Fabrication of biocompatible plasmonic sensing platforms via direct growth of metal nanomaterials on modified silk films

Abstract

Combining silk with noble metal nanomaterials offers a promising route to create biocompatible plasmonic platforms for biological and chemical sensing. However, popular approaches including direct mixing and metal sputtering have inherent limitations such as aggregation and poor morphology control that hinder their practical applications. In this work, we report for the first time the direct growth of silver and gold nanostructures on silk films via a seed-mediated approach. The silk films were chemically modified using both pre- and post-treatment methods, and the resulting growth behaviors of silver and gold were investigated. Compared with gold, silver growth exhibited more reproducible and intriguing results. The effects of various growth parameters were examined in detail, and a growth mechanism was proposed. As a proof of concept, silk films coated with silver nanoplates were demonstrated as surface-enhanced Raman scattering (SERS) substrates capable of detecting pH changes using 4-mercaptobenzoic acid (4-MBA) as a probe molecule. This study not only demonstrates the fabrication of a biocompatible, SERS-active substrate but also provides valuable insight into the growth behavior of metal nanocrystals on biomaterial-based substrates.

---

Introduction

Biocompatible nanocomposites, in which biomolecules are utilized as soft and flexible matrices to incorporate functional inorganic nanomaterials, are very attractive because the resulting materials can be mechanically strong and possess unique electronic and photonic properties. Among various biomaterials, silk fibroin (SF) is widely explored in a broad range of applications from tissue engineering to drug release to biosensing devices because of its remarkable optical and mechanical properties, biocompatibility, biodegradability and tunable fabrication into various forms such as films, hydrogels, microparticles and more.^1,2 SF is a natural protein isolated from the cocoon of the Bombyx mori silkworm. The protein is mainly composed of glycine–alanine–glycine–alanine–glycine–serine (GAGAGS) repeat units that self-assemble into crystalline domains that contribute to the high mechanical strength of silk fibers. While >75% of the amino acids in SF are non-reactive alanine and glycine, methods to chemically modify the serine (12.1%), tyrosine (5.3%), aspartic/glutamic acid (1.1%), and lysine (0.2%) residues have been developed to further tune the mechanical properties, hydrophilicity, surface charge, biological interactions, etc.^3–5 While SF alone has shown many promising applications, its potential can be further enhanced by incorporating functional inorganic nanomaterials to obtain functional silk nanocomposites.^6–12

Among assorted functional inorganic nanomaterials, metal nanocrystals – particularly gold and silver – have been intensively investigated during the past several decades due to their unique localized surface plasmon resonance (LSPR) properties, which allow them to be used in a wide range of applications such as catalysis, thermal therapy, optics, and surface-enhanced Raman scattering (SERS).^13–23 The LSPR, arising from the collective oscillation of conduction band electrons confined at the nanoscale, can cause significant enhancement of electromagnetic fields at or near the surface of these nanomaterials, which is a critical feature for SERS applications. As an ultra-sensitive analytical technique, SERS is capable of dramatically enhancing the Raman signals of analyte molecules adsorbed on substrates for trace level detection. Thus, combining metal nanocrystals with silk promises a fully biocompatible plasmonic sensing platform. Such potential has recently attracted interest from researchers aiming to take advantage not only of the LSPR from metal nanocrystals, but the various forms SF can take.^11,24–28 For example, Anucha et al. prepared a SERS substrate using SF hydrogels as the host matrix that was doped with pre-synthesized metal nanoparticles, for detection of methylene blue in water.²⁷ Instead of using hydrogel, Guo et al. spin-coated a SERS thin film, prepared by mixing aqueous gold nanoparticle solution and acid treated silk solution.¹¹ In their work, the fabricated metal/silk film demonstrated adequate SERS performance by detecting 4-(dimethylamino)pyridine and rhodamine 6G. They noted that the enhancement factor was not high and incorporating more complex structures into silk could result in further improvement. The Chen group developed a plastic SERS substrate coated with SF fibers decorated with multiple branched Au/Ag nanodendrites for monitoring lactic acid in human sweat.²⁸ In their work, they also mixed the silk solution with pre-synthesized dendrite Au/Ag nanomaterials and dried them in a template to form concave pyramid microstructures within a smooth film. Although this improved the enhancement factor, the intrinsic issues of fabricating metal/silk sensing platforms by mixing pre-synthesized nanomaterial with silk solution remain.

Approaches that directly mix pre-synthesized nanomaterials in silk solutions require solvent compatibility and high stability of nanoparticles in aqueous silk solution to maintain the nanoparticle properties. This is challenging as increasing concentration of nanoparticles can cause aggregation as well as potentially induce silk gelation.⁷ In addition, nanocomposites made by such an approach (embedding metal nanoparticles in silk matrices), have limited contact between analytes and metal nanoparticles, which is not ideal for a SERS substrate. To address this issue, some studies have focused on sputtering metal islands on the surface of a silk film to form a SERS sensor film.²⁴ However, the metal formed via sputtering methods tends to have less control of metal morphology at a large scale (i.e. nanogaps, low uniformity and roughness), as well as poor adhesion between the metal and silk. On the other hand, growth of metal nanomaterials directly on silk films can result in better control of the large scale metal morphology and adhesion of the metal to the silk surface. Until now, there has been no report on developing an effective approach to directly grow metal nanomaterials on silk films in large scale, which could be of interest to both fundamental metal nanocrystal growth studies and applied devices.

In this work, we demonstrate a novel method for direct growth of gold and silver nanomaterials on silk films via seed mediated growth. Chemical modification was used to tune the surface charge of the silk film, which can effectively impact the metal nanomaterial growth on large area substrates.^20,29 Ag growth was found to be more repeatable than Au growth. The impacts of varying Ag growth conditions, including the seeding step, reducing agent, and growth time were investigated, and a growth mechanism was proposed based on experimental evidence. Finally, a proof-of-concept experiment was conducted to demonstrate the ability of these nanocomposites to serve as a SERS platform to detect pH changes by using 4-MBA as a probe molecule.

Results and discussion

Silk film fabrication and chemical modification

An overview of both the pre- and post-treatment processes used to produce the modified silk films is shown in Fig. 1A. Silk is primarily composed of neutral and hydrophobic glycine and alanine amino acids.^3–5 Therefore, chemical modification was used to increase the number of charged groups in silk and enhance metal binding through electrostatic interactions. Silk contains an unusually high percentage of tyrosine residues (5.3 mol%, 277 tyrosines per silk protein), so these reactive aromatic groups were targeted for chemical modification using a well-established diazonium coupling reaction.³⁰ This reaction is amenable to a wide variety of aromatic amines, allowing installation of neutral, positively charged or negatively charged functional groups.^3–5

Fig. 1 (A) Method to produce “pre-modified” and “post-modified” silk films decorated with azobenzene groups. (B) Diazonium reaction used to chemically modify the tyrosine residues in silk to install positively charged amine groups.

Two methods were investigated to determine the best procedure for modification. In the first round of experiments, silk solution was cast into a film, dried and then annealed in 70% ethanol to render the films insoluble in water. The weight percent of silk solution used to cast the films varied from 2–5% w/v, resulting in thinner or thicker films. While both concentrations created silk films which were suitable for metal nanomaterial growth, the 5% w/v solution was determined to have greater strength and flexibility when manipulated after the metal structures have been grown on the surface. All further trials from this work used 5% w/v silk solution when performing film synthesis procedures.

To determine what type of surface modification would enhance nanomaterial growth, these films were then “post-modified” (Fig. 1A) with negatively charged sulfonic acid groups at one end and positively charged amine groups at the other (Fig. 2A and C). This was done by selectively dipping one end of the silk film in the respective diazonium reaction solutions. These dually-modified films were then used to determine whether a negatively charged surface or positively charged surface was more effective at enhancing the metal nanoparticle growth. The entire film was submerged in a solution containing gold seeds, followed by a gold growth solution (further details of the metal deposition procedure are discussed in the section below). A striking difference was observed in the film, where a metallic luster could be easily observed by eye only on the end of the film modified with the positively charged amine groups (Fig. 2B). The end modified with negative sulfonic acid groups did not have any observable metal deposition. This result was attributed to the fact that the gold seeds have negatively charged ligands that facilitate electrostatic binding to the amine-modified silk surface. Given this result, all further experiments were carried out with silk films modified with amine groups, as shown in Fig. 1B.

Fig. 2 (A) Photograph of silk films modified on one end with a negatively charged sulfonic acid and with a positively charged amine at the other. (B) Photograph of a dually-modified film after exposure to gold seeds and a gold growth solution. The metallic sheen is only observed on the positively charged end of the film, indicating a preference for metal growth on the amine-modified silk. (C) Cartoon illustrating the modifications made to silk using a diazonium reaction to introduce functionalized azobenzene derivatives.

For comparison, the diazonium reaction shown in Fig. 1B was also carried out on dissolved silk solutions. This “pre-modified” silk solution was then cast into a film, dried and then annealed in ethanol (referred to as “pre-modified films”, Fig. 1A).

Initial evaluation of nanomaterial growth on silk films

General procedure for gold and silver surface growth. Fig. 3A shows the approach taken to conduct gold and silver growth on both the pre- and post-modified silk films. First, citrate-capped gold nanoparticle seeds were synthesized, and then immobilized over the silk film through passive absorption from solution. Due to the positive charge of the silk film, the negatively charged gold seeds were deposited on the film via electrostatic attraction. The pre-placed seeds then facilitated both gold and silver deposition on the silk film after immersing the film into the growth solutions. To control the nanomaterial growth and morphology, sodium citrate and hydroquinone were both employed as reducing agents in the growth solution. The whole growth process was completed in 20 min (details can be found in the Experimental section).

Fig. 3 (A) Scheme of growth process for metal nanomaterial deposition on silk films. (B) Photos of representative films before and after metal growth. (C and D) SEM images of fabricated Au@silk and Ag@silk grown on post- or pre-modified films as noted (scale bar: 1 µm for all images).

Fig. 3B shows photos of representative silk films before metal deposition, and then after gold (Au@silk) and silver (Ag@silk) growth. In both cases, a dramatic difference in appearance by eye was noted as the films took on a metallic luster due to the metal nanomaterials grown on the surface. Fig. 3C and D shows the scanning electron microscopy (SEM) images of the fabricated Au@silk and Ag@silk films on pre-modified and post-modified silk films, revealing differences in the growth density, nanomorphology and size. In all cases, the gold growths took on a spiky, rounded ‘popcorn’ morphology with diameters in the range of 100–500 nm, while the silver grew into vertically oriented nanoplates with a thickness of ∼20 nm.

Comparison of pre- and post-modified silk films.
Characterization of Au@silk film. In general, regardless of whether a post-modification or pre-modification approach was taken, it was difficult to repeatedly get full coverage from the gold nanostructures and there was always some silk exposed. The sizes of the nanostructures varied across batches and the uniformity/coverage of the film was inconsistent. Elemental analysis was performed on post-modified films of Au@silk by energy dispersive X-ray spectroscopy (EDS) elemental mapping (Fig. 4A and B), which confirmed that the nanostructures were composed of elemental Au (golden color). In addition, carbon (red color) from the silk film was clearly detected between the Au nanostructures. Fig. 4C presents the X-ray diffraction (XRD) pattern of the Au@silk film with post-modification, which reveals its crystalline information. The main diffraction peaks were observed at ca. 38.1°, 44.3°, 64.5°, 77.7°, and 81.8°, corresponding to the (111), (200), (220), (311) and (222) planes, respectively, of face-centered cubic (fcc) gold crystal (JCPDS no. 04-0784). While these results do demonstrate the success of synthesizing gold nanostructures on modified silk films, addressing the issues with repeatability and uniformity are challenges that we will explore in future. Thus, the remainder of this manuscript will focus on silver nanostructures.

Fig. 4 (A) SEM image and EDS images, (B) EDS spectrum and (C) X-ray diffraction pattern of fabricated Au@silk on a post-modified film.

Characterization of pre- and post-modified films of Ag@silk. The post-modified silk films exposed to silver growth solutions had lower surface coverage of Ag nanostructures as compared to the pre-modified films. The images in Fig. 3C and Fig. 5A clearly indicate that large areas of the underlying silk film are still exposed. The reason for the difference between the pre- and post-modified films is unclear but may be due to a disparity in surface exposed amino acids within the two films. The assembly of the silk proteins within cast films has been known to change based on the conditions under which the film was formed,³¹ and the chemical modifications in the pre-modified solutions could also influence hydrogen bonding interactions during protein self-assembly. It is possible that there are less surface-accessible tyrosine residues for the post-modified films leading to a lower level of modification. We have attempted to measure the surface contact angle of the different films but have been unable to ascertain a significant difference in hydrophilicity between the pre- and post-modified samples.

Fig. 5 (A) SEM image and EDS images, (B) EDS spectrum and (C) X-ray diffraction pattern of a representative post-modified Ag@silk film.

In addition to lower surface coverage, the silver nanoplates tended to be thinner on the post-modified films (∼20 nm) and have petal-like structures indicating the growth was inefficient. The EDS mapping profile (Fig. 5A) showed that the two elements Ag (blue color) and carbon (red color), were distributed throughout the surface, confirming the successful synthesis of silver on the surface of the silk film. The EDS spectrum showed that the carbon content was higher compared to Ag content, indicating the Ag nanostructures had low coverage on the silk film (Fig. 5B). Fig. 5C presents an XRD pattern of the post-modified Ag@silk film revealing its crystalline information. There were four observed diffraction peaks at 2θ angles ∼38.1°, 44.3°, 64.5°, and 77.7°, corresponding to the (111), (200), (220), and (311) planes, respectively, of face-centered cubic (fcc) silver crystal (JCPDS no. 04-0783). In addition, the diffraction peak at 2θ angle of ∼81.5° assigned to the (222) lattice planes was quite weak and could not be well discerned. Overall, the intensity of the spectrum was low since the silver sparsely covered the silk film.

The same analysis was carried out on Ag@silk films grown on pre-modified silk, which in general had higher surface coverage and thicker nanoplates as shown in the SEM image of Fig. 3D and Fig. 6A. The EDS mapping profile (Fig. 6A) showed that the two elements Ag (blue color) and carbon (red color), were distributed throughout the surface, confirming the successful synthesis of silver on the surface of the silk film. Here, the Ag content was higher compared to the carbon content indicating that the Ag nanostructure had a high and dense coverage on the pre-modified silk film (Fig. 6B). The XRD pattern (Fig. 6C) contained the same four diffraction peaks as the post-modified Ag@silk films, however the (111)/(200) intensity ratio (0.19) was much lower than the standard file value (2.1) indicating abundant (200) facets. Moreover, the sharpness of the (200) peak revealed that Ag nanostructures had good crystallinity. The estimated crystallite size of (200) for the Ag nanostructure, as determined by the Debye–Scherrer formula, was ∼47.20 nm.

Fig. 6 (A) SEM image and EDS images, (B) EDS spectrum and (C) X-ray diffraction pattern of fabricated Ag@silk on a pre-modified silk film.

It is interesting to note that the (111)/(200) intensity ratio (0.19) from the fabricated Ag@silk film with pre-modification is lower than the ratio (3.16) from the fabricated Ag@silk film with post-modification and the standard file value (2.1).³² This indicates that the pre-modified film was predominantly enriched with (200) facets while the post-modified film was predominantly enriched with (111) crystal facets. The observations that the post-modified film had lower coverage and thinner Ag nanoplates suggests that the (111) facets of Ag nanoplates preferentially grew vertically. In contrast, the pre-modified film had higher surface coverage and thicker nanoplates suggesting that the (200) facets of the Ag nanoplates preferentially grew parallel along the silk film surface.

Optimization of silver growth conditions

Given the positive results with silver growth on pre-modified silk films discussed above, further experiments were conducted to optimize the procedure and help elucidate the metal growth mechanism. The effects of seeding, the reducing agent and reaction time were explored as detailed below.

Seeding step. Gold seeds were first deposited on pre-modified silk films to serve as active seeds in the growth step. Films were immersed in a solution containing pre-made citrate-coated gold nanoparticle seeds, and then dried with a stream of nitrogen. To study the impact of the seeding step on Ag growth, silk films with and without this seeding step were directly used for silver growth. Fig. 7A shows the dramatic difference in the silk films after the silver growth process. The film without seeding retains the initial orange-yellow color from the azo modification, with grey precipitate sparsely distributed on the surface. In contrast, the film that went through the seeding process shows uniform metallic gray color across the whole film, indicating the robust coverage of silver. It is worthwhile to mention that it was very challenging to image the gold seeds on the silk film due to the non-conductive nature of the silk film, so we were unable to obtain clear images showing the density of seeds on the film. However, the results clearly show that the seeding step plays a key role for the formation of highly dense Ag nanostructures on the silk film.

Fig. 7 (A) Photos of Ag@silk film with and without seeds. SEM images of Ag@silk film fabricated with only sodium citrate at (B) ×6k and (C) ×25k magnification and with only HQ at (D) ×200 and (E) ×2k magnification.

Reducing agent. In the growth solution, a mixture of reducing agents including sodium citrate and hydroquinone (HQ) was used for reducing Ag⁺ to Ag⁰, and to serve as capping agents. Citrate anions and HQ have long been used as the capping agents when synthesizing metal nanostructures, including Au and Ag.^20,33,34 Control experiments with sodium citrate or hydroquinone alone were conducted to study their roles on the formation of silver nanoplates. The silver nanostructures synthesized with sodium citrate alone or with HQ alone all showed isotropic irregular polyhedrons instead of plate-like structures (Fig. 7B–E). These results indicate that both capping agents are indispensable and play complementary roles for the growth of the Ag nanoplates. Furthermore, the polyhedron particles formed under sodium citrate alone had much smaller individual particles and higher density (Fig. 7B and C) on the silk film while the polyhedron particles formed under HQ alone were much bigger and had relatively lower density (Fig. 7D and E). Both agents are considered weak reducing agents, however, citrate anions can reduce Ag⁺ directly to Ag⁰. Thus, growth solution under sodium citrate can form Ag⁰ in solution and with a high concentration of Ag⁰, Ag⁰ can self-nucleate and form Ag particles. However, the redox potential of HQ is not sufficient to reduce isolated Ag⁺ ions and thus avoids secondary nucleation.^35,36 Instead, the gold seeds facilitate and overcome this barrier and mediate reduction of Ag⁺ ions and continue form nanostructures.^20,36 Thus, the resulting particles are much bigger and less dense on the film.

Growth time. To better understand the formation mechanism of Ag nanostructures, the morphology evolution of the Ag@silk film was monitored by removing films from the growth solution at different times. Within a 5 min growth time, the Ag nanostructures had already formed (Fig. 8A) and nearly covered the silk film (Fig. S1A). In addition, the silver nanostructures had small, blunt, plate-like structures indicating the initial stage of plate formation. When the growth time was increased from 10 min to 20 min (Fig. 8B–D), the plates became much taller and the edge of the plates were sharper while maintaining full coverage on the silk film (Fig. S1B–D).

Fig. 8 SEM images of Ag@silk film with various growth times: (A) 5 min; (B) 10 min; (C) 15 min; (D) 20 min. Scale bar: 1 µm. (E) X-ray diffraction (XRD) pattern of the Ag@silk film with various growth times.

Fig. 8E presents the XRD patterns of silk films with 0 min, 10 min and 20 min silver growth, which reveals their crystalline information. For silk films without silver on the surface, there were no clearly observed peaks. For samples with 10 min or 20 min silver growth, there were four diffraction peaks observed at 2θ angles ∼38.1°, 44.3°, 64.5°, and 77.7°, corresponding to the (111), (200), (220), and (311) planes, respectively, of face-centered cubic (fcc) silver crystal (JCPDS no. 04-0783). With the same XRD scanning conditions, the peak intensities from the sample with a growth time of 20 min were much stronger than the ones with 10 min. The (111)/(200) intensity ratio for the sample with a 10 min growth time was 0.30 and with 20 min growth time was 0.19 while the standard file value is 2.1.³² This indicated that both films had high coverage of silver and were predominantly enriched with (200). Comparing the two conditions, the sample with 10 min growth had thinner Ag nanomaterial on the film compared with the 20 min growth due to the preference for parallel growth of the (200) facets along the silk film.

Proposed growth mechanism

According to the observations and discussion above, we propose the following mechanism for the silver nanostructure formation on the surface of silk films (Fig. 9). In stage 1, the Au nanoparticle seeds electrostatically bind to the positively charged silk film and act as a nucleus for the silver growth. A mixture of sodium citrate and hydroquinone is then used to reduce Ag⁺ to Ag⁰. As stated above, sodium citrate can reduce Ag⁺ to Ag⁰ in solution while HQ only reduces Ag⁺ to Ag⁰ near the metal surface. With both reducing agents, the reduction of Ag⁺ to Ag⁰ is fast. When Ag⁰ is formed, it migrates to the surface of the silk film and aggregates with the existing Au seeds. At this stage, the growth from the seeds protrudes randomly in all directions. During stage 2, the Ag preferentially aggregates at the interface of the positively charged silk surface and the pre-existing metal particles due to stabilization from the negatively charged sodium citrate and HQ, which results in continuous silver film within a few minutes. The protrusions growing away from the surface are slower in the initial growth of the silver nanoplates. It has been previously reported that sodium citrate binds more strongly to (111) facets than the (200) facets of fcc Ag.^37,38 Thus, the (111) facets are expected to grow more slowly than the (200) facets as indicated in stage 2 of Fig. 9. This is consistent with our observations that the entire silk film surface was coated with silver nanostructures after 5 min, but the silver nanoplates were still small and blunt at this initial stage of growth (Fig. 8A and Fig. S1A).

Fig. 9 Proposed growth mechanism of Ag nanoplates. The black arrow indicates the growth direction, and the length of the arrows indicates the rate of growth.

As the growth proceeds to stage 3, the silver film becomes thicker due to the continuing aggregation of Ag, and the nanostructures growing away from the surface become larger and sharper. More importantly, a silver concentration gradient layer starts to form on the surface of film. The silver concentration on the surface of the silk film is lower than that of the bulk solution. As a result, the vertically standing nanoplates grow faster than the angled nanostructures resulting in sharper and larger nanoplates (Fig. 8B and Fig. S1B). As the vertically-standing nanostructures continue to grow higher into the bulk solution, the silver concentration will be further depleted, as illustrated in stage 4 in Fig. 9. The concentration of Ag near the surface will eventually be exhausted and the growth of horizontal nanostructures will completely stop, as illustrated in stage 5 in Fig. 9. Note, the formed Ag⁰ can also be consumed in solution by self-nucleation and growth due to the high concentration of Ag⁰. The solution formed material was found to form ‘desert rose’ shaped particles approximately 1 µm in diameter. These solution-formed particles randomly fall back to the Ag@silk film and were commonly observed in SEM images of films with longer growth times (Fig. S1C and D).

Evaluating use of silver metal films as SERS pH sensors

To test the sensing performance, a proof-of-concept experiment was carried out using the pre-modified Ag@silk film as a pH sensor. Mercaptobenzoic acid (4-MBA) was chosen as a pH probe molecule because it contains a thiol group that strongly binds to plasmonic metal nanoparticles and a carboxyl group that can respond to pH changes.³⁹ It is well known that many cellular activities are accompanied by changes in pH; therefore, a pH-responsive SERS probe is an important detection platform in biosensor application.⁴⁰ The Ag@silk film was first functionalized with 4-MBA by soaking overnight. During the process, the thiol group of 4-MBA can bind to the silver due to strong Ag–thiol interaction. Once bound, the carboxyl group (COOH) of 4-MBA is exposed to the environment and can undergo deprotonation as the pH changes from low to high (Fig. 10B, inset). To test the pH impact, a film was immersed into different solutions at pH values ranging from 1 to 12, and SERS spectra were obtained. Representative data from one film is shown in Fig. 10A where the normalized SERS spectra in the 350–1850 cm⁻¹ spectral region revealed strong Raman signals at different pH conditions. The 4-MBA SERS spectrum is dominated by two strong bands around 1080 and 1590 cm⁻¹, assigned to the aromatic-ring vibrations (ring breathing and axial deformation modes, respectively).^41,42 Because the 1080 cm⁻¹ peak is pH-independent, spectra were normalized to this peak intensity to control overall intensity variations among all spectra. The pH-sensitive vibration modes are related to the protonated or deprotonated state of the carboxyl group of 4-MBA which are observed at around 1425 cm⁻¹ (COO⁻ stretching) and 1705 cm⁻¹ (C=O stretching). The intensity of these two peaks was found to change in opposite directions with the rise of the pH value, as indicated by the dashed lines in Fig. 10A.

Fig. 10 (A) Full Raman spectra (350–1850 cm⁻¹) of Ag@silk films exposed to the pH values listed. (B) The ratio of peak intensity at 1707 cm⁻¹ (COOH) relative to 1419 cm⁻¹ (COO⁻) are plotted. Inset: a schematic diagram of 4-mercaptobenzoic acid (4-MBA) grafted on Ag@silk film and its protonation and deprotonation due to the pH change.

At a low pH, the COOH group is fully protonated resulting in a medium intensity band around 1705 cm⁻¹. With increasing pH, deprotonation occurs that increases the concentration of COO⁻ groups, indicated by the medium-intensity band at 1425 cm⁻¹. When the pH increased from 6 to 7, the COO⁻ peak had a noticeably larger increase. The relationship of the intensity ratio of 1707 cm⁻¹ (C=O stretching) to 1425 cm⁻¹ (COO⁻ stretching) as a function of pH value were plotted and fitted with a Boltzmann function. As shown in Fig. 10B, the Raman intensity has a sudden change between pH 5–7, which corresponds to the isoelectric range of the SERS pH sensor. While this range is higher than the pK_a of 4-MBA in solution (4.22), these values are in good agreement with previous SERS studies using 4-MBA,^43,44 and is likely due to the binding of 4-MBA to the metal surface. Overall, this experiment shows that Ag@silk films can be successfully employed as a biosensor platform. Further development of SERS sensors utilizing Ag@silk films are underway and will be reported in due time.

Conclusions

In summary, we have developed a general procedure to fabricate plasmonic nanocomposites where gold and silver nanostructures are grown directly on chemically-modified silk films. Compared with gold, silver was found to grow more consistently on the silk films and pre-modifying the silk before casting the film yielded a much higher coverage of silver nanoplates than growth on post-modified films. In addition, the impact of experimental conditions including the seeding step, reducing agent composition and the growth time on silver nanoplate formation were studied and the mechanism of silver nanoplate growth was proposed. Finally, a SERS-based pH sensor was successfully fabricated from an Ag@silk film. This study demonstrates a new method to directly grow metal nanomaterials on the surface of silk films, allowing novel nano-biocomposites that can be applied in a variety of fields including biosensing and bioengineering.

Experimental section

Chemicals and materials

All reagents and chemicals were purchased from Thermo Fisher Scientific, Sigma-Aldrich, and Spectrum and used without further purification. Tetrachloroauric acid trihydrate (HAuCl₄·3H₂O, ≥99.9%), hydroquinone (≥99%), sodium hydroxide (NaOH, 30 wt%), were purchased from Sigma-Aldrich. Sodium citrate dihydrate (≥99%) was purchased from Fisher. Nanopure water with a resistivity of 18 MΩ cm was used in all experiments. Silk cocoons were purchased from Treenway Silks, Lakewood, CO.

Silk film preparation

B. mori silk cocoons were processed into an aqueous solution as previously described.⁴⁵ To form films, silk solution (6–7% w/v) was pipetted onto a silicone mat in a ratio of 2 mL per 3 cm × 4 cm rectangle. The films were allowed to air dry in a laminar flow hood overnight. Then the dry films were submerged in 70% ethanol for 24 h to render them insoluble in water. Finally, the films were transferred to DI water slowly over a 2 h period by replacing small amounts of ethanol (approximately 3 mL) with water every 20 min.

Acid modification of pre-cast silk films. A 0.75 × 0.75 cm piece of silk film was cut using a razor blade and added to a Petri dish containing 16 mL of borate buffer (100 mM borate, 137 mM NaCl, pH 9), then placed on ice. Next, 4 mL of nanopure water, 304 mg (1.60 mmol) of p-toluene sulfonic acid (p-TSA), and 69.6 mg (0.26 mmol) procaine hydrochloride were combined, vortexed, and placed on ice for 10 min. To this solution, 136 µL (0.54 mmol) of 4 M sodium nitrite was added, vortexed and returned to ice for 15 min to form the diazonium salt. Finally, the reaction solution was poured into the Petri dish with the silk film and pipetted to mix the solution into the buffer. This was left on ice to react for 40 min. The film was then subsequently rinsed with nanopure water 3 times for 10 min each rinse. The films were stored in nanopure water proceeding with metal growth experiments.

Pre-acid modifying silk solution for films. Plain silk solution (∼5% w/v) was dialyzed against borate buffered saline solution (BBS, 100 mM, 137 mM NaCl, pH 9) for two days, changing the BBS once. To an Eppendorf tube, 1600 µL of the silk solution was added and placed on ice. To a separate tube, 400 µL of nanopure water, 161 mg (0.85 mmol) of p-toluene sulfonic acid, and 58.6 mg (0.22 mmol) of procaine hydrochloride were added and placed on ice for 10 min. An addition of 68 µL (0.27 mmol) of 4 M sodium nitrite was made to the solution in nanopure water, inverted several times and quickly returned to ice for 15 min. The reagent solution was added to the tube containing silk solution, inverted, and returned to ice for 5 min. The pH was adjusted as needed with 1 M NaOH to keep the solution at a pH between 9–9.5. The silk was immediately purified by passing it through a disposable size-exclusion column (Cytiva NAP-25) equilibrated with water. The purified solution was cast into a film as described above for the plain silk films.

Preparation of seed solution

Negatively charged citrate-coated gold nanoparticles (Au NPs, 13 nm in diameter) were prepared according to the well-known citrate reduction method as described in the literature.⁴⁶ HAuCl₄ (75 μL, 0.1 M) was added to nanopure water (30 mL) and brought to a boil on a stir plate. Then, sodium citrate (900 μL, 1 wt%) was added to the solution and left to boil for 40 min. A few minutes after the sodium citrate was added, the solution turned from clear to lavender-colored, before finally turning a wine red. The resulting seed solution was allowed to cool to room temperature, then stored under refrigeration for further use.

Growth of Au or Ag on silk films

The prepared silk films were first seeded by immersing into 0.5 mL of the as-synthesized gold seed solution for 20 min at room temperature. The films were then washed with nanopure water and dried with nitrogen before further use. To grow gold on silk films, HAuCl₄ (75 µL, 1.2 M) was added to nanopure water (9.6 mL) in a 20 mL vial. To grow silver on silk films, AgNO₃ (75 µL, 1.2 M) was used instead of HAuCl₄. The seeded silk film was immersed in this solution, then sodium citrate (22 μL, 3 wt%) was added, followed by hydroquinone (1 mL, 0.3 M). During the whole growth process, the solution was under gentle stirring. After 20 min, another addition of hydroquinone (500 μL, 1.0 M) was added. After another 2 min period, the silk substrate was then removed from the solution and rinsed with nanopure water before being blown dry with nitrogen. Films were stored in a 2 mL centrifuge tube wrapped with parafilm for further use.

SERS sample preparation

To prepare a sample for analysis with SERS, Ag@silk films were soaked overnight in a solution of 1 M 4-mercaptobenzoic acid in ethanol. After soaking, the films were air dried until the following day. The films were then soaked in a solution of nanopure water adjusted to the desired pH with 0.1 M sodium hydroxide or 0.1 M hydrochloric acid for 5 min to prepare for SERS measurements. After taking a measurement, the films were resoaked in another pH solution for 5 min and patted dry with a Kimwipe before reassessment. This was repeated for pH conditions 1 through 12, soaking the film in solutions from acidic to basic conditions.

Instruments and measurements

All SEM images and EDS elemental maps were obtained using a JEOL-7200F field emission SEM operated at 2–5 kV. The crystalline formation of metal nanomaterials was verified using a Rigaku Miniflex 6G benchtop diffractometer. The XRD patterns were measured in the range of 25–90° with a step of 0.04° and a scan rate of 0.50 min⁻¹ while spinning the samples at 10 rpm, employing characteristic Cu Kα radiation having a wavelength of k = 1.540593 Å, with a 40 kV voltage and 15 mA current. The full-width at half-maximum (FWHM) from different peaks was used in Scherrer's equation to determine the average crystallite size of the Ag nanomaterials. The SERS measurements were carried out with a Renishaw inVia Qontor confocal Raman microscope. The excitation wavelength was 532 nm, and the power was 200 µW. The spectra were collected with a 50× objective and with 5 s integration time.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting the findings of this study are presented in the main text and SEM images of large area Ag@silk films with various growth times are shown in the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d6nr00050a.

The raw data for the figures are available from the corresponding author upon reasonable request.

Acknowledgements

This research was supported by the National Science Foundation (CHE-2108842, CBET-2344490, DMR-1807878, CHE-REU-2243968), and a Henry Dreyfus Teacher Scholar Award (A. R. M.). We also are grateful for the following support from the College of Science and Engineering (CSE) at WWU: a summer seed grant through the Advanced Materials Science & Engineering Center (AMSEC), M. E. K. received the Jarvis Summer Research Fellowship and M. D. received the Arlan Norman Summer Research Fellowship. The SERS data acquisition on the confocal Raman microscope was made possible through an NSF-MRI award (CHE-2019208). SEM and XRD studies were conducted on instruments funded by the Joint Center for Deployment and Research in Earth Abundant Materials (JCDREAM). The authors thank Kyle Mikkelsen from AMSEC and Michael Kraft and Cassi King from Scientific Technical Services (STS) at Western Washington University for their technical support. The authors are grateful to all members in Bao group and Murphy group for valuable discussions.

References

1. L. D. Koh, J. Yeo, Y. Y. Lee, Q. Ong, M. Han and B. C. Tee, Mater. Sci. Eng., C, 2018, 86, 151–172.
2. C. Holland, K. Numata, J. Rnjak-Kovacina and F. P. Seib, Adv. Healthcare Mater., 2019, 8, 1800465.
3. A. R. Murphy and D. L. Kaplan, J. Mater. Chem., 2009, 19, 6443–6450.
4. S. A. L. Matthew and F. P. Seib, ACS Biomater. Sci. Eng., 2024, 10, 12–28.
5. J. K. Sahoo, O. Hasturk, T. Falcucci and D. L. Kaplan, Nat. Rev. Chem., 2023, 7, 302–318.
6. E. Kharlampieva, V. Kozlovskaya, R. Gunawidjaja, V. V. Shevchenko, R. Vaia, R. R. Naik, D. L. Kaplan and V. V. Tsukruk, Adv. Funct. Mater., 2010, 20, 840–846.
7. C. Belda Marín, V. Fitzpatrick, D. L. Kaplan, J. Landoulsi, E. Guénin and C. Egles, Front. Chem., 2020, 8, 604398.
8. T. Cohen-Karni, K. J. Jeong, J. H. Tsui, G. Reznor, M. Mustata, M. Wanunu, A. Graham, C. Marks, D. C. Bell, R. Langer and D. S. Kohane, Nano Lett., 2012, 12, 5403–5406.
9. S. H. Tsao, D. Wan, Y.-S. Lai, H.-M. Chang, C.-C. Yu, K.-T. Lin and H.-L. Chen, ACS Nano, 2015, 9, 12045–12059.
10. C. Belda Marín, C. Egles, V. Humblot, Y. Lalatonne, L. Motte, J. Landoulsi and E. Guénin, ACS Biomater. Sci. Eng., 2021, 7, 2358–2371.
11. C. Guo, G. N. Hall, J. B. Addison and J. L. Yarger, RSC Adv., 2015, 5, 1937–1942.
12. O. Akturk, K. Kismet, A. C. Yasti, S. Kuru, M. E. Duymus, F. Kaya, M. Caydere, S. Hucumenoglu and D. Keskin, RSC Adv., 2016, 6, 13234–13250.
13. J. Lee, J. C. Park and H. Song, Adv. Mater., 2008, 20, 1523–1528.
14. L. R. Hirsch, R. J. Stafford, J. A. Bankson, S. R. Sershen, B. Rivera, R. E. Price, J. D. Hazle, N. J. Halas and J. L. West, Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 13549–13554.
15. Y. Yang, J. Liu, Z.-W. Fu and D. Qin, J. Am. Chem. Soc., 2014, 136, 8153–8156.
16. M. Fernanda Cardinal, B. Rodríguez-González, R. A. Alvarez-Puebla, J. Pérez-Juste and L. M. Liz-Marzán, J. Phys. Chem. C, 2010, 114, 10417–10423.
17. H. Zhang, S. Duan, P. M. Radjenovic, Z.-Q. Tian and J.-F. Li, Acc. Chem. Res., 2020, 53, 729–739.
18. Y. Bao and A. Oluwafemi, Chem. Commun., 2024, 60, 469–481.
19. J. R. Crockett, M. Wang, J. E. Doebler, T. Pawale, X. Li and Y. Bao, Chem. Mater., 2022, 34, 9282–9293.
20. S. Cohen-Pope, J. R. Crockett, M. Wang, K. Flynn, A. Hoff and Y. Bao, J. Mater. Chem. C, 2020, 8, 12427–12436.
21. H. Wu, C. Kanike, A. Marcati and X. Zhang, Langmuir, 2024, 40, 4218–4227.
22. J. P. Singh, H. Chu, J. Abell, R. A. Tripp and Y. Zhao, Nanoscale, 2012, 4, 3410–3414.
23. Z. Li, X. Huang and G. Lu, J. Mater. Chem. C, 2020, 8, 3956–3969.
24. J. H. Choi, M. Choi, T. S. Ho, S. Kim, S. Choi, S. H. Choi and K. M. Byun, Opt. Express, 2022, 30, 7782–7792.
25. F. Xu, F. Ma, Z. Ding, L. Xiao, X. Zhang, Q. Lu, G. Lu and D. L. Kaplan, ACS Appl. Mater. Interfaces, 2019, 11, 42896–42903.
26. M. Lee, H. Jeon and S. Kim, Nano Lett., 2015, 15, 3358–3363.
27. C. B. Anucha, E. Guenin and A. Percot, Materials Today Sustainability, 2025, 31, 101169.
28. H. Ma, Y. Tian, A. Jiao, C. Wang, M. Zhang, L. Zheng, S. Li and M. Chen, Vib. Spectrosc., 2022, 118, 103330.
29. X. Zhao, X. Luo, C. G. Bazuin and J.-F. Masson, ACS Appl. Mater. Interfaces, 2020, 12, 55349–55361.
30. A. R. Murphy, P. S. John and D. L. Kaplan, Biomaterials, 2008, 29, 2829–2838.
31. C. B. Borkner, M. B. Elsner and T. Scheibel, ACS Appl. Mater. Interfaces, 2014, 6, 15611–15625.
32. C. Muhammed Ajmal, K. P. Faseela, S. Singh and S. Baik, Sci. Rep., 2016, 6, 34894.
33. B. Wiley, Y. Sun, B. Mayers and Y. Xia, Chem. – Eur. J., 2005, 11, 454–463.
34. Z. Zaheer and Rafiuddin, Colloids Surf., A, 2011, 384, 427–431.
35. A. Król-Gracz, P. Nowak, E. Michalak and A. Dyonizy, Acta Phys. Pol. A, 2012, 121, 196–199.
36. S. T. Gentry, S. J. Fredericks and R. Krchnavek, Langmuir, 2009, 25, 2613–2621.
37. V. Pawlik, S. Zhou, S. Zhou, D. Qin and Y. Xia, Chem. Mater., 2023, 35, 3427–3449.
38. A. Guo, N. Hall, T. Hamlett, J. R. Crockett, A. Talbott, T. Ogunrinola, A. Oluwafemi, M. Burke, Q. Chen and Y. Bao, J. Phys. Chem. C, 2025, 129, 14204–14213.
39. L. Zhang, Q. Zhao, Z. Jiang, J. Shen, W. Wu, X. Liu, Q. Fan and W. Huang, Biosensors, 2021, 11, 282.
40. L. Zhang, Q. Zhao, Z. Jiang, J. Shen, W. Wu, X. Liu, Q. Fan and W. Huang, Biosensors, 2021, 11, 282.
41. M. Potara, T. Nagy-Simon, A. M. Craciun, S. Suarasan, E. Licarete, F. Imre-Lucaci and S. Astilean, ACS Appl. Mater. Interfaces, 2017, 9, 32565–32576.
42. A. Michota and J. Bukowska, J. Raman Spectrosc., 2003, 34, 21–25.
43. M. Xie, F. Li, P. Gu, F. Wang, Z. Qu, J. Li, L. Wang, X. Zuo, X. Zhang and J. Shen, Cell Prolif., 2019, 52, e12618.
44. J. E. Park, N. Yonet-Tanyeri, E. Vander Ende, A.-I. Henry, B. E. Perez White, M. Mrksich and R. P. Van Duyne, Nano Lett., 2019, 19, 6862–6868.
45. K. G. Hausken, R. L. Frevol, K. P. Dowdle, A. N. Young, J. M. Talusig, C. C. Holbrook, B. K. Rubin and A. R. Murphy, Macromol. Chem. Phys., 2022, 223, 2200119.
46. C. Jiang, S. Markutsya and V. V. Tsukruk, Langmuir, 2004, 20, 882–890.

---

Footnote

† These authors contributed equally.

---