Functional nanoparticle architectures for sensoric, optoelectronic, and bioelectronic applications*

Tailored sensoric, electronic, photoelectrochemical, and bioelectrocatalytic functions can be designed by organized molecular or biomolecular nanoparticle hybrid configurations on surfaces. Layered receptor-cross-linked Au nanoparticle assemblies on electrodes act as specific sensors of tunable sensitivities. Layered DNA-cross-linked CdS nanoparticles on electrode supports reveal organized assemblies of controlled electronic and photoelectrochemical properties. Au nanoparticle-FAD semisynthetic cofactor units are reconstituted into apo-glucose oxidase (GOx) and assembled onto electrodes. The resulting enzymes reveal effective electrical contacting with the electrodes, and exhibit bioelectrocatalytic functions toward the oxidation of glucose to gluconic acid. Magneto-switchable electrocatalysis and bioelectrocatalysis are accomplished by the surface modification of magnetic particles with redox-relay units. By the attraction of the modified magnetic particles to the electrode support, or their retraction from the electrode, by means of an external magnet, the electrochemical functions of the magnetic particle-tethered relays can be switched between “ON” and “OFF” states, respectively. The magneto-switchable redox functionalities of the modified particles activate electrocatalytic transformations, such as a biocatalytic chemoluminescence cascade that leads to magneto-switchable light emission or the activation of bioelectrocatalytic processes. Capping of nanoparticles with molecular monolayers or thin films may introduce functional units such as recognition sites, catalytic elements, redox-active groups or bioactive components [1]. The surface modifier of the particles may then be used to couple the unique electronic, photonic, or catalytic properties of quantum-size nanoparticles with molecular or macromolecular functionalities to yield hybrid systems of new features [2]. For example, nucleic acid-functionalized Au nanoparticles, exhibiting a characteristic red color originating from the single-particle plasmon exciton, that are complementary to the two ends of a target DNA, turn blue upon hybridization with the target DNA due to the formation of an interparticle-coupled plasmon exciton [3]. Quantization of the double-layer capacitance charging of alkanethiol-functionalized Au nanoparticles is a property of the monolayer capping and nanoparticle core hybrid system. The capacitance generated by the ionic space charge upon the charging of the metal core was found to be controlled by the thickness of the dielectric medium of the monolayer, and the charging of nanocapacitors by single electrons occurs at quantized potential intervals (∆V = e/C), determined by the nanoparticle hybrid composition [4]. *Pure Appl. Chem. 74, 1489–1783 (2002). An issue of reviews and research papers based on lectures presented at the 2nd IUPAC Workshop on Advanced Materials (WAM II), Bangalore, India, 13–16 February 2002, on the theme of nanostructured advanced materials. ‡Corresponding author Alternatively, the functionalization of nanoparticles may allow the assembly of surface-confined architectures or patterns [5]. The present account outlines recent advances, originating from our laboratory, that demonstrate the surface modification of metallic, semiconductor, and magnetic particles with functional molecular or biomolecular units, and the use of these hybrid structures as key elements to construct novel sensoric, electronic, optoelectronic, and bioelectronic systems. RECEPTOR-CROSS-LINKED Au NANOPARTICLES ON ELECTRODES FOR SENSORIC APPLICATIONS A layer-by-layer assembly process of citrate-capped Au nanoparticles (12 ± 1 nm) on indium tin oxide (ITO) glass surfaces was developed [6], Scheme 1. First, a transparent ITO conductive glass support is functionalized with a 3-aminopropylsiloxane thin film. Electrostatic binding of the negatively charged citrate-capped Au nanoparticles, followed by the electrostatic association of the oligocationic bisparaquat-p-phenylene (1), yields the first nanoparticle layer. By a subsequent stepwise treatment of the surface with the Au nanoparticles and (1), a controlled number of Au nanoparticles associated with the surface are constructed. Figure 1A shows the absorption spectra of the system upon the build-up of the I. WILLNER AND B. WILLNER © 2002 IUPAC, Pure and Applied Chemistry 74, 1773–1783 1774 Scheme 1 Assembly of a layered receptor-cross-linked Au nanoparticle array for the electrochemical sensing of π-donor substrates. Fig. 1 Characterization of the (1)-cross-linked Au nanoparticle array upon the build-up of one to five layers (a) to (e): (A) the absorption spectra of the arrays; (B) cyclic voltammograms of the (1)-cross-linking units, recorded in 0.1 M phosphate buffer, pH = 7.2, under Ar, scan rate 100 mV⋅s–1; (C) cyclic voltammograms corresponding to the Au nanoparticles, recorded in 1.0 M H2SO4, under Ar, scan rate 50 mV⋅s –1. nanoparticle layers. In addition to the increase in the band at λ = 520 nm, characteristic of the plasmon absorbance of the Au nanoparticles, a new absorbance, λ = 650 nm is observed. The latter absorbance band is attributed to an interparticle-coupled plasmon exciton that originates from the clustering of the particles on the surface. As the number of aggregated particles is higher, the probability for such coupled plasmon excitations increases. Figure 1B shows the cyclic voltammograms of the bis-(paraquat)p-phenylene oligocationic cross-linker (1) units, observed upon the construction of the layered assembly. The electrical response of (1) increases almost linearly upon the build-up of the assembly, implying a three-dimensional conductivity of the array. By coulometric assay of the reduction wave of (1), the average surface coverage of (1) per layer is estimated to be 1.5 × 10–11 mole⋅cm–2. Figure 1C shows the electrical response of the Au nanoparticles (in 1.0 M H2SO4). These redox waves correspond to the reduction of the Au oxide layer and the surface reoxidation, respectively. The redox response of the Au nanoparticles increases linearly with the number of particle layers, and by coulometric assay of the reduction waves, we estimate that the average surface coverage of the nanoparticles per layer corresponds to 0.8 × 1011 particles⋅cm–2. Knowing the surface coverage per layer of (1), we find that ca. 100 oligocationic cross-linking units of (1) are associated with each particle. The molecular oligocationic cross-linker (1) exhibits π-acceptor properties, and π-donor substrates of appropriate size can bind to the cavity of (1). The resulting supramolecular complex is stabilized by π-donor–acceptor interactions. Formation of such supramolecular complexes on the nanoengineered three-dimensional array of nanoparticles would concentrate the π-donor substrate at the electrode surface, and enhanced sensitivity would be accomplished. Also, the sensitivity of the functional electrodes would be controlled by the number of nanoparticle layers associated with the conductive support. Furthermore, the three-dimensional conductivity of the array suggests that the π-donor encapsulated in the π-acceptor receptor units could be sensed electrochemically, provided that the π-donor compound is redox-active. Indeed, a series of π-donor substrates such as hydroquinone (2), dopamine (3), and adrenaline (4) was electrochemically sensed by such functional electrodes [6]. Figure 2 shows the cyclic voltammograms of the functional electrode consisting of 5 layers of (1)-cross-linked Au nanoparticles upon the analysis of different concentrations of adrenaline (4). The electrical response of (4) increases as the bulk concentration is elevated, and Figure 2 (inset) depicts the derived calibration curve. It should be noted that an analogous 5-layer array of Au nanoparticles cross-linked by N,N′-dimethyl-4,4′-bipyridinium dichloride (5), that was assembled on an ITO electrode, is insensitive toward the electrochemical detection of (4) in this concentration range. The dicationic cross-linker (5), exhibits substantially lower affinity to form a π-donor–acceptor complex with (4). Thus, this control experiment reveals that the successful electrochemical sensing of (4) by the nanoparticle-modified electrode, does not originate from the increase in the electrode surface area due to the formation of a rough© 2002 IUPAC, Pure and Applied Chemistry 74, 1773–1783 Functional nanoparticle architectures 1775 Fig. 2 Cyclic voltammograms corresponding to the analysis of adrenaline (4) by a 5-layer (1)-cross-linked array of Au nanoparticles. Concentration of (4) corresponds to: (a) 1 × 10–5 M; (b) 2 × 10–5 M; (c) 4 × 10–5 M; (d) 6 × 10–5 M; (e) 8 × 10–5 M. Data recorded under Ar in 0.1 M phosphate buffer, pH = 7.2, scan rate 100 mV⋅s–1. Inset: Calibration curve corresponding to the amperometric response of the array (at E° = –0.26 V vs. SCE) at variable concentrations of (4). ened surface, but rather from the specific concentration of (4) in the π-acceptor receptor sites associated with the functional electrode. ELECTRICAL CONTACTING OF ENZYMES BY SINGLE Au NANOPARTICLE ARCHITECTURES Redox-enzymes lack electron-transfer communication with electrodes. The barrier for electrical contacting of the enzymes with the electrodes may be explained by the Marcus equation, eq. 1, that defines the parameters controlling the electron-transfer rate between a donor–acceptor pair (∆G° and λ are the free-energy change and reorganization energy accompanying the electron-transfer process, d and do correspond to the actual distance separating the donor–acceptor pair and the van der Waals distance between the donor and acceptor, respectively, and β is the electron coupling constant). ket ∝ exp[–β(d – do)] ⋅ exp[(∆G° + λ) 2 / 4RTλ] (1) One may consider the redox-site in proteins and the electrode as a donor–acceptor pair. As the average size of proteins is ca. 40–120 Å, the redox site is embedded in a matrix that spatially separates the donor and acceptor components, leading to the lack of electron-transfer communication between the components. Electrical contacting of redox-enzymes with electrodes was achieved by tethering redoxrelays to the proteins, or the immobilization of the enzymes in redox-polymers. Although in these systems the bioelectrocatalytic functions of the enzymes are activated, the electron-transfer turnover rates between the enzymes and the electrode are far slower than the electron-transfer turnover rates with their native enzymes. We have reported that the alignment of redox proteins on surfaces by the surface reconstitution of an apo-flavoenzyme on an electron-relay/flavin adenine dinucleotide (FAD) monolayer, leads to an electrically contacted bioelectrocatalyst with an efficient electron-transfer turnover [7]. Recently, we found that a single Au nanoparticle that is nanoengineered into the protein acts as a nanoelectrode for the contacting of the redox site with macroscopic electrodes [8]. A Au nanoparticle (1.3 ± 0.2 nm) functionalized with a single N-succinimidyl active ester group is reacted with N6-(2-aminoethyl)-flavin adenine dinucleotide, NH2-FAD (6), Scheme 2. Apo-glucose oxidase, apoGOx, is then reconstituted with the cofactor-functionalized Au nanoparticles, and the resulting modified protein is linked to a bulk Au electrode functionalized with bis-p-xylenedithiol. Microgravimetric quartz-crystal microbalance measurements indicate that the surface coverage of the enzyme is ca. 1 × 10–12 mole⋅cm–2. Figure 3 shows the cyclic voltammograms of nanoparticle-functionalized enzyme electrodes at different concentrations of glucose. The resulting bioelectrocatalytic anodic current indiI. WILLNER AND B. WILLNER © 2002 IUPAC, Pure and Applied Chemistry 74, 1773–1783 1776 Scheme 2 Reconstitution of apo-GOx with an FAD-functionalized Au nanoparticle and the assembly of the reconstituted enzyme as a monolayer on an Au electrode. cates that the redox enzyme is electrically contacted with the electrode support. Control experiments reveal that a Au electrode modified with an FAD monolayer on which apo-glucose oxidase is reconstituted, yields an enzyme-electrode that lacks electrical communication with the electrode. Thus, the single Au nanoparticle attached to the FAD site of the enzyme acts as a nanoelectrode for the electrical contacting of the redox enzyme with the bulk electrode. A calibration plot that corresponds to the amperometric responses of the functional electrode at different glucose concentrations was derived. Knowing the surface coverage of the enzyme, and the maximum output current of the system, the electron-transfer turnover rate is estimated to be 2300 s–1. This unprecedented efficient electron-transfer communication between the redox protein and the electrode originates from the mediated electron transfer by a single Au nanoparticle that is conjugated to the protein assembly. This effective electrontransfer communication between the nanoengineered redox enzyme and the electrode has important consequences on the sensitivity and selectivity of the enzyme electrode. Glucose concentrations as low as 30 μM are sensed by the system, and the enzyme-electrode is not affected by common glucose-sensing interferants such as ascorbic acid, uric acid, or oxygen. NANOPARTICLE-FUNCTIONALIZED DNA STRUCTURES ON SURFACES Research in the area of nanoparticle-coupled DNA electronics attracts scientific efforts in two general directions: (i) The unique electronic and optical properties of nanoparticles may be used to generate labels of controllable properties. Indeed, metal nanoparticles or nanorods modified with nucleic acids, were used as functional labels for the sensing of DNA [9]. (ii) DNA may be used as a template for the generation of metallic or semiconductive nanostructures [10]. The possibility of synthesizing DNA of predesigned length and shape, the ability to “cut and paste” nucleic acids with appropriate biocatalysts, and the versatile chemical means to functionalize nucleic acids turn DNA into an attractive mold for the deposition of nanostructures [9]. Indeed, a primary study has demonstrated the generation of Ag or Pd wires on DNA templates [10]. We have used nucleic acid-functionalized CdS nanoparticles (2.6 ± 0.4 nm) to construct aggregated arrays of DNA [11], Scheme 3. The CdS particles were functionalized with the thiolated nucleic acids (7) and (8) that are complementary to the DNA (9) (ca. 20–24 nucleic acid residues are associated with each CdS nanoparticle). Scheme 3A shows the assembly of the layered aggregated CdS array on a glass support. Covalent coupling of (7) to the maleimide-functionalized siloxane film associated with the glass support yields the active interface for the assembly of the CdS nanoparticle structure. Hybridization of (9) with the surface is followed by the hybridization of the (8)-functionalized CdS nanoparticles to the single-stranded end of (9). The subsequent reaction of the first layer of CdS parti© 2002 IUPAC, Pure and Applied Chemistry 74, 1773–1783 Functional nanoparticle architectures 1777 Fig. 3 Cyclic voltammograms of the FAD-Au nanoparticle-reconstituted GOx electrode in the presence of variable concentrations of glucose corresponding to: (a) 0 M; (b) 1 × 10–3 M; (c) 1 × 10–2 M; (d) 2 × 10–2 M; (e) 5 × 10–2 M. Data recorded under Ar, 0.1 M phosphate buffer, pH = 7.0, scan rate 5 mV⋅s–1.