**1. Introduction**

Gold metal nanoparticles with diameters smaller than ~2 nm have received considerable interest in nanoscience because they can be controlled with atomic precision, and, due to discrete energy levels and molecular-like HOMO–LUMO transitions, exhibit optical features fundamentally di fferent from those of larger nanoparticles [1]. Commonly, the term nanoparticles (NPs) is used to refer to entities with diameters greater than 2 nm and an indefinite structure, whereas defined stoichiometric species with a diameter smaller than 2 nm are described as nanoclusters (NCs) [2]. The knowledge of the exact number of atoms present in gold nanoclusters is a very important aspect, as these clusters constitute the link between atomic and AuNP behavior. In this respect, AuNCs show discrete, size-dependent absorption and fluorescence emission from the UV to the near-IR spectral regions, together with significant quantum yield values [1,3,4]. Nonlinear optical (NLO) properties of gold NCs, such as third-order optical nonlinearity, two photon absorption (TPA), two-photon excited emission, and hyperpolarizabilities have been investigated [5–9]. Furthermore, AuNCs exhibit high photostability and biocompatibility, characteristics advantageous in biomedical applications, which can be improved by refining the synthesis, processing, and surface coating of the NCs [10–12]. Surface ligands, such as dendrimers, polymers, proteins, and oligomers play a key role in the stability of AuNCs in solution, and also a ffect their structure and optical properties [1,3]. In this framework, the use of biomolecules as capping ligands improves biocompatibility, and is considered an e fficient

way to transfer chirality to nanoclusters [13–15]. Indeed, three possible mechanisms for chirality transfer have been reported: (i) chiral growth, due to the formation of a chiral metal core influenced by the presence of chiral ligands, (ii) chiral polarization, due to electronic interaction between an achiral metal core and chiral ligands, and (iii) chiral footprint, due to the arrangemen<sup>t</sup> of the ligands on an achiral metal core [16,17]. In this respect, chiral emitting gold nanoclusters, in which the two enantiomers of histidine (His) have been used as both reducing agen<sup>t</sup> and protecting ligands, have been easily synthesized by a one-step approach [18,19]. The presence of suitable ligands bearing –NH2, –COOH, or polymerizable substituents open the way to further functionalization of NCs for applications in sensing, bioimaging, and energy transfer [20–22]. A central focus throughout materials research is control of the organization of the clusters on nanostructured surfaces, as well as understanding of the growth mechanisms at an atomic scale in order to obtain well defined and uniform architectures [23]. In fact, the quality of the metallic layers has a strong impact on the mechanical, electrical, and optical properties of the films. Several procedures to obtain gold nanostructure thin films have been reported, spanning from sputtering [24], lithographic methods [25], chemical vapor deposition (CVD) [26], electroless deposition (ELD) [27,28], and self-assembly approaches, which involve a variety of substrates [29,30]. In this framework, it has been reported that self-assembly of monodisperse gold colloid particles into monolayers on polymer-coated substrates produces highly reproducible macroscopic surfaces active for surface-enhanced Raman scattering (SERS) [31]. In fact, many spectroscopic studies on the interaction of AuNPs with porphyrins and their aggregates in solution and on surfaces have been reported [32–36]. In this particular area, our interest has focused on the J-aggregates of tetrakis(4-sulfonatophenyl)porphyrin (TPPS) [37–42]. These aggregates are formed under acidic conditions and the partially protonated porphyrins are arranged in a lateral stacking geometry, which leads to the occurrence of very peculiar optical properties [43,44]. Due to their manifold applications, various reports have dealt with the immobilization of J-aggregated porphyrins on substrates, highlighting the importance of the deposition step [45]. In particular, microscopic analysis has revealed the presence of di fferently shaped aggregates, such as rod, and nanotubular structures, depending on the experimental conditions [35,43,46]. Moreover, a detailed study has revealed that porphyrin concentration strongly a ffects the amount of J-aggregate adsorbed onto glass surfaces, showing a higher number of adsorbed entities at lower porphyrin concentrations [47]. In the past, J-aggregated porphyrin has been exploited for the design of inorganic/organic nanocomposites in combination with gold nanoparticles and spermine [48] or with gold nanorods [36,49]. Herein, we report on the ability of sub-nanometer sized gold clusters (Au10) capped with l- or d-histidine to induce the formation of chiral TPPS J-aggregates under rather mild acidic conditions. We describe a simple procedure to co-deposit gold nanoparticles and chiral porphyrin J-aggregates onto glass substrates by simple acidification of an aqueous solution of such Au10 clusters. Furthermore, these films have been shown to be active substrates for SERS, as demonstrated by the observation of intensified Raman signals for the co-deposited porphyrin J-aggregates. To the best of our knowledge, this is the first time that gold clusters composed of ten atoms (Au10) have been used in combination with porphyrins to build a nanohybrid assembly (Au10@Jagg) able to self-organize on a solid that can be potentially used as a SERS-active substrate for the detection of chemically- and biologically-relevant species.

#### **2. Materials and Methods**

*Chemicals*. Hydrogen tetrachloroaurate(III) hydrate (99.9%) was supplied by Strem Chemicals (Bischheim, France). d- and l-histidine (98%) were obtained from Sigma-Aldrich (Milan, Italy). The 5,10,15,20-tetrakis(4 sulfonatophenyl)porphyrin (TPPS) was purchased from Aldrich Chemicals (Milan, Italy), and its solutions of known concentration were prepared using the extinction coe fficient at the Soret maximum (ε = 5.33 × 10<sup>5</sup> M−<sup>1</sup> cm<sup>−</sup><sup>1</sup> at λ = 414 nm). Hydrogen peroxide (30%, Sigma Aldrich, Milan, Italy), NaBH4 (98%, Aldrich, Milan, Italy), potassium nitrate, and sulfuric (98%), hydrochloric (37%), and nitric (69%) acids (Fluka, Milan, Italy) were used. All the reagents were used without further purification and the solutions were prepared in dust free Milli-Q water (Merck, Darmstadt, Germany).

*Gold cluster synthesis*. Synthesis of gold clusters (Au10) was carried out according to a literature procedure [18]. Briefly, a solution of l- or d-histidine (0.1 M, 6 mL) was added to the solution of the metal precursor (HAuCl4, 10 mM, 2 mL) under continuous stirring at 25 ◦C for 2 h. After this time, the solution turned pale yellow and was used in this form.

*Plating procedure.* Small pieces of glass cover slides were thoroughly washed in an acid piranha solution (H2SO4:H2O2 4:1 v/v). After being rinsed in pure water, the slides were vertically immersed in 3 mL of solution containing gold clusters (1.5 mL) and HCl (pH 2.0). The metallic structures were left to grow on the substrates for 24 h at room temperature. Care was taken to remove any excess histidine possibly adsorbed onto the surface by dipping the slides in acidic water (pH 2.0). Finally, the substrates were dried under a gentle nitrogen flow.

*Au10@Jagg assemblies*. Glass slides, carefully cleaned according to the procedure described above, were vertically immersed in 3 mL of solution containing gold clusters (1.5 mL), porphyrin (up to a concentration of 5 μM), and HCl (pH 2.0). After an aging time of four days at room temperature, the slides were dried under a gentle nitrogen flow and the excess histidine was removed by quick immersion in water at pH 2.

*Spectroscopic and morphological characterization.* UV–vis spectra were collected on a diode-array spectrophotometer Agilent model 8452, subtracting the spectrum of a clean glass slide. Circular dichroism (CD) spectra were recorded with a Jasco model J-720 spectropolarimeter. Atomic force microscopy (AFM) measurements were performed using a NanoSurf easyScan2 microscope operating in non-contact mode at room temperature, with a resolution of 512 × 512 pixels and a moderate scan rate (1–2 lines/s). Commercial Si-N-type tips (AppNano mod. ACLA) with resonance frequencies of 145–230 kHz were used. Fluorescence emission and resonance light scattering (RLS) experiments were performed on a Jasco mod. FP-750 spectrofluorimeter. A synchronous scan protocol with a right angle geometry was adopted for collecting RLS spectra [50], which were not corrected for the absorption of the samples. Raman and SERS spectra were obtained at room temperature using a Renishaw RM2000 single-grating spectrograph apparatus, equipped with an Ar<sup>+</sup> laser at 514.5 nm and a near-infrared diode laser at 785 nm. Measurements were made in backscattered geometry using a 50× microscope objective filtered by a notch holographic filter, dispersed by a single grating (1200 lines mm<sup>−</sup>1), into a charge-coupled device (CCD) detector cooled to −70 ◦C by the Peltier effect. Spatial resolution was 2 mm and the spectral resolution was 3 cm<sup>−</sup>1. Laser power at the sample was in the range 17–185 μW (514.1 nm exc.) and 200 μW–2 mW (785 nm exc.). No sample degradation was observed under these conditions. To improve the signal-to-noise ratio, a number of spectra were accumulated and summed only if no spectral differences were noted. Raman and SERS spectra were calibrated with indene and CCl4 as standards. In order to measure Raman intensities, nitrate and sulfate were used as internal standards. Therefore, to this end, porphyrin aggregation was fostered by means of the addition of KNO3 (150 mM) and HNO3 (32 mM) to the porphyrin solution (5 μM). Due to the overlap of the 993 cm<sup>−</sup><sup>1</sup> band of sulfate and a porphyrin mode, the 1053 cm<sup>−</sup><sup>1</sup> band of NO3− was used to normalize the spectra.
