Development of ordered metal nanoparticle arrangements on solid supports by combining a green nanoparticle synthetic method and polymer templating for sensing applications

Abstract

In this paper, we present a one step, simple, robust and “green” methodology to fabricate high-density ordered arrays of uniform Au nanoparticles (NPs) and Au NP clusters at room temperature over large areas which are suitable for high-performance surface enhanced Raman spectroscopy (SERS). The method is based on the template-guided self-assembly process undergone by polystyrene-b-poly(4-vinylpyridine) copolymer (PS-b-P4VP), in which gold salt is incorporated within the micellar cores (P4VP) and subsequently reduced by HEPES buffer salt in a reduction process at room temperature to give single and clustered NP arrays with hexagonal order. Sizes of P4VP domains are nearly constant (ca. 27 nm) and contain clusters of tiny Au NPs of ca. 4 nm separated 2–3 nm each other. Moreover, the clusters can be transformed either on singly dispersed anisotropic (“nanocrescent-like”) or spherical NPs only by their exposure to O₂ plasma. Excellent SERS performance with high signal intensities (as evidenced by high enhancement factors >8 × 10⁵) and excellent reproducibility were found for the cluster arrays. This is because of the uniform size and gap distance of the gold clusters in large areas. The anisotropic and isotropic dispersed NP metallic substrates also displayed good sensitivities but with relatively slightly lower enhancement factors (ca. 10⁴ to 10⁵). All these metallic substrates can be achievable without the use of any expensive equipment or clean room processing enabling the potential obtention of low-cost and high-throughput production of chips for (bio)sensing applications.

---

Introduction

In the past few years, important efforts in the biomedical area have been made to design and obtain new biosensors which enable the early detection of diseases and their suppression through the early administration of suitable therapy. In this regard, the application of metal nanostructures with plasmonic response derived from block copolymer self-assembly/lithography techniques (BCL) opens up the possibility of designing new platforms with high, modulable and tunable sensing response. To do that, it is possible to take advantage of the well-known optical properties¹ of metallic nanoparticles (NPs) in the visible (Vis) and near-infrared (NIR) spectra ranges, where they exhibit one or more intense UV-vis absorption bands originating from the collective resonant oscillations of their conduction electrons upon incidence of electromagnetic radiation of suitable wavelength (localized surface plasmon resonances, LSPR).^2,3 Hence, the application of metal NPs with plasmonic response for detection has become extremely popular, either based on plasmon shifts due to refractive index changes when a biomolecular recognition process takes place near the NP surface (LPSR-based sensors),^4–6 or on the direct identification of the bioanalyte through its Raman scattering fingerprint when adsorbed onto it (surface enhanced Raman scattering, SERS, sensors).⁷ Most of these new biosensors require the ordered assembly of NPs onto substrates in order to ensure high detection limits, safety and reproducibility in multiple tests. Thus, it is interesting to develop a flexible method that allows nanostructured assembled metallic substrates to be obtained with the possibility of tuning the fabrication conditions to achieve the requirements of any type of plasmonic sensor. In this regard, self-assembly techniques compared to nanopatterning methods^8–12 are attractive since the spontaneous organization of nanoscale building blocks allows for the large-scale, parallel production of periodic nanostructures at low costs. In this way, block copolymers (BCP) can form periodically ordered nanostructures in thin films with typical dimensions between 5–50 nm.¹³ The size and shape of the BCP microdomains and, thus, the resulting nanostructure of the BCP scaffold, can be controlled by manipulating the polymer chain lengths and composition, the volume fraction of each block, the temperature and the composition of the surrounding atmosphere. Their orientation and lateral ordering can be improved by different methods, such as graphoepitaxy, chemical patterning, applied electrical and shear fields, temperature or solvent annealing.¹⁴ Among all the strategies for controlling the location and distribution of metallic nanoparticles,^15–19 the in situ NP synthesis process is a very useful technique due to the low aggressiveness of the required chemistry within the block copolymer template. This approach involves the selective loading of metal ions on specific blocks domains usually followed by an oxygen plasma treatment,²⁰ or the immersion in a reduction medium to form NPs with different morphologies.^21,22

In this manner, either LPSR or SERS sensors derived from BCL have been produced in laboratory. For example, Russell et al.²³ reported the fabrication and optical characterization of dense and ordered arrays of spherical gold and gold/silver metal NPs. The metal arrays were produced by reducing metal salts in BCP templates of poly styrene-block-poly(4-vinylpyride) (PS-b-P4VP) or polystyrene-block-polyoxyetilene (PS-b-PEO) BCPs thin films. Spatz et al.²¹ obtained highly ordered arrays of spherical gold NPs with tunable particle sizes and interparticle spacing by combination of BCL and chemical reduction. In a similar approach, Liz-Marzán et al.²⁴ also created arrays of Ag NPs combining BCL and chemical reduction. In this case, the growth of the Ag NP islands was controlled to ensure the growth of the NPs in close proximity one each other with the objective of creating a dense array of hot spots which dramatically increases the SERS efficiency. Kim et al.²⁵ developed a method to fabricate an ultrahigh-density array of silver nanoclusters suitable for SERS substrates by the reduction of incorporated silver nitrate in the P4VP block of a PS-b-P4VP template. These authors could control the gap distance between neighboring silver nanoclusters to promote the formation of hot spots²⁶ which contribute to SERS enhancement factors (EF) of ca. 10⁸. Recently, Feldmann et al.²⁷ have claimed a two-step approach for the fabrication of quasi-hexagonal ordered arrays of star-shape gold nanoparticles substrates for SERS by combination of plasma treatment and an overgrowth solution, obtaining a high Raman signal intensity due to the strongly enhancement of the electromagnetic field at the tips of these anisotropic NPs,^28,29 whereas Polleux et al.³⁰ have obtained hexagonal arrays of Au nanorings by means of photochemical reduction under UV irradiation (185 nm) of gold ions inside swelled P2VP blocks of PS-P2VP block copolymers with high absorption–scattering rations providing excellent photothermal properties.

Nevertheless, despite the achievements made in terms of simple fabrication techniques, improvement of detection limits, and/or reusability there are still many aspects that need further consideration as, for example: (i) the development of synthetic processes using simple, economic, scalable and more friendly-for-environment techniques/solvents; (ii) the achievement of well-ordered NP arrays in large scales; and (iii) the maintenance of such ordered arrangements when subjected, for example, to overgrowth processes for signal improvement. Thus, herein we propose a one step, simple, cheap and reproducible methodology developed at room temperature making use of block copolymer lithography for obtaining suitable polymeric templates to synthesize well-ordered arrays of gold NPs and nanoclusters suitable to be used as SERS substrates through an in situ one-step approach by the use of 2-[4-(2-hydroxyethyl)-1-piperazinyl]ethane-sulfonic acid (HEPES), a Good's buffer broadly used in tissue culturing and in (bio)chemistry laboratories.^31,32 On one hand, we used the block copolymer self-assembly properties of PS-b-P4VP copolymers and the affinity of a gold metal precursor possesses to the P4VP microdomains³³ to create well-ordered metal polymeric hybrid thin films from polymer and chloroauric acid (HAuCl₄) mixed solutions. These ordered hybrid films were used as templates for the in situ NP synthesis within the micellar cores by addition of HEPES as a simultaneous reducing and stabilizing agent thanks to the presence of a piperazine ring in its molecular structure.^34–37 Reduction with HEPES enabled to obtain hexagonally packed ordered NP clustered arrays which find application as very sensible SERS substrates, as demonstrated by using 4-nitrothiophenol as a model analyte due to its affinity to gold surfaces. Enhancement factors of up to ca. 8 × 10⁵ were observed thanks to the increased electromagnetic field as a consequence of the formation of hot spots between nearly-located tiny NPs inside the clusters. In addition, short exposure to oxygen plasma allows the partial melting of the cluster arrangements to obtain single arrays of anisotropic nanoparticles with a “nanocrescent-like morphology”, also with larger EF values (5-fold) than classical spherical NP ones thanks to the enhancement of electromagnetic filled at the tips of these anisotropic structures.

Experimental

Materials

An asymmetric block copolymer, polystyrene-block-poly(4-vinylpyridine) (PS-b-P4VP, molecular weight: 41 500 g mol⁻¹ for PS and 17 500 g mol⁻¹ for P4VP, respectively) was purchased from Polymer Source Inc. and used as received. 2-[4-(2-hydroxyethyl)-1-piperazinyl]ethanesulfonic acid (HEPES), chloroauric acid (HAuCl₄), toluene, 4-nitrothiphenol (4-NBT) and tetrahydrofurane (THF) were from Sigma Aldrich. Silicon wafers substrates were from Ted-Pella. Water was of Milli-Q grade.

PS-b-P4VP@HAuCl₄ thin film preparation and characterization

Thin films were prepared following the method developed by Russell et al.²³ Briefly, PS-b-P4VP block copolymer was dissolved in a toluene/THF mixture (80/20, v/v) at 70 °C for 2 h and cooled to room temperature to yield a 0.5 wt% polymer solution. Different aliquots of a HAuCl₄ stock solution (50 mM) were added to the previous polymer solution to obtain different HAuCl₄/P4VP molar ratios R = 0.3, 0.9, 1.5 and 2.5. The final mixtures were stirred for 24 h to ensure the complete Au salt coordination of metal ions and poly(4-vinylpyridine) chains.³⁸ In a typical procedure, thin films with loaded-metal salt were prepared by spin-coating of the previous mixed solution (PS-b-P4VP@HAuCl₄) onto silicon wafers at 2000 rpm for 60 s if not otherwise stated. The films were kept under vacuum overnight at room temperature to obtain the final ordered templates. Different HEPES concentrations (10 mM, 450 mM and 1.5 M, pH 7.4) were prepared and used for gold ion reduction by immersing the (PS-b-P4VP@HAuCl₄) thin films in HEPES solutions for different time periods at room temperature. Afterwards, samples were rinsed with water to eliminate non-coordinated NPs and dried under nitrogen. To remove the polymer template 15 s exposure to oxygen plasma (50 W, 0.2 mbar, Plasma Surface Technology model Femto PECCE, Diener Electronics) followed by 24 h UV irradiation at 254 nm were used. Further O₂ plasma exposure led to the appearance of additional nanostructures like anisotropic “nanocrescent”-like and spherical NPs (see below). The ordered metallic arrays were characterized using a field-emission scanning electron microscope ZEISS FESEM ULTRA Plus operating at 3 kV. Transversal film cross-sections were measured using a focused ion beam and scanning electron microscope instrument model Helios NanoLab 600 (FEI, Oregon, USA). UV-vis spectroscopy measurements were performed in a CARY 100 Bio UV-visible (Agilent Technologies, Santa Clara, USA) spectrophotometer. AFM images were acquired in non-contact mode using a XEI-100 instrument from Park Systems and crystal silicon cantilevers with a spring constant of 40 N m⁻¹ and a resonant frequency of 300 kHz (ACTA, AppNano).

Film preparation for SERS measurements

SERS substrates were prepared by incubation of thin films in 4-NBT solutions for 20 min. Unbound and weakly adsorbed analyte molecules were cleaned with ethanol and the films dried. Raman spectra were measured upon excitation with a 633 nm laser line. Inelastically scattered light was collected with a Renishaw Invia Reflex system, equipped with a confocal optical microscope, high resolution gratings (1200 g mm⁻¹), a two dimensional CCD detector and an x-y-z motorized stage with 100 nm of resolution. Spectra were collected by focusing the laser line onto the samples, using a 100× objective (N.A. 0.95), with accumulation times of 5 s and laser power at the sample of 5 mW. Bulk Raman spectra were acquired with a laser intensity of 75 mW and 120 s acquisition times, averaged different location onto the control substrates and background corrected.

Results and discussion

It has been previously shown that HEPES can reduce Au ions in aqueous solution even at low concentrations (∼6 ppm) thanks to the presence of a piperazine ring in its molecular structure able to generate nitrogen-centered free radicals.³⁴ In this manner, anisotropic branched Au NPs such as multipods with different tip numbers³⁵ or branched nanoshells³⁹ were previously obtained through a kinetically controlled synthetic processes by simply tuning the reaction conditions such as temperature, precursor salt or reductant concentrations. Moreover, the oriented attachment of primary nuclei led to the formation of Au nanoflowers with 10 or more tips per particle,³⁶ whose size and number of tips could be controlled by modification, for example, of the HEPES/Au molar ratio. Also, the presence of HEPES as a coating provides a relatively “clean” surface where postsynthesis surface modifications may be carried out for biological applications under fully biocompatible conditions. In the light of these previous works and the environmentally-benign character of the used buffer salt, we here simultaneously combine block copolymer lithography and an in situ Au NP synthesis and stabilization promoted by HEPES within the polymeric template to create well-ordered arrays of single and clustered Au NPs arrays in thin films able to be used as efficient SERS substrates.

Hence, Au single NPs and clustered NP arrays were fabricated by reduction with HEPES of HAuCl₄ salt inside the P4VP domain of PS-b-P4VP reverse micelles spun coated onto silicon substrates using the steps illustrated in Fig. 1. In a typical procedure, PS-b-P4VP copolymer was dissolved into a mixed toluene/THF (80/20 v/v) solvent to give a 0.5 wt% micellar solution. Once spherical micelles were formed in solution, gold salt was incorporated into the micellar P4VP core via coordination with the pyridine groups of the copolymer. Next, the micellar solutions were usually spun-coated at 2000 rpm on silicon/glass wafers if not otherwise stated. The obtained hybrid films displayed an excellent 2D quasi-hexagonally periodic order of P4VP domains surrounded by a PS matrix. It is well-known that the periodicity of the reverse micellar templates can be varied in steps of a few nanometers by varying either the copolymer molecular weight, the copolymer concentration or the spinning speed.²⁵ To this end, the spin-coating speeds were also varied in the range 500–3000 rpm to achieve polymeric templates with systematically decreasing periodicities from 51 to 34 nm but the size of the micellar template features remains constant (not shown). Subsequently, a mild reduction step using HEPES as the reductant and stabilizing agent to obtain single or clustered gold NPs was performed (Fig. S1†). Furthermore, the films were subjected to either oxygen plasma and/or UV irradiation to ensure complete removal of the polymeric material when required; however, caution should be considered in this step since shape changes and/or loss of NP order in the arrangements could emerge if optimal conditions were not applied (see below). In addition, the structural and geometrical attributes of the resulting clustered and single NP arrays could be modified by changing some of the synthetic conditions such as the HEPES concentration, incubation time, the [HAuCl₄]/[P4VP] molar ratio or the time of plasma exposure/UV light irradiation amongst others, and were thoroughly characterized using plan-view and cross-sectional SEM, and AFM measurements.

Fig. 1 Scheme showing the “green” synthetic process for the obtention of the different kinds of nanostructured gold substrates.

Effect of incubation time and cleaning process

Although the reduction and protective role of HEPES for the solution synthesis of anisotropic Au NPs has been previously observed,⁴⁰ we firstly tested if such reduction process might take place to reduce Au ions embedded into spun-coated polymeric micelles to form 2D ordered micellar hybrid assemblies. To do that, HAuCl₄-loaded PS-P4VP micellar thin films at molar ratio R = 0.9 were incubated for different times (from 10 to 180 min) in a 10 mM HEPES solution. As observed in Fig. 2a, after 10 min of incubation only very few bright spots corresponding to very dispersed Au NPs within some P4VP micellar cores were observed. At longer incubation times (20 min, and clearer at 40 min) small and bright spots could be roughly discerned within the micellar cores, which would correspond to tiny NPs of sizes between 4 ± 1 nm (Fig. 2b and c). A polymeric film layer precludes an optimal and closer visualization of these tiny NPs through electron microscopy. After 60 and 120 min some small NPs were still observed but an increasing number of slightly larger NPs can be observed within the micellar cores (size ca. 10–15 nm) which might result from the overgrowth and fusion of neighboring smaller NPs (Fig. 2d and e). At longer incubation times (>120 min), no additional significant changes in the hybrid films were observed (Fig. 2f).

Fig. 2 SEM images of PS-b-P4VP/HAuCl₄ films in HEPES after (a) 10 min, (b) 20 min, (c) 40 min, (d) 60 min, (e) 120 min, and (f) 180 min of incubation (R = 0.9, [HEPES] = 10 mM). Scale bars are 200 nm.

To better visualize the formation of NP clusters within the polymeric micellar templates, a combination of oxygen plasma and UV light irradiation was used to remove the organic polymeric layer. In this regard, it is necessary to bear in mind that both energy sources possess reduction capability of the metal salt so that in order to exclusively consider the reduction effect of HEPES a very limited exposure time (specially to O₂ plasma) should be applied at most. It was found that a 15 s-exposure of the hybrid films to O₂ plasma was successful to allow a clear visualization of the obtained NP cluster arrays without a significant contribution to the gold salt reduction, whilst longer exposure times led to structural modifications within the clusters (see below for further details). Exposition and subsequent reduction of hybrid substrates only under O₂ plasma led to the formation of disordered clustered NPs and partially-ordered single NPs arrays as the exposition time increases, respectively (Fig. S1†). On the other hand, Au salt-loaded polymeric substrates exclusively exposed to UV light irradiation contributed to the formation of some larger and dispersed NPs within the polymeric templates (Fig. S2†). Finally, it was found that a combination of 15 s O₂ plasma exposure followed by UV light irradiation for 24 h allowed the complete removal of the polymeric template while preserving the order, size and composition (in terms of the number of constituent NPs) of the formed nanoclusters as observed by energy dispersive X-ray spectroscopy (EDX, Fig. S3†). Further UV exposure led to a progressive loss of the ordered nanoclusters arrays with an increase in their heterogeneity (Fig. S4†).

Effect of HEPES concentration

HEPES concentration was found to be a key aspect for the obtention of well-defined arrays of NP clusters. Fig. 3 shows different hybrid PS-b-P4VP/HAuCl₄ films with a [HAuCl₄]/[P4VP] molar ratio of 0.9 reduced at different HEPES concentrations for 40 min followed by a very short oxygen plasma exposure (15 s) to partially eliminate the polymeric film to improve visualization. As observed, the reduction process with HEPES led to an enhanced contrast of polymeric P4VP micelle cores. An increasing amount of very small gold nanoparticles in the form of clusters could be detected as the HEPES concentration increases from 10 mM to 1.5 M. The cluster sizes are ca. 27 ± 4 nm, which are relatively smaller than the polymeric micellar diameters (ca. 36 ± 6 nm) as a consequence of the shrinking of polymeric chains upon drying and plasma degradation, and are very uniform over the entire substrate (wafer scale). The cluster NP arrays retained the quasi-hexagonal order of the polymeric template independently. One interesting feature is that each Au nanocluster contains several tiny gold nanoparticles of sizes ranging from 4 ± 1 nm, with intercluster distances of ca. 46 ± 7 nm (Fig. 3a and b and S5†). As observed, the availability of HEPES seemed to promote the formation of new nuclei against the growth of existing particles inside the micellar cores. This effect was also indirectly confirmed when applying oxygen plasma for more than 15 s, leading to the progressive attachment and fusion of closely-located smaller Au NPs inside clusters (see below). At the largest HEPES concentration (1.5 M) larger NPs were obtained probably as a consequence of a quicker reduction process favoring the growth of existing NPs, which may also lead to some fusion of adjacent NPs (Fig. 3c).

Fig. 3 SEM images of HEPES concentration effect on the HAuCl₄ reduction process of hybrid PS-P4VP polymeric films at a HAuCl₄/P4VP ratio R = 0.9: (a) 10 mM, (b) 450 mM (c) 1.5 M (15 s of O₂ plasma exposure and 24 h UV radiation λ = 254 nm were applied). Insets are magnifications of some regions of the images for a better visualization. Scale bars are 200 nm.

Effect of O₂ plasma exposure

Once a suitable incubation time and reductant concentration for the hybrid films were chosen, the hybrid thin films were further treated by oxygen plasma at room temperature for certain time intervals in order to simultaneously eliminate the polymeric template and analyze both the contribution of O₂ plasma to the reduction process of the metal salt (specially at the lowest HEPES concentration used) and its effect on the size and shape of the resulting nanostructures, which should affect the optical properties of the ordered NP cluster arrays. Fig. 4 and S6–S8† show the effect of O₂ plasma exposure times on the films incubated at different HEPES concentrations for 40 min.

Fig. 4 SEM images of the O₂ plasma exposure effect on the formation of single NPs and clustered NP arrays: (a) 0 s (b) 15 s, (c) 60 s, (d) 120 s, (e) 150 s and (f) 180 s for hybrid films of R = 0.9 incubated for 40 min in 10 mM HEPES. Insets are magnifications of some regions of the images for a better visualization. Scale bars are 200 nm.

For hybrid thin films reduced at the smallest HEPES concentration (10 mM) exposure to O₂ plasma for short times (15 to 60 s) led to the observation of quasi-hexagonally packed Au nanocluster arrays formed by singly dispersed Au NPs with mean sizes of ca. 4 ± 1 nm and particle interdistances of 46 ± 7 nm, as mentioned previously (Fig. 4b and c). The enhanced contrast of the nanoclusters with O₂ plasma exposure arises from the progressive elimination of the polymer component, as commented previously. Here, it is necessary to consider that the resulting NP clusters exhibit a 3D geometry as they directly inherit the shape of the micellar core template, as observed from cross-sectional HRTEM and AFM images (Fig. S9†); therefore, the 3D geometry can imply artifacts in the plan-view TEM images allowing that the NPs from different planes can sometimes appear fused.⁴¹ As a consequence of the close proximity of the small NPs comprising the clusters (gap NP center to center distance ∼7 nm), the possibility to form hot spots under electromagnetic irradiation which leads to an enhancement of the SERS signal should occur (see below). Moreover, it is also worth noting that the visual inspection of the nanocluster arrays shows their uniformity across the entire coated area. SEM analysis was used to spot any uncovered areas to quantify the yield of the cluster formation. On the basis of the observation that no uncovered areas could be spotted in FESEM visualization at different zones of the substrate and observed defects were found to be due to spin-coating artifacts at the extreme corners of the substrate, we quantified the yield of NP cluster formation to be >85%. This is an aspect of high importance toward the utility of these arrays for applications such as SERS probes, as the uniformity of the cluster arrays would contribute toward a low signal intensity fluctuation.

On the other hand, longer O₂ plasma exposures (120 s) involved the progressive fusion of adjacent Au NPs within the cluster giving rise to anisotropic nanoparticles with a shape resembling to that of nanocrescents, which is a very interesting morphology to configure ordered NP arrays for SERS as a result of their capacity to enhance the local electromagnetic field at their tips.⁴² This kind of anisotropic structure might be originated from the fusion of a density of small NPs at the interface between the micellar core and the PS shells within the clusters as a consequence of the lower diffusion of HEPES molecules from the interface to the center of the P4VP micellar core.⁴³ In some cases, when the salt diffusion is more regular the nanocrescent structure is not formed and more circular, irregular shapes are obtained. This shape resembles to the “potatoid-like” one recently obtained by Polleux et al. through reduction of the loaded metal salt inside the polymeric micelles by a photochemical approach using a 185 nm-UV light.³⁰ Additional O₂ plasma exposure (>120 s) of the thin substrate films led to the progressive conversion of the nanocrescents to quasispherical Au NPs while keeping the quasi-hexagonal order. Furthermore, the higher the exposure time the smaller the final NP size was if compared to that of the original template size (micelle core). The interparticle distances upon progressive melting of the nanoclusters increases with the plasma exposure time in agreement with previous works.²⁵

Similar shape transitions within the clusters upon plasma exposure at similar time points were observed for the reduction of the Au salt in 450 mM HEPES solution (Fig. S6†). Conversely, reduction of substrates within a 1.5 M HEPES solution involved a faster process leading firstly to the formation of tiny NPs inside the micellar cores, which immediately started to aggregate upon application of a short O₂ plasma interval (15 s, Fig. S7a†). Further exposition to plasma of ca. 1 min also led to the formation of anisotropic nanocrescent-like structures, which are formed by two or three of gold domains (NPs) fused together (Fig. S7c†), followed by the development of quasispherical NPs at larger plasma exposure times, as commented before.

Effect of the HAuCl₄/P4VP molar ratio

Au salt-loaded polymeric films at different HAuCl₄/P4VP molar ratios (R = 0.3, 0.6, 0.9, 1.5, and 2.5) were reduced in 10 mM HEPES solution to study the effect of gold salt concentration on the size and shape of the resulting NP arrays. Clearly, the metal salt concentration plays a role in the final size and shape of the resulting nanostructures. Under similar plasma time exposures (15 s), low R values (R = 0.3 and 0.6, Fig. 5a and b) make the available Au salt concentration is not enough to saturate the P4VP core in the original template, being concentrated in a small portion of the microdomain enabling the formation of small ordered spherical nanoparticles, with sizes of ca. 13 ± 3 nm and 15 ± 2 nm, respectively. At larger mole ratios (R = 0.9 and 1.5), the P4VP microdomains contain more gold salt, which under HEPES reduction form very small NPs uniformly distributed and confined in the interior of the P4VP microdomains leading to a high population of nanoclusters with quasi-hexagonal order and sizes of ca. 27 ± 4 nm and 24 ± 4 nm, respectively (Fig. 5c and d). Finally, for the highest molar ratio analyzed (R = 2.5), the P4VP microdomains seem to be fully saturated by the gold salt, resulting in the formation of fewer and larger NPs (of ca. 2 to 5 individual NPs with sizes of ca. 7–12 nm) within clusters of sizes of ca. 25 ± 4 nm upon HEPES reduction (Fig. 5e and f).

Fig. 5 SEM images of the HAuCl₄/P4VP molar ratio effect on the formation of metal nanostructures incubated for 40 min in 10 mM HEPES under 15 s oxygen plasma: (a) R = 0.3, (b) R = 0.6, (c) R = 0.9, (d) R = 1.5, (e) R = 2.5, and (f) size distribution of sample (e). Insets are magnifications of some regions of the images for a better visualization. Scale bars are 200 nm.

Nanoparticle arrays as efficient SERS substrates

Metal (specially Au or Ag) NPs organized in patterned ensemble arrays have been attracting increasing attention due to the enhanced field resulting from surface plasmon coupling effect, as mentioned previously. The single NP and clustered arrays fabricated in the present study provide an optimal platform for investigating the correlation between the SERS activity and the NP arrangement of Au nanostructures and the potential influence of coupling effects due to the presence of hot spots when several NPs are in close proximity within a cluster array. We carry out SERS measurements using 4-nitrobenzenthiol (4-NBT) as a target molecule by exploiting selected single and clustered NP arrays supported on reconstructed BCP inverse micelle films. A series of Raman spectra of 4-NBT (1 × 10⁻⁶ M) adsorbed onto the different metallic substrate arrays are displayed in Fig. 6a. The SERS spectra reveal the characteristic peaks of 4-NBT at 1079, 1107 and 1307 cm⁻¹. It can be observed that Raman signals are significantly enhanced for substrates with single NP arrays with anisotropic shapes, that is, in the form of nanocrescent, in comparison to the classical spherical NP arrays. Moreover, Raman signals of the clustered NP arrays are even larger than that of individual isotropic and anisotropic NPs. The SERS signals also increased with decreasing the inter-distance between NP nanoclusters because of the electromagnetic coupling between them largely increases with decreasing the gap (intercluster) distance (Fig. 6b).⁴⁴ This enhanced SERS intensity with decreasing the cluster interdistance is consistent with the progressive red-shift of the peak position to ca. ∼590 nm in the UV-vis absorption spectra (Fig. 6c). In this regard, the extinction spectra reveal a single broad peak red-shifted by up to ∼70 nm in comparison to that of isolated NPs. Since each extinction spectrum averages optical properties from numerous clusters, the standard deviation in the number of tiny NPs composing the cluster and the interparticle and intercluster separations are likely to contribute to the broadening of the observed spectral line width.⁴⁵ A systematic red-shifting of the resonance peak position as well as an increase in the peak intensity is noted with a decrease in intercluster gaps as a consequence of strong coupling of localized surface plasmons between adjacent metal nanostructures,^46,47 although some contribution from enhanced scattering cannot be neglected. Plasmonic coupling is known to be a sensitive function of the interparticle separation as well as the number of nearest neighbors.⁴⁸ In the case of the NP nanoclusters shown here, the average intracluster NP separation is lower than the inner particle radius, so an excellent coupling is expected. Moreover, the extinction spectra of the cluster arrays are in sharp contrast to those here obtained for ordered single dispersed isotropic particles, which display extinction maxima at 522, and anisotropic ones (with a main peak at 548 nm and a shoulder at ca. 750 nm, see Fig. 6d), or those corresponding to random clustered nanoparticle aggregates for which a predominant peak corresponding to the isolated NPs appears at ∼520 nm with only a broad low-intensity peak corresponding to clusters above 600 nm are observed.⁴⁸ As a consequence, our ordered NP cluster arrays can be expected to exhibit low variation in intra- and intercluster separations in comparison to random nanoparticle clusters reported in the literature. Also, since the excitation wavelength used in the SERS measurements was 633 nm, the red-shift in the UV-vis spectra would favor a progressive better coupling between the incident laser light and localised-surface plasmons resulting in a higher SERS intensity.⁴⁹

Fig. 6 (a) SERS spectra of 4-NBT adsorbed onto single NP (), “nanocrescent-like” NPs () and NP clusters (). (b) SERS spectra of NP cluster substrates with intergap distances of 51 (), 46 (), 39 () and 34 nm (). (c) Extinction spectra of the different Au NP substrates: dispersed single Au NPs (); “nanocrescent-like” NPs (); and NP clusters with gap interdistances of 51 (), 46 (), 39 () and 34 nm (). (d) SERS enhancement factor of 4-NBT for singly dispersed spherical NPs (), anisotropic “nanocrescent-like” dispersed NPs (), and clustered NP arrays ().

The SERS enhancement factor (EF) for the different types of substrates was calculated based on the equation

EF = (I_surface/I_ref) × (N_ref/N_surface) (1)

where I_surface and I_ref are SERS intensities of the samples and reference, and N_surface and N_ref are the number of SERS active molecules in the samples and reference, respectively, taking the band at 1307 cm⁻¹ for comparison of different substrates. 4-NBT-coated substrates without metal NP arrays were used for reference measurements. Nevertheless, it is necessary to consider that quantification of the EF in a SERS substrate is not trivial and needs some assumptions since the number of adsorbed molecules is poorly defined.⁵⁰ Hence, in this study we assumed that 4-NTB molecules are uniformely adsorbed on the metal nanostructures, following the methodology previously described by Capasso et al.⁵¹

The overall EF values were of the order of ca. 9 × 10³, 4 × 10⁴ and 1.3 × 10⁵ for the spherical, “nanocrescent-like” and clustered NP arrays respectively at gap distances of ca. 49, 49 and 51 nm. Hence, the EF for the nanocluster arrays obtained from 15 s plasma exposure at R = 0.9 and nanocrescent NPs was measured to be fifteen and five-fold higher than that of the singly dispersed NP arrays, respectively. SERS EF values reported here are comparable to those previously obtained, for example, using ordered hexagonal arrays of Au nanospheres as substrates (∼9 × 10⁴),⁵² and silver (2.0 × 10⁵)²⁵ and gold clustered arrays (8.8 × 10⁴ to 2.0 × 10⁵)⁵³ obtained by plasma and/or chemical reduction with sodium borohydride. The relatively lower EF efficiency found in some cases can be ascribed to the presence of some HEPES moieties at the Au NP surface and the dielectric BCP layer between the NPs and the clusters, respectively, which is still remaining under very short O₂ plasma exposures (if larger times are applied fusion of the clusters starts as shown above) and may interfere partially with the access of the 4-NTB molecules and the near-field LPSR coupling between neighbouring Au nanocluster domains.

The SERS results here obtained for the nanoclusters may be explained by the existence of a large variety of small Au NPs containing isolated hot spots with higher electromagnetic enhancement and by the shift of the LPSR coupling band to longer wavelengths facilitating an optimal matching with the wavelength of the He–Ne laser excitation source (633 nm). Moreover, there exists an increase in EF as the interdistance between nanoclusters decreases thanks to the enhancement of electromagnetic coupling between the nanoclusters and the formation of hot spots, which increases the Raman spectra (Fig. 6d). The relatively high SERS enhancements exhibited by cluster arrays are therefore consistent with experimental⁵⁴ and theoretical⁵⁵ investigations, which have shown the dependence of SERS enhancement on the density of hot-spots. Conversely, for substrates with higher density arrays of spherical or anisotropic (nanocrescents-like) NPs, EFs were still lower than for clustered NP arrays. For the present SERS nanocluster substrates, it seems that a critical gap distance of ca. 34 nm starts to provide a significant increase in EF, which is larger than for the singly dispersed isotropic and anisotropic NPs. This indicated that the Raman intensity was affected by not only from gap distance but also the internal gold nanocluster structure,^25,54,56 that is, the interaction of the tiny NPs within the clusters results in a contribution to the SERS intensity. Namely, the arrays of gold nanoclusters increased plasmon coupling among individual NPs within the nanoclusters, which is similar to the multiscale SERS signal enhancement.^55,56 Since the Raman intensity reflects from all of the nanoclusters within the laser spot, the measured intensity was quite reproducible thanks to the large number of nanoclusters contributing Raman spectra within the laser spot (∼1.54 μm²) as well as very good uniformity in the gap distance and gold nanocluster size. Nevertheless, further reduction of the intergap distance should lead to a larger SERS enhancement as observed for silver nanoclusters obtained by oxygen plasma reduction through modification of the molecular weights of the copolymers used as templates.^25,57

Conclusions

In summary, 2-D Au ordered arrays of single isotropic, anisotropic and NP cluster arrays with controlled lateral configuration have been obtained using inverse micelle monolayers of PS-PVP block copolymers by a one step “green” chemical reduction method using HEPES salt as reductant and stabilizing agent and O₂ plasma/UV irradiation for polymer removal. The structure, evolution and optical characteristics of the obtained arrays were analyzed considering the influence of different synthetic parameters by means of SEM, AFM and UV-vis techniques. Using 4-NBT as a molecular probe, it was found that the produced Au cluster arrays displayed the largest overall SERS enhancement factor (ca. 8 × 10⁵) as a consequence of the reduced separation (2–4 nm) of tiny individual NPs forming the cluster, which enables to act as hot spots for enhancing SERS signal. Meanwhile, anisotropic NP ones showed EF values 2–3 times larger than the isotropic ones. Additional increments in EF might be obtained either using other PS-P4VP copolymers with much shorter PS blocks or larger polymer concentrations. The ease of the present one-step methodology without the need of any reductant agent (except HEPES) or energy source for reduction (such as oxygen plasma^21,23,25 or light irradiation³⁰) and the possibility of easy post-functionalziation makes the present approach really interesting for different applications, specially for biological ones as, for example, Raman imaging of cell membranes.⁵⁷ Hence, SERS substrates here obtained involve a significant improvement in terms of ease and cost of fabrication over other methodologies reported in literature given that the cluster NP arrays with the best SERS enhancements can be obtained without use of any expensive equipment or clean-room environment, and/or harsh processing conditions.

Acknowledgements

Authors thank Ministerio de Economía y Competitividad (MINECO) for research project MAT 2013-40971-R and Xunta de Galicia for projects EM 2013-046 and GPC2015-007, respectively. S. B. greatly acknowledges MINECO for her Ramon y Cajal fellowship. A. P. and E. V. A. are grateful to MINECO for their FPU fellowships.

Notes and references

1. (a) J. Pérez-Juste, I. Pastoriza-Santos, L. M. Liz-Marzan and P. Mulvaney, Coord. Chem. Rev., 2005, 249, 1870–1901; (b) C. Langhammer, Z. Yuan, I. Zoric and B. Kasemo, Nano Lett., 2006, 6, 833–838; (c) V. Myroshychenko, J. Rodríguez-Fernández, I. Pastoriza-Santos, A. M. Funston, C. Novo, P. Mulvaney, L. M. Liz-Marzán and F. García de Abajo, Chem. Soc. Rev., 2008, 37, 1792–1805.
2. C. L. Haynes and R. P. Van Duyne, J. Phys. Chem. B, 2001, 105, 5599–5611.
3. K. H. Su, Q. H. Wei, X. Zhang, J. J. Mock, D. R. Smith and S. Schultz, Nano Lett., 2003, 3, 1087–1090.
4. (a) A. P. Alivisatos, Nat. Biotechnol., 2004, 22, 47–52; (b) R. Wilson, Chem. Soc. Rev., 2008, 37, 2028–2045.
5. C. Sönichssen, B. M. Reinhard, J. Liphardt and A. P. Alivisatos, Nat. Biotechnol., 2005, 23, 741–745.
6. B. Sepúlveda, P. Angelomé, L. M. Lechuga and L. M. Liz-Marzán, Nano Today, 2009, 4, 244–251.
7. (a) L. Brus, Acc. Chem. Res., 2008, 41, 1742–1749; (b) J. P. Camden, J. A. Dieringer, J. Zhao and R. P. Van Duyne, Acc. Chem. Res., 2008, 41, 1653–1661.
8. J. Henzie, J. Lee, M. Lee, W. Hasan and T. W. Odom, Annu. Rev. Phys. Chem., 2009, 60, 147–165.
9. S. M. Jeon, K. Y. Jang, S. H. Lee, H. W. Park and B. H. Sohn, Langmuir, 2008, 24, 11137–11140.
10. (a) E. M. Hicks, S. Zou, G. C. Schatz, K. G. Spears and R. P. Van Duyne, Nano Lett., 2005, 5, 1065–1070; (b) D. Wouers and U. S. Schubert, Angew. Chem., Int. Ed., 2004, 43, 2480–2495.
11. R. D. Piner, J. Zhu, F. Xu, S. Hong and C. A. Mikin, Science, 1999, 283, 661–663.
12. Y. Xia, J. A. Rogers, K. E. Paul and G. M. Whitesides, Chem. Rev., 1999, 99, 1823–1848.
13. (a) F. S. Bates and G. H. Fredickson, Phys. Today, 1999, 52, 32–38; (b) A. V. Ruzette and L. Leibler, Nat. Mater., 2005, 4, 19–31.
14. (a) S. B. Darling, Prog. Polym. Sci., 2007, 32, 1152–1204; (b) J. Bang, U. Jeong, D. Y. Ryu, T. P. Russell and C. J. Hawker, Adv. Mater., 2009, 21, 4769–4792; (c) I. W. Hamley, Prog. Polym. Sci., 2009, 34, 1161–1210.
15. (a) J. J. Chiu, B. J. Kim, E. J. Kramer and D. J. Pine, J. Am. Chem. Soc., 2005, 127, 5036–5037; (b) C. Minelli, C. Hinderling, H. Heinzelmann, R. Pugin and M. Liley, Langmuir, 2005, 21, 7080–7082.
16. (a) R. B. Thompson, V. V. Ginzburg, M. W. Matsen and A. C. Balazs, Science, 2001, 292, 2469–2472; (b) R. B. Thompson, V. V. Ginzburg, M. W. Matsen and A. C. Balazs, Macromolecules, 2002, 35, 1060–1071.
17. (a) E. Glogowski, R. Tangirala, T. P. Russell and T. Emrick, J. Polym. Sci., Part A: Polym. Chem., 2006, 44, 5076–5086; (b) Z. Li, H. Sai, S. C. Warren, M. Kamperman, H. Arora, S. M. Gruner and U. Wiesner, Chem. Mater., 2009, 21, 5578–5584.
18. Y. Zhao, K. Thorkelsson, A. J. Mastroianni, T. Schilling, J. M. Luther, B. J. Rancatore, K. Matsunaga, H. Jinnai, Y. Wu, D. Poulsen, J. M. Fréchet, A. P. Alivisatos and T. Xu, Nat. Mater., 2009, 8, 979–985.
19. (a) S. Föster and M. Antonietti, Adv. Mater., 1998, 10, 195–217; (b) J. P. Spatz, T. Herzog, S. Mömer, P. Ziemann and M. Möller, Adv. Mater., 1999, 11, 149–153.
20. J. P. Spatz, S. Mömer, C. Hartmann, M. Möller, T. Herzog, M. Krieger, H.-G. Boyen, P. Ziemann and B. Kabius, Langmuir, 2000, 16, 407–415.
21. T. Lohmueller, E. Bock and J. P. Spatz, Adv. Mater., 2008, 20, 2297–2302.
22. J. Chai and J. M. Buriak, ACS Nano, 2008, 2, 489–501.
23. P. A. Mistark, S. Park, S. E. Yalcin, D. H. Lee, O. Yavuzcetin, M. T. Tuominen, T. P. Russell and M. Achermann, ACS Nano, 2009, 3, 3987–3992.
24. A. Sánchez-Iglesias, P. Aldeanueva-Potel, W. Ni, J. Pérez-Juste, I. Pastoriza-Santos, R. A. Álvarez-Puebla, B. N. Mbenkum and L. M. Liz-Marzán, Nano Today, 2010, 5, 21–27.
25. W. J. Cho, Y. Kim and J. K. Kim, ACS Nano, 2012, 6, 249–255.
26. R. F. Aroca, Surface Enhanced Vibrational Spectroscopy, Wiley, New York, 2006.
27. L. Osinkina, T. Lohmüller, F. Jäckel and J. Feldmann, J. Phys. Chem. C, 2013, 117, 22198–22202.
28. N. Pazos-Perez, S. Barbosa, L. Rodríguez-Lorenzo, P. Aldeanueva-Potel, J. Pérez-Juste, I. Pastoriza-Santos, R. A. Álvarez-Puebla and L. M. Liz-Marzán, J. Phys. Chem. Lett., 2010, 1, 24–27.
29. S. Barbosa, A. Agrawal, L. Rodríguez-Lorenzo, I. Pastoriza-Santos, R. A. Alvarez-Puebla, A. Kornowski, H. Weller and L. M. Liz-Marzán, Langmuir, 2010, 26, 14943–14950.
30. F. Kundrat, G. Baffou and J. Polleux, Nanoscale, 2015, 7, 15814–15821.
31. N. E. Good, G. D. Winget, W. Winter, T. N. Connolly, K. Izana and R. M. M. Singh, Biochemistry, 1966, 5, 467–477.
32. N. E. Good and K. Izawa, Methods Enzymol., 1972, 24, 53–68.
33. C. Lee, S. H. Kim and T. P. Russel, Macromol. Rapid Commun., 2009, 30, 1674–1678 , and references therein.
34. A. Habib, M. Tabata and Y. G. Wu, Bull. Chem. Soc. Jpn., 2005, 78, 262–269.
35. J. Xie, J. Y. Lee and D. I. C. Wang, Chem. Mater., 2007, 19, 2823–2830.
36. J. Xie, Q. Zhang, J. Y. Lee and D. I. C. Wang, ACS Nano, 2008, 2, 2473–2480.
37. S. Park, J. Y. Wang, B. Kim and W. Chen, Macromolecules, 2007, 40, 9059–9063.
38. J. P. Spatz, V. Z. H. Chan, S. Mößmer, F. M. Kamm, A. Plettl, P. Ziemann and M. A. Möller, Adv. Mater., 2002, 14, 1827–1832.
39. A. Topete, M. Alatorre-Meda, E. M. Villar-Alvarez, A. Cambon, S. Barbosa, P. Taboada and V. Mosquera, ACS Appl. Mater. Interfaces, 2014, 6, 11142–11157.
40. S. Krishnamoorthy, R. Pugin, J. Brugger, H. Heinzelmann and C. Hinderling, Adv. Funct. Mater., 2006, 16, 1469–1475.
41. F. L. Yap, P. Thoniyot, S. Krishnan and S. Krishnamoorthy, ACS Nano, 2012, 6, 2056–2070.
42. K. Li, L. Clime, B. Cui and T. Veres, Nanotechnology, 2008, 19, 145305.
43. L. Shan, S. Punniyakoti, M. J. Van Bael, K. Temst, M. K. Van Bael, X. Ke, S. Bals, G. Van Tendeloo, M. D'Olieslaeger, P. Wagner, K. Haenen and H.-G. Boyen, J. Mater. Chem. C, 2014, 2, 701–707.
44. F. J. García-Vidal and J. B. Pendry, Phys. Rev. Lett., 1996, 77, 1163–1166.
45. C. L. Haynes and R. P. Van Duyne, J. Phys. Chem. B, 2001, 105, 5599–5611.
46. T. Atay, J. H. Song and A. V. Nurmikko, Nano Lett., 2004, 4, 1627–1631.
47. P. K. Jain, W. Huang and M. A. El-Sayed, Nano Lett., 2007, 7, 2080–2088.
48. J. Schmitt, P. Mächtle, D. Eck, H. Möhwald and C. A. Helm, Langmuir, 1999, 15, 3256–3266.
49. K. A. Willets and R. P. Van Duyne, Annu. Rev. Phys. Chem., 2007, 58, 267–297.
50. E. C. Le Ru, E. Blackie, M. Meyer and P. G. Etchegoin, J. Phys. Chem. C, 2007, 111, 13794–13803.
51. E. J. Smythe, M. D. Dickey, J. Bao, G. M. Whitesides and F. Capasso, Nano Lett., 2009, 9, 1132–1138.
52. W. Lee, S. Y. Lee, R. M. Briber and O. Rabin, Adv. Funct. Mater., 2011, 21, 3424–3429.
53. Y. H. Jang, K. Chung, L. N. Quan, B. Spacková, H. Sípova, S. Moon, W. J. Cho, H.-Y. Shin, Y. J. Jang, J.-E. Lee, S. T. Kochuveedu, M. J. Yoon, J. Kim, S. Yoon, J. K. Kim, D. Kim, J. Homola and D. H. Kim, Nanoscale, 2013, 5, 12261–12271.
54. H. Xu, J. Aizpurua, M. Käll and P. Apell, Phys. Rev. E, 2000, 62, 4318–4324.
55. L. Yang, B. Yan, W. R. Premasiri, L. D. Ziegler, D. D. Negro and B. M. Reinhard, Adv. Funct. Mater., 2010, 20, 2619–2628.
56. B. Yan, A. Thubagere, W. R. Premasari, L. D. Ziegler, D. Dal Negro and B. M. Reinhard, ACS Nano, 2009, 3, 1190–1202.
57. G. Zito, G. Rusciano, G. Pesce, A. Dochshanov and A. Sasso, Nanoscale, 2015, 7, 8593–8606.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra04925g

---