Surface Modification of ZnO Nanorods with Hamilton Receptors

A new prototype of a Hamilton receptor suitable for the functionalization of inorganic nanoparticles was synthesized and characterized. The hydrogen bonding receptor was coupled to a catechol moiety, which served as anchor group for the functionalization of metal oxides, in particular zinc oxide. Synthesized zinc oxide nanorods [ZnO] were used for surface functionalization. The wet-chemical functionalization procedure towards monolayer-grafted particles [ZnO-HR] is described and a detailed characterization study is presented. In addition, the detection of specific cyanurate molecules is demonstrated. The hybrid structures [ZnO-HR-CA] were stable towards agglomeration and exhibited enhanced dispersability in apolar solvents. This observation, in combination with several spectroscopic experiments gave evidence of the highly directional supramolecular recognition at the surface of nanoparticles.


Introduction
Supramolecular mediated self-assembly of individual building blocks represents a powerful method for both controlling implementation of organic and inorganic nanomaterials into hierarchically ordered 3D structures and generation of new nano-hybrid materials [1][2][3]. The aim to benefit from properties of thermodynamically unstable inorganic nanoparticles and to influence their characteristics through surface modifications is determined by stabilization of nano-sized matter [4]. A dense coverage of the surface through chemical functionalization leads to inhibition of Ostwald ripening [5]. However chemical transformations based on substitution reactions of covalently bound ligands on the particle surface can be problematic because desorption of the initial ligand creates a defect in the shell providing highly reactive surfaces, which can promote irreversible changes of the nanoparticles like agglomeration and growth [6]. A combination of stable covalent surface attachment with the possibility of reversible modification of particle properties is accomplished by the concept of supramolecular self-assembly of surface bound ligands with functionalities [7,8] Such an approach combines the advantages of both, avoiding the tendency of nanoparticles to agglomerate with the potential of reversible attachment of specific guests towards the generation of hybrid architectures. Among other supramolecular forces, dynamic hydrogen bonding-based systems are used for the non-covalent association of individual modules [9]. Such strategies involve the formation of ordered nanoparticle aggregates employing polymers [10], biomacromolecules, such as DNA [11] and proteins [12]. A well-established artificial hydrogen bonding motif is the Hamilton receptor (Figure 1), which can easily and strongly bind cyanuric and barbituric acid derivatives due to six complementary hydrogen bonding interactions [13]. In such a highly specific, rigid and directional host-guest complex association constants range between 10 3 and 10 6 M −1 in apolar solvents [14,15]. Thus, many supramolecular architectures employing this key-lock principle were investigated, including self-assembled dendrimers [16][17][18], polymers [19,20], self-assembled monolayers [21,22], dynamic systems [23,24] and charge-transfer architectures [25][26][27][28][29]. Recently, we could demonstrate a successful functionalization of single-wall carbon nanotubes with the Hamilton receptor [30]. However, the covalent grafting of inorganic nanoparticles with such receptors for the detection of specific organic cyanurate/barbiturate functionalities has not been described so far. Such a construction represents a unique approach for the self-assembly of individual modules towards organic-inorganic hybrid architectures. In this paper we describe, for the first time, a chemical functionalization of nanoparticles with Hamilton receptors and subsequent noncovalent attachment of organic cyanuric acid derivatives ( Figure 2). We report on the synthesis of a hitherto unknown Hamilton receptor 4 covalently equipped with a catechol anchor group, suitable for the binding to metal oxides, in particular zinc oxide [31,32]. One-dimensional zinc oxide nanorods [ZnO] were synthesized and used for surface functionalization, because they provide good stability at the nanoscale, are intensively investigated and as wide-bandgap semiconductor material relevant for many applications in electronics and optoelectronics [33,34]. The particles were coated in a wet-chemical process with the Hamilton receptor to afford monolayer-grafted nanorods [ZnO-HR]. In a detailed characterization study the conditions for stable monolayer grafting were determined. The hybrid material [ZnO-HR] featured stability against agglomeration and growth through stable monolayer surface functionalization and was accessible for further noncovalent surface modification. The unique concept was applied for the formation of newly organized organic-inorganic hybrid architectures. The coupling of tailor designed alkyl-ester substituted cyanurates 5 was accomplished to yield the hybrid structures [ZnO-HR-CA], which featured enhanced dispersability of the ZnO nanorods in apolar solvents.

Synthesis of Building Blocks
A Hamilton receptor covalently equipped with a catechol anchor group suitable for the functionalization of zinc oxide surfaces was prepared as outlined in Scheme 1. First, an amine-Hamilton receptor 2 was synthesized according to a literature procedure [16]. A coupling of this amine 2 with protected protocatechuic acid 1 under modified Steglich conditions using 1-(3-dimethylaminopropyl)-3ethylcarbodiimide (EDC) and 1-hydroxybenzotriazole (HOBt) as coupling agents resulted in Hamilton receptor derivative 3 covalently linked to the catechol anchor. Column chromatography on silica was required to isolate highly pure material. Since catechol has a strong affinity for the attachment to SiO2, the dihydroxy functionality had to be protected. We chose the diphenylmethane-protecting group, as it can be removed very easily by treating with trifluoroacetic acid at room temperature, resulting in a quantitative conversion to the target molecule 4.   As metal oxide core system, one dimensional zinc oxide nanorods [ZnO] were synthesized according to procedures reported earlier [33]. The ZnO nanorods [ZnO] of up to 90 nm in length and up to 20 nm in diameter were obtained by precipitation of zinc acetate dihydrate, Zn(OAc)2·2H2O with potassium hydroxide, KOH in methanol and subsequent refluxing for 49 h at 65 °C. X-ray diffraction (XRD) and high resolution transmission electron microscopy (HRTEM) confirm the presence of ZnO nanorods with its hexagonal Wurtzite structure (B4-type) that are elongated along the crystallographic c-axis. The specific surface area of [ZnO] was determined by BET analysis to be 19.76 m 2 /g.

Chemical Surface Functionalization
The nanoparticles [ZnO] were initially coated in a wet-chemical process with catechol Hamilton receptor 4 to afford the hybrids [ZnO-HR]. The particles were functionalized from 0.15 wt % dispersions in methanol. After sonication and centrifugation all samples were washed three times before analysis. A distinct decrease of Hamilton receptor absorption after functionalization was observed in UV-Vis analysis of the supernatant, indicating surface adsorption. Next, TLC plate experiments revealed interactions of [ZnO] with Hamilton receptor 4. While pure Hamilton receptor 4 was eluted, a mixture of 4 with [ZnO] stayed at the baseline. In initial experiments, the nanorods [ZnO] were exposed to 25.0 mM Hamilton receptor solutions. Such high amounts of ligands, added far above stoichiometric conditions in respect to monolayer surface coverage resulted in multilayer functionalized nanoparticles, as followed from a TGA-MS experiment ( Figure 4). There are two factors that explain such an observed multilayer formation. First, Hamilton receptors are very likely to form hydrogen bonding aggregates and second catechol is known to form multilayers if applied in too high concentrations during the functionalization process [31,35]. However, the high amount of adsorbed Hamilton receptor enabled a detection of mass fragments in a TGA-MS experiment ( Figure 4). In the region of weight loss of the functionalized sample [ZnO-HR] at ~350 °C , simultaneously measured mass spectra revealed particular peaks of ion current for m/z = 57, 109 and 207. These mass fragments could be assigned to fragment ions of 4 (Hamilton receptor arm). In contrast, no notable weight loss was observed when measuring the pure nanorods [ZnO].  For the purpose of attaching cyanurates directly to surface bound Hamilton receptors, monolayer functionalized nanoparticles were required. The findings outlined above suggested that concentrations during functionalization needed to be decreased. To determine the requirements for complete monolayer surface coverage the particles were now exposed to concentrations from 0.1 to 1.5 mM. The obtained hybrids [ZnO-HR] were analyzed in TGA measurements. The weight loss for each sample was converted into grafting density employing the equation [36] (with θ: grafting density; wt: mass loss; MW: molecular weight; SSA: specific surface area): A plot of grafting density vs. concentration resulted in a typical Langmuir isotherm for monolayer adsorption ( Figure 5). By converting the data points into a linear plot (1/θ vs. 1/c) a linear fit of the experimental results could be performed. From the resulting equation a specific adsorption constant for this monolayer functionalization was calculated to be 11.5 × 10 3 , describing the equilibrium between ZnO nanorods, the Hamilton receptor molecule and the solvent methanol (calculation in detail in ESI). As another result the theoretical maximum monolayer grafting density for this molecule was determined to be 1.1 molecules/nm 2 , which is consistent with previously reported grafting densities of catechols on zinc oxide [8,31]. Through insertion of these both values into the Langmuir-equation, we obtained a calculated binding isotherm ( Figure 5). These systematic binding experiments revealed that applying concentrations of 0.75 to 1.5 mM provided monolayer coverage.

Supramolecular Self-Assembly
The supramolecular binding on the surface was accomplished by adding cyanurates in chloroform. Prior to the addition, [ZnO-HR] were transferred into the apolar solvent through centrifugation of the dispersions and subsequent redispersion in chloroform. Then, cyanurate 5 was added in 1 mM concentration to yield the final bilayer coated nanoparticles [ZnO-HR-CA]. The addition was accompanied by individualization of the coated nanoparticles and, therefore, affected the stability of the nanoparticle dispersions. An improved dispersability of [ZnO-HR-CA] resulted because (i) the steric stabilization of the nanoparticles was improved through the alkyl ester moiety and (ii) mainly due to the formation of the highly directional six-fold hydrogen bonding complex which broke up the weaker HR-HR hydrogen bonding interactions. The significant improvement of dispersability is depicted in Figure 8. After eight hours, pure nanoparticles [ZnO] were completely sedimentated in chloroform. The sedimentation of [ZnO-HR] was slightly reduced, while the final hybrids [ZnO-HR-CA] showed good stability in chloroform. As reference experiment to neglect any interactions of the cyanurate with [ZnO], the pure particles were added to a cyanurate solution, which had no impact on their stability.

Materials and Methods
The ZnO nanorods [ZnO] were prepared according to procedures reported earlier [33]. In brief zinc oxide nanoparticles were prepared by precipitation from solutions of Zn(OAc)2 and KOH in methanol. To 125 mL of zinc acetate dihydrate solution, 65 mL KOH solution was dropped within 10 min at a refluxing temperature of 65 °C. Subsequent refluxing of the dispersions for 49 h led to the formation of ZnO nanorods. To remove remaining salt all the samples were washed three times with methanol. In a typical functionalization procedure 0.15 wt % solutions of ZnO nanorods in methanol (25 mL) were sonicated for two hours in different concentrations of catechol 4. The functionalized particles [ZnO-HR] were centrifuged for isolation (3500 rpm; 15 min) and washed three times before analysis. The subsequent supramolecular coupling step was accomplished by addition of cyanuate 5 in a concentration of 1.0 mM to 0.15 wt % particle dispersions in chloroform. For analysis, the cyanurate-attached particles [ZnO-HR-CA] were centrifuged (3500 rpm; 15 min) and redispersed in chloroform.

Conclusions
In conclusion, we described a versatile concept for the construction of nano hybrid architectures through noncovalent grafting of cyanurates to Hamilton receptor modified nanoparticle surfaces. The approach combines unique characteristics of stable covalent surface attachment with advantages of noncovalent supramolecular chemistry. A tailor-designed Hamilton receptor 4 covalently equipped with a catechol anchor group was synthesized and grafted to freshly prepared ZnO nanorods [ZnO]. A detailed functionalization study towards monolayer grafted particles [ZnO-HR] revealed a maximum monolayer grafting density of 1.1 molecules per nm 2 . These building blocks were accessible for noncovalent attachment of cyanurates 5 to yield the hybrid architectures [ZnO-HR-CA], which displayed enhanced dispersability in apolar solvents. The results outline the versatility of this unique construction principle and open up an exciting option for the formation of well-oriented organic-inorganic hybrid architectures for applications in nano-electronics and biotechnology. Further extension and elaboration of such assembly protocols including variation of the inorganic cores (e.g., TiO2, SiO2, Au) and cyanurate-functionalities (e.g., chromophores; semiconductors) are currently underway in our laboratory.