^22c This suggests that the T7 and S7 sequences possess higher affinity for Pt surfaces over the BP7A sequence to allow for the selective shape growth. Overall, the Pt nanostructures synthesized using facet specific peptides provides precise control over the particle morphology that is likely to be of great interest for catalysis; however, their reactivities remain unstudied.

3. Peptide-fabricated multicomponent nanocatalysts

While monometallic nanoparticles can drive catalytic reactions, as described above, nanoparticles composed of multiple metallic components are known to enhance the functionality of the system.²⁸ This enhancement effect likely arises from two possibilities, electronic changes to the system or the geometric arrangement of the two components, both of which arise from the intimate contact between the two metals in the system.^28b,c,29 To generate this type of structure using peptides, Naik and colleagues have employed multifunctional peptides that are formed by the fusion of two materials binding sequences.^7dFig. 5c presents a schematic for the approach used for the formation of bimetallic Pd/Au nanoparticle catalysts using a bifunctional peptide. In this particular study, the Flg-A3 peptide was employed, which is engineered to contain two materials binding domains: one that binds Au (A3-AYSSGAPPMPPF) and another short sequence for binding secondary metal ions such as Pd⁴⁺ (Flg-DYKDDDDK). Interestingly, the order in which the two domains were incorporated into the fusion sequence demonstrated altered materials binding capabilities to their respective targets.^7d For instance, when two different fusion sequences were generated, Flg-A3 and A3-Flg that positioned either the Flg or A3 sequence as the N-terminus, respectively, different binding capabilities were observed for the secondary Pd material. Here, both sequences were used to fabricate Au nanoparticles where the A3 sequence should ideally bind to the Au and expose the Flg domain to solution for Pd binding. Using two sets of nanoparticles capped with the different peptides, altered binding of a two dimensional Pd surface was evident where the Flg-A3-capped Au nanoparticles demonstrated substantially higher Pd binding as compared to the other system.^7d As a result, the Flg-A3 sequence was employed to prepare the bimetallic Pd/Au nanoparticles.

Fig. 5 Peptide-based synthesis of bimetallic catalysts. Part (a) presents a low magnification TEM image of Au–Pd bimetallic nanoparticle, while part (b) presents an HR-TEM image of the materials displaying the crystal lattices of Pd present along the Au nanoparticle surface. Part (c) presents the synthetic scheme used to achieve these materials via the use of a fusion peptide sequence. (Reproduced with permission from ref. 7d, copyright 2010 Wiley-VCH Verlag GmbH & Co. KGaA.)

Using the fusion sequence, Au nanoparticles were generated that presented the Flg domain to solution.^7d The Flg region is known to bind metal ions; therefore, addition of Pd⁴⁺ to the system should position the ions and subsequent Pd nanoparticle directly at the Au nanoparticle surface. Upon NaBH₄ reduction of the Pd⁴⁺ ions, the formation of Pd nanoparticles of ∼3.0 nm occurs, depositing these materials directly onto the Au nanoparticle surface (Fig. 5a and b).^7d The formation of such bimetallic structures is due to the specificity of the interactions of materials-binding peptides and their surface display. In this arrangement, an intimate interaction is present between the Pd and Au components at the interface between the two particles. To that end, the more electronegative Au species is likely to pull electron density from the Pd materials, thus making them slightly electropositive to increase their reactivity for catalytic reactions such as olefin hydrogenation.

With confirmation of the bimetallic structure of the peptide-generated materials, their catalytic activity was examined for the hydrogenation of 3-buten-1-ol to 1-butanol.^7d The reaction was carried out by adding the nanoparticle catalyst with 1.15 × 10⁻³ moles of the reactant in a septum sealed flask. Subsequently, H₂ was purged into the flask for 5.00 min and then the system was filled with H₂ to a slight positive pressure. Aliquots of the solution were removed from the flask at different time intervals, centrifuged to remove the nanoparticles, and the products were analyzed by ¹H NMR. Using the peptide-capped materials, a TOF value of 1016.0 mol product (mol Pd × h)⁻¹ was observed, which is a two fold enhancement over the TOF value observed using commercially available monometallic Pd nanoparticles.^7d This increase in catalytic activity for the bimetallic nanocatalysts is directly due to the electronic effects associated with the two metals in the system, where peptide-based methods were used to guide the formation of the unique structure.

To expand peptide-based approaches to additional multicomponent catalysts, Slocik et al. demonstrated a versatile peptide-mediated synthesis of an integrated CdS–Pt nanoparticle system.^7c These materials were used to mimic the enzyme nitrate reductase for the catalytic reduction of nitrate to nitrite. The system is comprised of CdS quantum dots that are surface decorated by catalytic Pt nanoparticles, where the structure is achieved by using the cysteine modified Flg-A3 peptide (termed Flg-A3C). Here, the Flg-A3C sequence caps the growing CdS quantum dots via the thiol side chain of the cysteine residue. This then exposes the Flg-A3 portion of the sequence to solution to nucleate the Pt nanoparticles on the metal sulphide surface.^7c To that end, the peptide interface enforces the close proximity of the Pt nanoparticles to CdS for rapid electron transfer upon exposure of the system to light with a wavelength of 200–400 nm. The tyrosine residues of the interfacial peptides enable it to act as an electron donor for the trapped holes at the CdS surface eliminating the requirement of an external electron donor. TEM analysis of the composite materials revealed an average size of ∼1.5 nm for the Pt nanoparticles arranged around the CdS surface that resulted in the formation of chain-like structures,^7c which was unanticipated.

Once the composite structure was confirmed, the Pt–CdS materials were employed as catalysts for nitrate reduction using photochemical-based approaches.^7c From this analysis, the activity of peptide-fabricated Pt–CdS complex was observed to be 23 fold higher as compared to the enzyme driven system. Control experiments were conducted where a lower nitrate reduction activity was observed with CdS nanoparticles as compared to the CdS–Pt complex, but higher than that of the enzyme. Further experiments with the metallic Pt nanoparticles alone demonstrated no nitrate reduction capabilities. This suggests that the proximity of the Pt nanoparticles with the CdS surface plays a critical role in enhancing the reactivity. This could be further explained in terms of a chemical process where the electrons photoexcited inside the Pt penetrate the CdS surface and these trapped optically excited electrons get involved in the slow nitrate reduction reaction occurring at the nanocrystal surface.^7c Thus the peptide-based CdS/Pt materials serve the dual purpose by providing a versatile route for achieving high catalytic activity for nitrate reduction as well as eliminating the need of electron mediators, external electron donors, and even enzymes for certain applications.

4. Peptide scaffolds for non-spherical, linear nanocatalysts

While most materials directing peptides drive nanoparticle catalyst fabrication schemes using free peptides in solution, some sequences are able to self assemble to generate large scaffolds that are capable of acting as a template for the fabrication of non-spherical inorganic nanostructures.^3b,23 One such self-assembling peptide is the R5 sequence (SSKKSGSYSGSKGSKRRIL)³⁰ isolated from the diatom Cylindrothica fusiformis for the production of biosilica.^30a The R5 peptide can precipitate metal oxides such as SiO₂,³⁰ TiO₂,³¹ and TiP₂O₇,³² thus it possesses a unique materials directing capability. Furthermore, the R5 can self assemble in solution via the C-terminal RRIL motif, resulting in the formation of large peptide aggregates on the order of ∼825 nm that display a high localized concentration of amines.^30b These amines can bind and sequester metal ions, such as Pd²⁺, within the interior of the framework, which, upon reduction, results in the formation of zerovalent metallic nanostructures.^3b Based upon the amount of metal ions loaded within the scaffold, a variety of different shaped materials can be formed from spherical nanoparticles to linear nanoribbons and nanoparticle networks (NPNs), whose structure is controlled by the peptide branches.^3b As such, the R5 acts as a template for materials fabrication.

Fig. 6g presents the proposed mechanism for the formation of spherical and linear Pd nanostructures using the R5 as a biomolecular template.^3b The formation of the observed structures directly depends on the Pd : peptide ratio used in the materials fabrication process. Here, at low ratios, spherical structures are formed; however, as the ratio increases, linear and subsequently branched materials are generated.^3b,23 For instance, as shown in Fig. 6a, at a low Pd : peptide ratio of 60 (termed the Pd60 sample), spherical nanoparticles with a diameter of 2.9 ± 0.6 nm were observed. On increasing the ratio to 90 (Pd90, Fig. 6b), one-dimensional nanoribbons were detected with an average width of 3.9 ± 0.8 nm, while for the highest ratio of 120 (Pd120, Fig. 6c), dense and highly branched NPNs with an average width of 4.1 ± 1.2 nm were observed.^3b For higher ratios (>Pd120), precipitation of bulk Pd black was noted upon addition of NaBH₄, indicating that saturation of peptide template with Pd²⁺ occurred. High-resolution TEM (HR-TEM) images of the materials are presented in Fig. 6d–f where atomic structures associated with face center cubic (fcc) Pd are clearly evident. For instance, in the spherical Pd60 samples, nanoparticles with lattice fringes of 2.3 Å are noted that correspond to the Pd (111) fringe.^3b Imaging of the Pd120 NPNs demonstrated a polycrystalline structure with both Pd (111) fringes and hexagonal atomic packing, consistent with fcc Pd.^3b Furthermore, it is evident that interparticle necking is present to give rise to the linear, branching, and bulging points within the NPN structure. Based on these materials morphologies, it is likely that the peptide and metal loading control the final shape. At the lowest Pd : peptide ratio (Pd60), the generated Pd nanoparticles are well dispersed within the framework that prevent aggregation; however, as the ratio increases, more particles are contained within the template, thus diminishing the interparticle distance, leading to aggregation in a linear fashion.^3b As a result, the Pd materials are thus contained within the peptide framework, where both the metallic surface area and biological scaffold are likely to contribute to the catalytic functionality.

Fig. 6 Analysis of the Pd nanostructures fabricated using the R5 peptide template. Parts (a, b, and c) present low magnification TEM images of the Pd60, Pd90, and Pd120 materials, respectively, while parts (d, e, and f) present the corresponding HR-TEM image of the materials. Part (g) displays the synthetic scheme for the fabrication of different Pd nanostructures using the R5 peptide. The top path employs a low Pd : peptide ratio to generate Pd nanoparticles, while the bottom path uses a higher Pd : peptide ratio to fabricate Pd NPNs. (Reprinted with permission from R. Bhandari and M. R. Knecht, ACS Catal., 2011, 1, 89–98. Copyright 2011 American Chemical Society.)

To determine the catalytic functionality of the materials, two very different reactions were studied: Stille coupling and 4-nitrophenol reduction.^3b,23 These two reactions were selected as they follow drastically different pathways where Stille coupling likely employs an atom leaching process,¹⁹ while the reduction of 4-nitrophenol occurs directly on the Pd surface.^6a,33 Together, these two reactions can more fully probe both the catalytic functionality and the effects of composite structure, which can potentially be used to tune reaction rates and selectivity. For the Stille reaction, 4-iodobenzoic acid and PhSnCl₃ were reacted in the presence of the three different templated Pd nanostructures in water at room temperature over 24.0 h to produce biphenylcarboxylic acid. For all the Pd samples employed (Pd60, Pd90, and Pd120), quantitative product yields were obtained down to a Pd loading of 0.01 mol%, indicating a high degree of catalytic activity.^3b

To further probe the structure/catalytic function relationship of the materials, TOF analyses were conducted for all three systems using a variety of aryl halide substrates, as shown in Fig. 7. Using 4-iodobenzoic acid, the overall highest TOF values were observed, where TOFs of 452 ± 16 mol product (mol Pd × h)⁻¹, 334 ± 38 mol product (mol Pd × h)⁻¹, and 437 ± 13 mol product (mol Pd × h)⁻¹ were determined for the Pd60, Pd90, and Pd120 systems, respectively.^3b For this reaction, on average, the Pd60 nanospheres and Pd120 NPNs demonstrated increased TOFs as compared to the Pd90 system, suggesting that materials structural effects are likely modulating the reactivity. Beyond the structural effects, a shift in the chemical reactivity was also observed when the composition of the reagent was altered.²³ This difference in reactivity is likely due to the chemical structure, halide identity, and electronic effects of the catalyst/reagent complex. For instance, when using 3-iodobenzoic acid, a decrease in the TOF was observed as compared to 4-iodobenzoic acid, which could be attributed to the spacing of the functional groups. Due to the linear structure of the inorganic materials, both the halogen and carboxylic acid functional groups can potentially bind to the Pd surface, thereby partially inhibiting the oxidative addition step of the Stille reaction to lead to lower TOF values. Employing either 4-iodophenol or bromo-based substrates further decreased the TOF value due to changes in the electronic character and halogen identity of the reagent following known trends.^23,34 Interestingly, while decreased TOF values were observed for the different reagents, the same structural effects were evident for each reaction with higher TOFs for the Pd60 and Pd120 materials as compared to the Pd90 nanoribbons.

Fig. 7 Stille coupling TOF value comparison for the three Pd materials generated using the R5 template with multiple different aryl halides. (Reprinted with permission from R. Bhandari and M. R. Knecht, ACS Catal., 2011, 1, 89–98. Copyright 2011 American Chemical Society.)

The observed Stille reactivity trends for the Pd60, Pd90, and Pd120 materials are likely an effect of the composite structure, thus similar effects should be observed from different reactions. As such, the reduction of 4-nitrophenol was studied where the reaction occurs directly on the Pd surface.^23,33a The rate of the reaction was calculated by monitoring the decrease of the intensity of the 400 nm peak in the UV-vis spectrum corresponding to the 4-nitrophenol reagent employing pseudo first order kinetics. Such kinetics are justified as the concentration of the second reagent, NaBH₄, is in significant excess and considered to be constant throughout the reaction.^6a,35Fig. 8 presents the UV-vis analysis for the reduction of 4-nitrophenol using the Pd60 materials at 20.0 °C.²³ Here a rapid decrease in the 400 nm peak was observed, along with a linear increase in the 300 nm peak corresponding to the 4-aminophenol product, indicating a high degree of reactivity for these particles. Similar results were observed for the Pd90 nanoribbons and Pd120 NPNs with high catalytic capabilities for the surface-based reduction reaction. This reaction was subsequently studied for all three systems at temperatures between 10.0 °C and 60.0 °C from which pseudo first order rate constants (k) could be extracted using well-known methods.³⁵Fig. 8b plots the variation in the k values as a function of temperature for all three catalytic nanosystems. From the plot, it is evident that the rate constants are directly proportional to temperature with a linear increase in k values at higher temperatures. Furthermore, across the entire scale, higher rate constants are observed for both the Pd60 and Pd120 systems as compared to the Pd90 nanoribbons.²³ As such, the same trend is observed for the 4-nitrophenol reduction reaction as compared to Stille coupling where higher reactivity is observed for the Pd60 nanospheres and Pd120 NPNs over the Pd90 nanoribbons, further indicating that the composite structure of the materials is critically important to their overall catalytic functionality.

Fig. 8 Analysis of the 4-nitrophenol reduction reaction catalyzed by the Pd materials encapsulated within the peptide scaffold. Part (a) presents the time resolved UV-vis spectra from the reaction driven by the Pd60 nanoparticles at 20 °C where the absorbance at 400 nm arises from the substrate, while the absorbance that grows in at 300 nm is attributed to the 4-aminophenol product. Part (b) displays a plot for the k values for all three catalytic systems as a function of temperature. (Reprinted with permission from R. Bhandari and M. R. Knecht, ACS Catal., 2011, 1, 89–98. Copyright 2011 American Chemical Society.)

Taken together, the observed results for two very different reactions indicate higher reactivity for the spherical particles and NPNs as compared to the linear nanoribbons.^3b,23 Two critical structural factors of these biotemplated materials could give rise to these results: the catalytic surface area of particles and the reagent penetration depth within the peptide template, shown in Fig. 6g. Here, for the Pd60 spherical nanoparticles, the surface area is the highest as compared to the linear materials, but they are also highly dispersed within the peptide template leading to a higher reagent penetration depth required to reach and react at the metallic surface. For the Pd120 system, the NPNs are quite dense with lower surface area as compared to the nanospheres; however, these structures are formed at the highest Pd concentration per template, thus the catalytic materials are positioned very close to the peptide/water interface. To that end, their depth within the scaffold is significantly minimized as compared to the other structures to substantially diminish reagent diffusion effects for increased reactivity.^3b,23 As a result of the composite structure, one rate-determining factor is optimized for both the Pd60 and Pd120 systems, but both are diminished for the Pd90 system, thereby limiting the reactivity. Such capabilities were confirmed using control reactions that probed surface area effects and diffusion effects in solution;²³ however, these results demonstrate a unique template-based approach to bio-inspired nanocatalyst synthesis that could be used to tune the functionality and selectivity of the system, which is presently being studied.

5. Nanozymes

While most bionanocatalysts exploit the metallic nanoparticle to drive the reaction, the surface-bound ligands can also be designed as the catalytic component as enzyme mimics.^7c,36 Enzymes are catalytically active proteins widely recognized for their energy-saving mechanisms, catalyzing biological processes and chemical reactions under both in vivo and in vitro conditions.³⁷ Unfortunately, while enzymes are highly reactive and substrate-selective, they can be highly sensitive to their reaction environment. For instance, protein denaturation as a result of the solution dielectric and reaction temperature can minimize or even destroy all activity.³⁸ Mimicking and controlling enzymatic activities via nanotechnology can provide the ability to change, improve, and expand such catalytic processes to more advantageous reaction conditions.³⁹ This is achieved through the design of nanoparticle-based enzyme mimics, termed nanozymes. For these materials, nanoparticle ligands are engineered to reflect the active site structure of an enzyme along the metallic surface, where multiple reactions can simultaneously occur on one nanoparticle.^36a,40 As a result, this provides a colloidal dispersion of active sites throughout the solution to more quickly and efficiently catalyze large-scale reactions. Furthermore, the ligands and nanozyme motifs can be engineered to be reactive at elevated temperatures and non-aqueous conditions where enzyme activity would be unfavored.⁴¹ This capability to provide highly stable, multiple enzymatic sites on a single particle, combined with the power of colloidal suspension, can lead to catalytic processes that are optimized for difficult multi-step reactions.

To develop a nanozyme particle, highly specific ligands and their interactions with both the nanoparticle core and other ligands on the particle surface are crucial. Furthermore, it is potentially possible to include multiple different catalytic active sites on the nanoparticle surface, as selected for by the specific ligands, thus allowing for cascade-based processes to occur on a single particle. Synthesizing such materials with multiple enzyme-mimicking sites creates the environment for a substrate molecule to move from one site to the next without requiring solution mixing and/or Brownian motion to limit the overall rate of the cascade. This provides the potential to make catalytic processes faster and more accessible in a colloidal suspension. The most commonly used nanoparticles for nanozyme fabrication are Au monolayer protected clusters (MPCs).^36a,40,42 Such materials are employed due to their simple synthesis, well-developed thiol-based functionalization chemistries, high stability under a variety of stringent conditions, and their low toxicity.⁴³ Furthermore, thiols bound to Au MPCs are readily exchanged with secondary thiols in solution, which is controlled by equilibrium-based processes,⁴⁴ thus allowing for facile attachment of enzyme-mimicking ligands in selected ratios. This ease of synthesis, functionalization, and high versatility creates an ideal target for the design of multifunctional nanozyme systems.

A selection of nanozymes has been fabricated for a variety of enzyme mimicking capabilities. For instance, azacrown-functionalized Au nanoparticles were synthesized, due to the ligand coordination capabilities with Zn²⁺, as shown in Fig. 9a.⁴⁵ This was selected to mimic the Zn²⁺-bound porphyrin ring active site of many RNase-based or ribonuclease enzymes that are capable of cleaving phosphodiester bonds.⁴⁶ After addition and binding of Zn²⁺ to the surface ligands, these materials were shown to be functionally similar to the target RNases, cleaving the phosphate ester 2-hydroxypropyl p-nitrophenyl phosphate (HPNP) as a model RNA-like compound. The HPNP model was chosen because of its standard use and ease of characterization after cleavage.⁴⁵ The reactivity was studied in a methanol–water (6 : 4) solution, in which the Zn²⁺ containing nanozymes are soluble in the pH range of 4.5–7.2. The reaction was monitored by UV–vis for the formation of 2,4-dinitrophenolate, the product, at 400 nm at 25 °C. Kinetic analysis of the cleavage reaction demonstrated that the greatest activity arises from particles where the azacrown moieties were fully loaded with Zn²⁺.⁴⁵ As a control comparison, free ligands were used as monomeric catalysts in solution, showing lower activity than the nanozyme system.⁴⁵ This suggests that the multimeric structure of the nanozyme surface facilitates a high degree of reactivity, as compared to monomeric species in solution. This is highly important for catalytic activity, especially where cascade-based pathways are designed on a single particle surface.

Fig. 9 Ligand structural attributes of Au nanozymes functionalized with (a) azacrown groups, (b) Cu²⁺-based ligands, and (c) amino acids for catalytic activity as enzyme mimics. (Image 9b reprinted with permission from Springer Science + Business media, Journal of Biological Inorganic Chemistry, Self-assembled gold nanocrystal micelles act as an excellent artificial nanozyme with ribonuclease activity, 14, 2009, 653–662.)

Beyond Zn-based inorganic active sites in RNAse enzymes, a second common metal is Cu.⁴⁶ As such, Cu²⁺ was incorporated onto a Au nanoparticle surface via the binding of the Cu-specific N1,N1-bis(2-aminoethyl)-N2-dodecylethane-1,2-diamine ligand, which coordinates the metal ion (Fig. 9b).^36a As a result, the Au nanoparticle surface mimics the active site structure of other, non-Zn containing RNases that depend upon a Cu²⁺ active site. This unique Cu-binding ligand is incorporated onto the Au nanozyme surface via an alkyl tail to expose the Cu-based species to the reaction solution. When Cu²⁺ ions were introduced to the nanoparticles, a ligand to metal charge transfer band was observed for the metal ion binding the target ligand, demonstrating that complexation had occurred.^36a Once the active structure is fully formed, the materials demonstrated a high degree of enzyme-like activity via the hydrolysis of the phosphodiester bond in HPNP. For the Au nanozymes, a k_cat value of 2.1 × 10⁻³ s⁻¹ was observed, which was four orders of magnitude greater than the k_cat achieved for the free ligand (2.8 × 10⁻⁷ s⁻¹).^36a Further studies of the Cu loading on the nanoparticle surface demonstrated that an optimal concentration exists, such that electrostatic effects and lack of competition of close-proximity Cu²⁺ ions is required for maximum efficiency.^36a

While non-biological ligands such as azacrowns can be used to generate nanozymes, combinations of amino acids on nanoparticle surfaces can also be used to mimic enzyme active sites for catalytic reactivity.^40,42 A specific example of this effect was a histidine-phenylalanine dimer functionalized nanozyme where the peptide was coordinated to the surface through an alkyl-thiol-based ligand, as shown in Fig. 9c.⁴² This combination of amino acids was selected as a mimic of the catalytic active site of aspartic esterase for cleaving biological esters.⁴⁷ Catalytic activity of the nanozyme was tested via the cleavage of the ester bond in 2,4-dinitrophenylbutanoate (DNPB). The cleavage of dinitrophenyl from the ester group is easily followed spectrophotometrically, and can thus be readily quantitated. Using this nanozyme system at a reaction temperature of 25 °C in a 10% CH₃OH/H₂O solution, hydrolysis of DNPB was monitored with respect to changes in pH. Over the pH range of 3–10, the dipeptide-functionalized Au nanozyme showed a significantly higher reactivity (second order rate constant = 38.8 M⁻¹ s⁻¹) as compared to free Ac-His-Phe-OH dipeptide in solution, which is not shown to be reactive.⁴² The enhanced reactivity of the peptide-based nanozyme was likely due to the strong nucleophilic contribution for both the imidazole and carboxylate groups in the overall catalytic reaction that was enhanced due to the stabilizing factor of the nanoparticle. Of particular interest, the di-amino acid functionalized Au nanozyme is able to catalytically cleave esters with optimum efficiency below pH 5, where aspartic esterases begin to lose catalytic activity due to the sensitivity of the enzyme structure to the reaction medium.⁴² As a result, efficient bio-based enzymatic activity could be translated to non-biological conditions for desirable catalytic functionality.

Beyond amino acid dimers, attachment of short peptide chains to the particle surface can also be used to mimic enzyme active sites.⁴⁰ Note that it is important that the selected sequence have minimal to no interactions with the particle surface, such as that observed with the materials binding peptides described above. This is important such that the enzyme mimicking peptide is extended from the surface in a linear fashion to generate the desired catalytic activity.⁴⁰ This linearity is achieved through the attachment of an alkyl thiol group at either the N- or C-termini of the sequence, thus preferential binding to the particle occurs via the thiol moiety. Using this structural motif, a nanozyme has been developed using a dodecamer sequence (LGYKAHFAGRGR) that simulates carboxyl esterase activity, which successfully cleaves the ester bond in DNPB.⁴⁰ The peptide sequence was selected and designed based upon the side chain functionalities for catalysis, with arginine groups and a free carboxylate, as well as a histidine surrounded by hydrophobic moieties.⁴⁰ This combination of basic and acidic groups has the potential to catalytically cleave esters. The preliminary dodecapeptide showed catalytic improvement at a pH of 10, suggesting a base-catalyzed system due to the high pK_a residues.⁴⁰ This contrasts with the dimer peptide functionalized nanozyme discussed above that was optimized for efficiency at acidic pH. This further expands enzyme-like activity to additional reaction conditions, which is important for reaction design.

By controlling the surface structure of nanoparticles, nanozymes have incorporated catalytic activity without requiring the inorganic core to participate in the reaction. By further tuning these nanozymes, their catalytic ability could potentially surpass the already high activity and selectivity of enzymes while functioning over a more broad range of pH and temperature. Their higher stability, shelf life, homogenous distribution, and ease of synthesis make them ideal candidates for catalysis. While the development of nanozymes is a new direction, these materials will likely continue to develop and expand such that catalysis and the mimicking of enzymatic activity are potentially controllable and tuneable to meet specific catalytic needs on the nanoscale level. Furthermore, by combining the catalytic effects of both the ligands and the metallic core, a highly efficient catalytic system could potentially be realized for multistep pathways that incorporate both catalytic and enzymatic functionalities.

6. Conclusions

It is clear that new methods are required to overcome the energy-consumptive catalytic technologies currently in use. By combining the efficiency of biological systems with the size advantages of nanomaterials, highly efficient and energy-neutral catalytic technologies are being realized where peptide-based materials have shown remarkable potential. While this research is in an early phase of development, the future remains bright that bio-inspired structures could yield new directions for efficient catalytic materials. Additional research is certainly required, especially at the peptide/materials interface. A reference table that includes all of the peptides discussed here is available in the ESI.† By controlling the arrangement and materials specificity of the individual peptide sequences, new materials could be realized where the peptide mediates the size, shape, composition, structure, and catalytic functionality of the materials. Many international laboratories are currently conducting such research where the results are showing significant promise.

Acknowledgements

This material is based upon work supported by the National Science Foundation under Grants No. CBET-1033334 and DMR-1145175. Furthermore, acknowledgement is made to the donors of the American Chemical Society Petroleum Research Fund for partial support of this research.

Notes and references

1. U.S. Department of Energy, Basic Research Needs: Catalysis for Energy, 2007.
2. (a) D. Astruc, Inorg. Chem., 2007, 46, 1884–1894; (b) V. S. Myers, M. G. Weir, E. V. Carino, D. F. Yancy, S. Pande and R. M. Crooks, Chem. Sci., 2011, 2, 1632–1646.
3. (a) R. Coppage, J. M. Slocik, M. Sethi, D. B. Pacardo, R. R. Naik and M. R. Knecht, Angew. Chem., Int. Ed., 2010, 49, 3767–3770; (b) A. Jakhmola, R. Bhandari, D. B. Pacardo and M. R. Knecht, J. Mater. Chem., 2010, 20, 1522–1531.
4. M. De bruyn and R. Neumann, Adv. Synth. Catal., 2007, 349, 1624–1628.
5. (a) H. Ye, R. W. J. Scott and R. M. Crooks, Langmuir, 2004, 20, 2915–2920; (b) O. M. Wilson, R. W. J. Scott, J. C. Garcia-Martinez and R. M. Crooks, J. Am. Chem. Soc., 2004, 127, 1015–1024.
6. (a) S. Behrens, A. Heyman, R. Maul, S. Essig, S. Steigerwald, A. Quintilla, W. Wenzel, J. Bürck, O. Dgany and O. Shoseyov, Adv. Mater., 2009, 21, 3515–3519; (b) A. Housni, M. Ahmed, S. Liu and R. Narain, J. Phys. Chem. C, 2008, 112, 12282–12290; (c) T. Yang, Z. Li, L. Wang, C. Guo and Y. Sun, Langmuir, 2007, 23, 10533–10538.
7. (a) R. Coppage, J. M. Slocik, B. D. Briggs, A. I. Frenkel, H. Heinz, R. R. Naik and M. R. Knecht, J. Am. Chem. Soc., 2011, 133, 12346–12349; (b) D. B. Pacardo, M. Sethi, S. E. Jones, R. R. Naik and M. R. Knecht, ACS Nano, 2009, 3, 1288–1296; (c) J. M. Slocik, A. O. Govorov and R. R. Naik, Angew. Chem., Int. Ed., 2008, 47, 5335–5339; (d) J. M. Slocik and R. R. Naik, Adv. Mater., 2006, 18, 1988–1992.
8. (a) L. J. Rothberg and E. M. Nelson, Langmuir, 2011, 27, 1770–1777; (b) G. Borghs, Y. N. Cheng, M. Wang and H. Z. Chen, Langmuir, 2011, 27, 7884–7891.
9. M. B. Dickerson, K. H. Sandhage and R. R. Naik, Chem. Rev., 2008, 108, 4935–4978.
10. S. Mann, Biomineralization: Principals and Concepts in Bioinorganic Materials Chemistry, Oxford University Press, New York, 2002.
11. D. E. Morse and R. L. Brutchey, Angew. Chem., Int. Ed., 2006, 45, 6564–6566.
12. (a) R. R. Naik, S. J. Stringer, G. Agarwal, S. E. Jones and M. O. Stone, Nat. Mater., 2002, 1, 169–172; (b) S.-W. Lee, C. Mao, C. E. Flynn and A. M. Belcher, Science, 2002, 296, 892–895; (c) S. R. Whaley, D. S. English, E. L. Hu, P. F. Barbara and A. M. Belcher, Nature, 2000, 405, 665–668.
13. J. M. Slocik, M. O. Stone and R. R. Naik, Small, 2005, 1, 1048–1052.
14. C. B. Mao, D. J. Solis, B. D. Reiss, S. T. Kottmann, R. Y. Sweeney, A. Hayhurst, G. Georgiou, B. Iverson and A. M. Belcher, Science, 2004, 303, 213–217.
15. (a) L. F. Drummy, S. E. Jones, R. B. Pandey, B. L. Farmer, R. A. Vaia and R. R. Naik, ACS Appl. Mater. Interfaces, 2010, 2, 1492–1498; (b) M. B. Dickerson, S. E. Jones, Y. Cai, G. Ahmad, R. R. Naik, N. Kröger and K. H. Sandhage, Chem. Mater., 2008, 20, 1578–1584; (c) G. Ahmad, M. B. Dickerson, Y. Cai, S. E. Jones, E. M. Ernst, J. P. Vernon, M. S. Haluska, Y. Fang, J. Wang, G. Subramanyam, R. R. Naik and K. H. Sandhage, J. Am. Chem. Soc., 2007, 130, 4–5.
16. (a) R. L. Brutchey and D. E. Morse, Chem. Rev., 2008, 108, 4915–4934; (b) J. L. Sumerel, W. Yang, D. Kisailus, J. C. Weaver, J. H. Choi and D. E. Morse, Chem. Mater., 2003, 15, 4804–4809.
17. (a) Y. J. Lee, H. Yi, W. J. Kim, K. Kang, D. S. Yun, M. S. Strano, G. Ceder and A. M. Belcher, Science, 2009, 324, 1051–1055; (b) K. T. Nam, D. W. Kim, P. J. Yoo, C. Y. Chiang, N. Meethong, P. T. Hammond, Y. M. Chiang and A. M. Belcher, Science, 2006, 312, 885–888.
18. J. M. Slocik, J. S. Zabinski, D. M. Phillips and R. R. Naik, Small, 2008, 4, 548–551.
19. D. B. Pacardo, J. M. Slocik, K. C. Kirk, R. R. Naik and M. R. Knecht, Nanoscale, 2011, 3, 2194–2201.
20. (a) C.-L. Chen, P. J. Zhang and N. L. Rosi, J. Am. Chem. Soc., 2008, 130, 13555–13557; (b) C.-L. Chen and N. L. Rosi, J. Am. Chem. Soc., 2010, 132, 6902–6903.
21. M. Sethi, D. B. Pacardo and M. R. Knecht, Langmuir, 2010, 26, 15121–15134.
22. (a) C. Y. Chiu, Y. J. Li, L. Y. Ruan, X. C. Ye, C. B. Murray and Y. Huang, Nat. Chem., 2011, 3, 393–399; (b) Y. J. Li and Y. Huang, Adv. Mater., 2010, 22, 1921–1925; (c) L. Ruan, C.-Y. Chiu, Y. Li and Y. Huang, Nano Lett., 2011, 11, 3040–3046; (d) L. M. Forbes, A. P. Goodwin and J. N. Cha, Chem. Mater., 2010, 22, 6524–6528.
23. R. Bhandari and M. R. Knecht, ACS Catal., 2011, 1, 89–98.
24. R. B. Pandey, H. Heinz, J. Feng, B. L. Farmer, J. M. Slocik, L. F. Drummy and R. R. Naik, Phys. Chem. Chem. Phys., 2009, 11, 1989–2001.
25. A. C. Templeton, S. W. Chen, S. M. Gross and R. W. Murray, Langmuir, 1999, 15, 66–76.
26. A. K. Diallo, C. Ornelas, L. Salmon, J. R. Aranzaes and D. Astruc, Angew. Chem., Int. Ed., 2007, 46, 8644–8648.
27. (a) M. A. El-Sayed, Acc. Chem. Res., 2001, 34, 257–264; (b) C. Burda, X. Chen, R. Narayanan and M. A. El-Sayed, Chem. Rev., 2005, 105, 1025–1102; (c) B. Wiley, T. Herricks, Y. Sun and Y. Xia, Nano Lett., 2004, 4, 1733–1739.
28. (a) N. Toshima and T. Yonezawa, New J. Chem., 1998, 22, 1179–1201; (b) R. W. J. Scott, O. M. Wilson, S.-K. Oh, E. A. Kenik and R. M. Crooks, J. Am. Chem. Soc., 2004, 126, 15583–15591; (c) C.-H. Chen, L. S. Sarma, J.-M. Chen, S.-C. Shih, G.-R. Wang, D.-G. Liu, M.-T. Tang, J.-F. Lee and B.-J. Hwang, ACS Nano, 2007, 1, 114–125; (d) M. H. Chowdhury, S. Chakraborty, J. R. Lakowicz and K. Ray, J. Phys. Chem. C, 2011, 115, 16879–16891; (e) R. W. J. Scott, A. K. Datye and R. M. Crooks, J. Am. Chem. Soc., 2003, 125, 3708–3709.
29. N. Toshima, M. Kanemaru, Y. Shiraishi and Y. Koga, J. Phys. Chem. B, 2005, 109, 16326–16331.
30. (a) N. Kröger, R. Deutzmann and M. Sumper, Science, 1999, 286, 1129–1132; (b) M. R. Knecht and D. W. Wright, Chem. Commun., 2003, 3038–3039.
31. S. L. Sewell and D. W. Wright, Chem. Mater., 2006, 18, 3108–3113.
32. K. E. Cole, A. N. Ortiz, M. A. Schoonen and A. M. Valentine, Chem. Mater., 2006, 18, 4592–4599.
33. (a) N. Pradhan, A. Pal and T. Pal, Langmuir, 2001, 17, 1800–1802; (b) J. Zeng, Q. Zhang, J. Chen and Y. Xia, Nano Lett., 2009, 10, 30–35.
34. J. K. Stille, Angew. Chem., Int. Ed. Engl., 1986, 25, 508–524.
35. Z. V. Feng, J. L. Lyon, J. S. Croley, R. M. Crooks, D. A. Vanden Bout and K. J. Stevenson, J. Chem. Educ., 2009, 86, 368–372.
36. (a) Z. M. Zhang, Q. A. Fu, X. Q. Li, X. Huang, J. Y. Xu, J. C. Shen and J. Q. Liu, JBIC, J. Biol. Inorg. Chem., 2009, 14, 653–662; (b) D. Kisailus, M. Najarian, J. C. Weaver and D. E. Morse, Adv. Mater., 2005, 17, 1234–1239.
37. H. Vangelderen, J. M. Mayer and B. Testa, Biochem. Pharmacol., 1994, 47, 753–756.
38. G. A. Petsko, Nature, 1999, 401, 115–116.
39. M. E. Davis, A. Katz and W. R. Ahmad, Chem. Mater., 1996, 8, 1820–1839.
40. L. Pasquato, P. Pengo and P. Scrimin, Supramol. Chem., 2005, 17, 163–171.
41. X. Yuan, M. Iijima, M. Oishi and Y. Nagasaki, Langmuir, 2008, 24, 6903–6909.
42. P. Pengo, S. Polizzi, L. Pasquato and P. Scrimin, J. Am. Chem. Soc., 2005, 127, 1616–1617.
43. (a) A. M. Brynskikh, Y. L. Zhao, R. L. Mosley, S. Li, M. D. Boska, N. L. Klyachko, A. V. Kabanov, H. E. Gendelman and E. V. Batrakova, Nanomedicine, 2010, 5, 379–396; (b) M.-C. Daniel and D. Astruc, Chem. Rev., 2004, 104, 293–346.
44. M. J. Hostetler, A. C. Templeton and R. W. Murray, Langmuir, 1999, 15, 3782–3789.
45. F. Manea, F. B. Houillon, L. Pasquato and P. Scrimin, Angew. Chem., Int. Ed., 2004, 43, 6165–6169.
46. S. J. Lippard and J. M. Berg, Principals of Bioinorganic Chemistry, University Science Books, Mill Valley, California, 1994.
47. D. B. Northrop, Acc. Chem. Res., 2001, 34, 790–797.

---

Footnotes

† Electronic supplementary information (ESI) available: Table of peptides. See DOI: 10.1039/c1cy00350j

‡ These authors contributed equally to this work.

---