Controllable formation of high density SERS-active silver nanoprism layers on hybrid silica-APTES coatings

Accepted Manuscript Title: Controllable formation of high density SERS-active silver nanoprism layers on hybrid silica-APTES coatings Author: J. Pilipavicius R. Kaleinikaite M. Pucetaite M. Velicka A. Kareiva A. Beganskiene PII: DOI: Reference:

S0169-4332(16)30642-0 http://dx.doi.org/doi:10.1016/j.apsusc.2016.03.169 APSUSC 32935

To appear in:

APSUSC

Received date: Revised date: Accepted date:

22-9-2015 28-1-2016 22-3-2016

Please cite this article as: J.Pilipavicius, R.Kaleinikaite, M.Pucetaite, M.Velicka, A.Kareiva, A.Beganskiene, Controllable formation of high density SERS-active silver nanoprism layers on hybrid silica-APTES coatings, Applied Surface Science http://dx.doi.org/10.1016/j.apsusc.2016.03.169 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

CONTROLLABLE FORMATION OF HIGH DENSITY SERS-ACTIVE SILVER NANOPRISM LAYERS ON HYBRID SILICA-APTES COATINGS Pilipavicius J a , Kaleinikaite R a , Pucetaite M b , Velicka M b , Kareiva A a , Beganskiene A a a

Department of Inorganic Chemistry, Faculty of Chemistry, Vilnius University, Naugardukas str. 24, Vilnius, 03225 Lithuania

b

Department of General Physics and Spectroscopy, Faculty of Physics, Vilnius University, Sauletekis ave. 9, Vilnius, 10222 Lithuania E.mail – [email protected], phone no.: +37052193184

KEYWORDS: Silver nanoprism, APTES, sol-gel, SERS, coatings ABSTRACT In this work sol-gel process for preparation of the uniform hybrid silica-3-aminopropyltriethoxysilane (APTES) coatings on glass surface is presented from mechanistic point of view. The suggested synthetic approach is straightforward, scalable and provides the means to tune the amount of amino groups on the surface simply by changing concentration of APTES in the initial sol. Deposition rate of different size silver nanoprisms (AgNPRs) on hybrid silica coatings of various amounts of APTES were studied and their performance as SERS materials were probed. The acquired data shows that the deposition rate of AgNPRs can be tuned by changing the amount of APTES. The optimal amount of APTES was found to be crucial for successful AgNPRs assembly and subsequent uniformity of the final SERS substrate – too high APTES content may result in rapid non-stable aggregation and non-uniform assembly process. SERS study revealed that SERS enhancement is the strongest at moderate AgNPRs aggregation level whereas it significantly drops at high aggregation levels. 1

INTRODUCTION

Over the past few decades numerous significant achievements were made in noble metal nanoparticle research. The field unites scientists from different research areas, such as biomedical engineering [1] and material science [2]. Success in the development and production of new metal nanostructures opened new possibilities for molecular sensing [3], novel optical materials [4, 5], integrated circuits [6-8], communications [9] and cancer treatment [10-12]. Most of these applications are based on unique optical properties of metal nanoparticles such as localized surface plasmon resonance (LSPR). Depending on size and origin of the particles LSPR of spherical noble metal nanoparticles typically occurs in near-UV and visible light region. By introducing anisotropy into particle shape, it is possible to shift LSPR to Near-IR region. Anisotropic silver nanoparticles, e.g. silver nanoprisms (AgNPRs), exhibit strong LSPR in 600 – 1200 nm range, which is important for application in surface-enhanced Raman spectroscopy (SERS). Currently many studies are directed towards plasmon driven SERS materials [13-18]. Often such materials include nanoparticle colloids or nanostructured thin films of noble metals, like silver or gold. General strategies for their preparation employs chemical synthesis of anisotropic nanoparticle colloids or lithography methods for highly ordered nanostructured surfaces. [3, 19-21]. Successful application of SERS nanomaterials in determination of otherwise untraceable amounts of molecules is well known. [22-25]. However, more quantitative analysis and higher versatility in the determination of various molecules is still in need. Nanomaterials prepared by lithography stands out for their reproducible SERS enhancements whereas materials made by chemical approaches have considerably higher SERS enhancements and are less expensive. Nanolithography techniques require expensive equipment and are hardly scalable. These limitations therefore stimulate the interest in development of novel and straightforward synthetic methods for highly ordered nanostructured materials in more economical ways. Production of uniform colloidal silver nanoparticle layers on a solid surface could be the method of choice, since it combines high enhancement typical to colloidal nanoparticles as well as increased repeatability due to uniform SERS substrate layers. Such substrates can be prepared by spin-coating of Ag colloid previously reported by Torres, V., et al. [26], however the method has its limitations and to obtain highly ordered nanoparticles on a solid substrate is complicated. Xue, C., et al. [27] obtained high uniformity

AgNPR layers by employing Langmuir-Blodgett method. Yet this method requires silanization and further hydrophobization of AgNPRs, which limits diversity of molecules applicable on such SERS substrates. An alternative method [28] offers synthesis based on direct AgNPRs self-assembly on solid surface functionalized with amino groups. Due to the strong interactions between nitrogen atom and silver nanoparticle surface it is possible to deposit high-density AgNPRs layers on the amino group rich solid surface. There exist many methodologies to prepare such surfaces of high affinity for Ag nanoparticles with majority of them based on solution or gas phase silanization using APTES [29]. Surfaces silanized by these methods usually produce only sparsely distributed AgNPRs. High fidelity self-assembly of AgNPRs directly depends on the uniformity of silanized surface and amount of attached amino groups, as reported earlier by us and others [30]. This aminization method requires activation of the surface by aggressive chemicals such as concentrated hydrogen peroxide and sulphuric acid (or ammonia), which limits substrate choice. Moreover, amount of attached amino groups and their distribution are hardly controllable due to short lived activated surfaces, non-perfect activation process and purity of used reagents. In this study we aim to investigate influence of various parameters to the deposition process which would guide to develop a reproducible and scalable method for preparation of uniform AgNPRs layers on a glass substrate. Our approach is based on fabricating hybrid silica-APTES ((3-Aminopropyl)triethoxysilane) coatings on glass substrates via the sol-gel process, followed by assembly of AgNPRs. In this way uniform aminized surfaces are accessible in a more reproducible manner as compared to direct surface aminization. The main advantage of the proposed method is the possibility to control AgNPRs deposition through the number of amine groups introduced to the surface. Uniformity, density and deposition rate of the assembled AgNPRs layers are important parameters, which will be affected by amino group distribution. In this work, we have synthesized AgNPRs via the seed-mediated method described by Aherne D. [31], which enables to prepare AgNPRs of different sizes and thus investigate their deposition process on coatings containing different amounts of APTES. The obtained AgNPRs aggregation level and size effects to their performance in SERS is also reported. 2

MATERIALS AND METHODS

2.1 Materials Tetraethyl orthosilicate (TEOS, 98%, Sigma-Aldrich), N- (3-Aminopropyl)triethoxysilane (APTES, 99%, Sigma-Aldrich), Silver nitrate (AgNO3, 99%, Acros), Poly(sodium 4-styrenesulfonate) (PSSS, 1000000 Mw, Sigma-Aldrich), Sodium Borohydride (NaBH4, 99%, Sigma-Aldrich), L-Ascorbic Acid (99%, Acros), Hydrochloric acid (HCl, 37%, Acros) were used without further purification. Absolute ethanol (99.9 %) prepared by keeping rectified ethanol on 3Å molecular sieves (30% w/v) overnight and then distilling. As substrates in the coating process Soda Lime glass microscope slides 76x26 mm (DG) were used. Coatings were performed using dip-coater (KSV Instruments DipCoater D). Commercial detergent RBS25 (Carl-Roth) was used for substrate cleaning. Deionised-distilled water was used for preparation of all aqueous solutions. 2.2 Synthesis of hybrid silica-APTES sols and coating procedure Hybrid APTES-silica sols in ethanol were prepared by acid catalyzed TEOS and APTES hydrolysis and condensation. To obtain different amount of amino groups on the coating various TEOS/APTES ratios were used in sol preparation step. For this purpose TEOS, APTES, HCl, and ethanol were mixed at molar ratios a : b : 0.04 : 38, respectively (where a and b indicats different molar fractions of TEOS and APTES). Sols were aged at room temperature for 24 hours. Silica coatings without APTES were prepared by mixing TEOS, HCl, H2O and ethanol in molar ratio 1 : 0.04 : 4 : 38, respectively. In order to remove dust and agglomerates the prepared sols were filtered through 0.2 µm Nylon membrane filter before coating procedure. Before dip-coating substrates were washed with 1% (v/v) RBS25 commercial detergent solution, followed by rinsing thoroughly with deionized water and drying in laminar flow cabinet. Coatings were prepared by Dip-Coating technique at the constant 40 mm/min withdrawal speed. Coatings were dried for 24 h at room temperature in laminar flow cabinet. Prepared coatings were cut to 15x25 mm size rectangles. Water contact angle was measured by KSV Instruments CAM200 optical goniometer. Coatings on a silicon substrate were prepared in the same way and the layer thickness was analyzed with rotating analyzer Spectral ellipsometer J.A. Woollam M2000X equipped with Complete EASE software. CaF2 monocrystal plate was used as a substrate for registering FT-IR spectra of coatings by using Perkin Elmer Frontier FT-IR spectrometer equipped with ATR module.

2.3 Synthesis of AgNPR colloid AgNPRs were fabricated by the seed-mediated method [31] which includes two steps – producing silver nanoseeds and then growing nanoprisms. Synthesis of silver seeds. Silver seeds obtained by mixing aqueous trisodium citrate (20 mL, 25 mM), PSSS (1 ml, 0.5 g/L) and freshly prepared NaBH4 (1.2 mL, 10 mM) solutions followed by addition of aqueous AgNO3 (20 mL, 0.5 mM). The addition rate of AgNO3 was 2 mL/min. Addition rate was maintained using syringe pump. Growing AgNPRs. AgNPRs are produced by mixing 35 ml of distilled water, aqueous ascorbic acid (1.050 mL, 10 mM) solution, silver seeds solution described above, followed by addition of aqueous AgNO3 (21 mL, 0.5 mM) at the rate of 5 mL/min. Excitation spectra of prepared AgNPRs solutions were registered by UV-Vis spectrometer Perkin Elmer Lambda 35. Particle size was determined using SEM Hitachi FE-SEM SU-70. Samples for SEM were prepared by drop-coating of AgNPRs solution onto silicon substrate and drying in air. Average particle size (tip-edge length) was calculated from at least 150 particles. 2.4 Deposition of AgNPR layers For deposition of AgNPRs layers, silica-APTES coatings were placed in plastic containers and 5 ml of AgNPR colloidal solution were added to each container and left to stand at room temperature for specified time. Coatings with deposited AgNPRs layers were gently washed with distilled water and left to dry in air at room temperature. In order to obtain reliable results, each deposition experiment was repeated at least 4 times. 2.5 SERS measurements Aqueous uric acid (Sigma-Aldrich) solutions of various concentrations were prepared by diluting a 1 mM stock solution (16.8 mg in 98 ml of water) with DI water. 2 ml of 2 % NaOH (Thermo Scientific) solution was added to aid dissolve the uric acid completely. Samples were prepared by immersing the substrates into uric acid solution for 1 h then rinsing with DI water and drying at ambient conditions. -1

SERS spectra in 4000-70 cm spectral range were recorded on FT-Raman spectrometer MultiRAM (Bruker Optik GmbH, Ettlingen, Germany) equipped with Nd:YAG (1064 nm) excitation laser of 100 µm diameter of the beam at its focal point. Liquid-nitrogen-cooled Ge diode and Si diode detectors were used to record the SERS signal. Gold plated mirror objective (focal length – 33 mm) was employed. For each spectrum 128 interferograms were acquired and averaged, followed by Fourier transformation. The data was further processed by applying Blackman-Harris 3-Term apodization function and zero filling factor of 2. -1 The spectral resolution of 4 cm was used. Laser power was set to 100 mW. Five to seven random spots in every sample were analysed. 3

RESULTS AND DISCUSSIONS

3.1 Preparation of hybrid silica-APTES coatings Hybrid silica coatings containing various amounts of APTES were prepared by increasing APTES molar fraction from 0.02 to 0.3 while decreasing TEOS molar fraction from 0.98 to 0.7, respectively. To obtain uniform coatings precise control of hydrolysis and condensation processes of the sol is neccessery. Due to natural APTES ability to induce hydrolysis-condensation reaction absolute ethanol was used and no additional water was added to the initial sol. When APTES molar fraction was kept above 0.06 visually uniform coatings were obtained. However, low magnification SEM images (Suppl. Fig S1) revealed the surface being coated with micron-sized droplets rather than forming a homogenous layer. Coatings prepared with the molar fraction of APTES above 0.06 are homogenous with a few minor defects caused by dust and glass substrates imperfections. It highly suggests that at low concentration of APTES, hydrolysis and condensation processes are not sufficient and may result in a non-uniform coating with segregated domains. Layer thickness and water contact angle measurements of the prepared hybrid coatings as a function of TEOS/APTES molar ratio are presented in Fig. 1. The obtained coatings are relatively thin and tend to thicken gradually from 13 to 22 nm as APTES molar fraction raises. The highest water contact angle was obtained of the coating with 0.06 molar fraction of APTES and it starts to decrease gradually by adding more APTES. Water contact angle is highly sensitive to the surface properties as polarity, porosity, and roughness. Normally, silica surface modification with APTES, due to less polar amine groups and increased roughness, results in increased surface hydrophobicity. Therefore, gradual decrease of contact angle might be related to the reduced surface roughness and/or decreased amount of amine groups. To clarify coatings were further studied by FTIR.

FTIR spectra (fig. 2) of the prepared coatings show high intensity of the corresponding of Si-O-Si -1 -1 vibrations at 1050 cm and medium intensity of Si-OH stretches at 789 cm . The characteristic primary -1 amine vibrations in the region 3400 – 3300 cm are hidden under absorbed H2O vibrations. Other weak NH2 -1 group vibrations at 1549 cm became more intense with increasing amounts of APTES. Moreover, the -1 intensity of the broad signal in the range 3500-3000 cm resulting from adsorbed H2O and well as Si-OH -1 stretch at 789 cm are reduced with increasing APTES molar fraction, which indicates decreased amount of -1 free Si-OH groups. The weak intensity of N-H stretch at 689 cm becomes visible at quite high APTES molar ratios, i.e. 0.1 and higher. The observation of increased characteristic N-H vibrations and reduced intensity of Si-OH bands by adding more APTES, let us to conclude, that increasing APTES amounts indeed results in increased amino group content on the glass surfaces and hence successful surface silanization procedure. 3.2 Synthesis of AgNPR colloids and characterization For the synthesis of AgNPRs seed-mediated method was selected, which allows easy control of AgNPRs size by changing seed concentration in the initial particle synthesis step. Three colloidal solutions of different size AgNPRs were prepared as described in the experimental section. In table 1 presented data shows, that lower amount of seed solution results in larger particle size and thus subsequent red-shifted LSPR peak. But narrower particle size distribution was obtained with higher amount of seed solution used. Prepared colloids were stable at ambient conditions and no dissolution or colour change were noticed within several weeks. Prismatic shape particles (Suppl. Fig S2-S4) dominate in all synthesized AgNPRs solutions although trace amounts of disc or hexagon shaped particles could be detected. Larger particles distinguish by more prismatic shape and higher size deviations while smaller particles have somewhat more irregular shape, but fewer size deviations. The size and shape deviations are inherent to the selected synthesis method where intentional crystal defects of Ag seeds determine the final shape of particles. Therefore, to reduce these deviations by seed-mediated synthesis is challenging. An alternative photo-mediated AgNPRs synthesis gives particles with narrower distribution in size and shape, however, it suffers from rather low reproducibility and are harder to scale up. On the other hand, high reproducibility of the seed-mediated methodology allows access to larger quantities of AgNPRs colloids necessary for deposition studies and weights out the wider size and shape distribution. The excitation spectra of the synthesized AgNPRs (Fig. 3) show characteristic high absorbance LSPR peaks of AgNPRs. Most intense LSPR peaks at longer wavelengths can be associated with dipole mode in particles, which is particularly sensitive to particles shape and size and their dielectric surrounding. Less intense quadrupolar oscillations are noticeable in the spectrum of largest particles in 550 – 750 nm range. Octupole mode at 333 nm is observed in all samples, which is the least sensitive to particle shape and size. An additional peak at 400 nm, observed in first two samples with smallest particles (JPSP001 and JPSP002), may be assigned to the residue of small spherical shape particles which did not grow into AgNPRs. Moreover, broadening of dipole mode peak is observed in the case of larger particles, which is attributed to higher polydispersity and scattering effects. Nevertheless, high-intensity dipole mode LSPR peaks and SEM images of the prepared colloidal sols confirms high yield of the synthesized AgNPRs. 3.3 Deposition of AgNPR layers on silica-APTES coatings Medium size AgNPR particles sample (AgNPR002) was selected to investigate the influence of APTES amount to the deposition rate, uniformity and surface pattern of deposited nanoparticle layers. Fig 4 represents dipole mode LSPR absorbance peak intensity as a function of AgNPRs deposition time. The deposition process was monitored for 24 hours. No deposition of AgNPRs onto glass substrate without APTES was observed. The deposition rates on coatings with APTES of molar fraction in the series of 0.060.1 are very similar and show consistent growth of particle layer over 24 hours monitored. Deposition rate highly increases when 0.2 molar fraction of APTES was used, but process slows down after 12 hours. With APTES molar fraction of 0.3 strong aggregation of AgNPRs took place throughout whole volume of the sample causing greyish precipitates on the substrate surface as well as container walls. Excessive aggregation could be seen already after 3 hours since the initial stage. AgNPRs layers on coatings with 0.060.1 molar fraction of APTES are visually uniform in all surface area (suppl. Fig S5) and due to strong LSPR in longer wavelengths have solid blue colour. A constant slight increase in colour intensity is also noticeable with increasing amount of APTES in this series. AgNPRs layer on coating with the highest amount of APTES (0.3) have greyish blue colour and contain both uncovered areas and islands of AgNPRs aggregates (Suppl. Fig. S6). With higher APTES content, AgNPRs tend to form micron-sized aggregates with high loss of LSPR. These observations clearly show that amount of APTES plays crucial role not only on the deposition rate of AgNPRs layer but also on the stability of colloidal AgNPRs solution. Since the best coatings with evenly aggregated AgNPRs were obtained with APTES molar ratio of 0.1 these conditions were chosen for further study on the effect of particle size for their deposition. In this case,

deposition process was monitored for a longer period of time – up to 48 hours. Extinction curves of the obtained AgNPRs layers as a function of deposition time are presented in Fig. 5a. At the beginning smaller AgNPRs tend to deposit on the surface more rapidly while larger particles show more constant growth of AgNPR layer. However, the overall deposition rates of all size particles are similar, resulting in just a slight difference in absorbance after 48 hours. It can be explained by different mobility of smaller particles compared with larger particles. Small AgNPRs are more mobile, therefore the probability them to be attached to the surface is higher. On the other hand, eventually deposition slows down as surface gets more covered by AgNPRs. As a result, the surface covering after 48 hours is similar in both cases despite small particle aggregation starts off faster. LSPR peak position of all AgNPRs deposited glass slides fall into 900 nm – 1100 nm range. Red-shift of LSPR peak was observed over time as amount of deposited AgNPRs increased (Fig. 5b and Suppl. Fig S7-S9). Surprisingly, the largest shift was detected when a colloid of the smallest particles (AgNPR01) was used, while using large particles resulted in less significant shift. These shifts can be associated with the coupling of LSPR dipole mode peak arising from AgNPRs aggregation when two or more plasmonic particles are in close distance - phenomena already described in detail by Kotkowiak M. [32]. SEM images of the obtained AgNPRs layers (Fig. 6) correlate well with deposition rate graphs. As compared to larger particles (AgNPR02 and AgNPR03), small particles (AgNPR01) deposit quicker resulting in high surface covering in just 3 hours since the start of the experiment. After 24 hours surface was highly covered by small particles while uncovered areas were still visible in the case of large particles. Full surface covering with large particles was achieved after 48 hours. After the same deposition duration small and medium (AgNPR01 and AgNPR02) particles started to form large irregular shape particles on top of uniform AgNPRs layer. Note, that these conglomerates were not detected in the stock solutions of AgNPRs which means that they were formed during the deposition process. The exact mechanism of formation of the irregularly shaped particles remains unclear and requires more detailed study. Deposition process does not stop after full surface covering - AgNPRs assembly continues and forms additional layer. It strongly implies, that electrostatic interaction between substrate decorated with amino groups and free AgNPRs is very strong. Particle adsorption and aggregation mechanism has already been studied in the case of spherical gold nanoparticles [33]. It was suggested that adsorbed particles have random mobility over the aminized surface and that hydrophobic nature of metal nanoparticle are main driving forces leading to their aggregation. Our results shares analogy to the proposed mechanism. 3.4 SERS of deposited AgNPR layers With new AgNPRs coatings in hand their performance in SERS was investigated. As the sample analytical object we selected uric acid. Owning to 4 nitrogen atoms in its molecular structure uric acid absorbs easily on silver nanoparticle surface. Uric acid is non-resonant molecule and does not enhance Raman signal itself. Thus, it is suitable analyte for evaluation of SERS enhancement of deposited AgNPRs layers. Furthermore, determination of uric acid by SERS has already been presented in other research works which results can be used for reference [34, 35]. Development of sensitive methods for uric acid sensing is extremely important, since its concentration deviations in the blood stream (>0.4 mM) might indicate serious medical conditions such as preeclampsia. The series of glass substrates with different AgNPRs layers was subjected to SERS enhancement studies. Uric acid concentration was kept very low (0.1 mM) and no Raman signal could be detected of the plain sample. Fig. 7 represents SERS intensity versus deposition time of layers consisting different size AgNPRs. No signal was observed with the coatings deposited only for 3 hours. Only a weak signal was detected on coatings deposited for 6 hours with the smallest particles in the series (AgNPR01). In the case of AgNPR02 and AgNPR03, SERS signal was observed only with coatings which were deposited for 12 hours. Coatings assembled from small AgNPRs (AgNPR01) show rather low SERS enhancement, which decreases even more with increasing AgNPRs deposition time. Very similar behaviour can be seen in the case of medium size particles. Highest SERS enhancement was observed on coatings with large particles (AgNPR03) deposited for 24 hours. Though the layer of small and medium AgNPRs have LSPR peak approximately at 1000 nm, which is near to the wavelength of the laser used, the SERS signal intensity remains low. High SERS enhancement is indeed associated with LSPR of individual particle and not to collective LSPR of aggregated particles. On the other hand, solutions of small particles have a lower fraction of well-defined shape AgNPRs, which might also be responsible for their lower SERS sensitivity. It suggests that small aggregated particles offer smaller surface area for interaction with analyte molecule than individual particles, which in turn results in lower intensity of SERS signal. Analytical enhancement factors (AEF) were calculated using the reported method [36]. In the most successful case in the series i.e. the sample 4 AGNPR003 with 24 hours deposition AEF was found to be 2.7 x 10 . As to compare to other reports the AEF

value obtained is rather low, however, present study reveals the influence of amine group content, deposition time and particle size on AgNPRs behaviour when particles are in contact to aminized surface.

4

CONCLUSIONS

In this work, uniform hybrid silica coatings containing different APTES amounts were prepared via sol-gel process and utilized for the assembly of AgNPRs. It was found, that amount of APTES is crucial for deposition rate and behaviour of AgNPRs - high APTES concentration leads to rapid aggregation and particle dissolution. Nevertheless, deposition and aggregation rates can be slightly adjusted when moderate amount of APTES is used (0.06-0.2 molar ratio of APTES). Excitation spectra and SEM images of deposited AgNPRs layers showed that in comparison with large particles (120 nm), small AgNPRs (34 nm) have higher deposition and subsequent aggregation rates which result in broadening as well as strong red-shift of LSPR dipole mode band. The deposition rate of large particles is lower which gives less aggregated AgNPRs layer and only slight shift of LSPR band. SERS study revealed, the highest enhancement occurs on coatings containing large particles where LSPR band is near laser excitation wavelength (1064 nm). Due to the coupling of LSPR dipole mode of aggregated particles, final LSPR position of coatings deposited from small particles was near laser excitation wavelength. Thus, the registered SERS signal was much weaker comparing with large particles as excessive aggregation reduce the SERS performance. It lets to conclude that SERS enhancement depends more on individual particle LSPR and not collective LSPR of aggregated particles. Although comparing to other reports [37-39] AEF achieved in this study is considerably lower, it provided a deeper insight into AgNPRs assembly events on aminized surfaces and its importance to SERS enhancement.

Acknowledgement. The study was funded by the European Community's social foundation under Grant Agreement No. VP1-3.1-ŠMM-08-K-01-004/KS-120000-1756 References

[1] P.K. Jain, X. Huang, I.H. El-Sayed, M.A. El-Sayad, Review of some interesting surface plasmon resonanceenhanced properties of noble metal nanoparticles and their applications to biosystems, Plasmonics, 2 (2007) 107-118. [2] T.K. Sau, A.L. Rogach, F. Jackel, T.A. Klar, J. Feldmann, Properties and Applications of Colloidal Nonspherical Noble Metal Nanoparticles, Adv. Mater., 22 (2010) 1805-1825. [3] A. Shiohara, Y.S. Wang, L.M. Liz-Marzan, Recent approaches toward creation of hot spots for SERS detection, J Photoch Photobio C, 21 (2014) 2-25. [4] N. Marquestaut, Y. Petit, A. Royon, P. Mounaix, T. Cardinal, L. Canioni, Three-Dimensional Silver Nanoparticle Formation Using Femtosecond Laser Irradiation in Phosphate Glasses: Analogy with Photography, Adv Funct Mater, 24 (2014) 5824-5832. [5] G.T. Forcherio, D.K. Roper, Optical attenuation of plasmonic Au-PDMS nanocomposite thin-film devices, Nanophotonic Materials Xi, 9161 (2014). [6] J. Kim, Z. Wang, W.S. Kim, Stretchable RFID for Wireless Strain Sensing With Silver Nano Ink, Ieee Sens J, 14 (2014) 4395-4401. [7] H.I. Park, S. Lee, J.M. Lee, S.A. Nam, T. Jeon, S.W. Han, S.O. Kim, High Performance Organic Photovoltaics with Plasmonic-Coupled Metal Nanoparticle Clusters, Acs Nano, 8 (2014) 10305-10312. [8] H. Youn, T. Lee, L.J. Guo, Multi-film roll transferring (MRT) process using highly conductive and solutionprocessed silver solution for fully solution-processed polymer solar cells, Energ Environ Sci, 7 (2014) 27642770. [9] V. Bastys, I. Pastoriza-Santos, B. Rodríguez-González, R. Vaisnoras, L.M. Liz-Marzán, Formation of Silver Nanoprisms with Surface Plasmons at Communication Wavelengths, Adv Funct Mater, 16 (2006) 766-773. [10] N. Bhalla, N. Formisano, A. Miodek, A. Jain, M. Di Lorenzo, G. Pula, P. Estrela, Plasmonic ruler on fieldeffect devices for kinase drug discovery applications, Biosens Bioelectron, 71 (2015) 121-128. [11] C.S. Kumar, M.D. Raja, D.S. Sundar, M.G. Antoniraj, K. Ruckmani, Hyaluronic acid co-functionalized gold nanoparticle complex for the targeted delivery of metformin in the treatment of liver cancer (HepG2 cells), Carbohyd Polym, 128 (2015) 63-74.

[12] W.P. Savarimuthu, P. Gananathan, A.P. Rao, E. Manickam, G. Singaravelu, Protoporphyrin IX-Gold Nanoparticle Conjugates for Targeted Photodynamic Therapy-An In-Vitro Study, J Nanosci Nanotechno, 15 (2015) 5577-5584. [13] J.H. Zhao, J. Lin, X.H. Li, G.N. Zhao, W.J. Zhang, Silver nanoparticles deposited inverse opal film as a highly active and uniform SERS substrate, Appl Surf Sci, 347 (2015) 514-519. [14] L.J. Tang, S. Li, F. Han, L.Q. Liu, L.G. Xu, W. Ma, H. Kuang, A.K. Li, L.B. Wang, C.L. Xu, SERS-active Au@Ag nanorod dimers for ultrasensitive dopamine detection, Biosens Bioelectron, 71 (2015) 7-12. [15] O. Olea-Mejia, M. Fernandez-Mondragon, G. Rodriguez-de la Concha, M. Camacho-Lopez, SERS-active Ag, Au and Ag-Au alloy nanoparticles obtained by laser ablation in liquids for sensing methylene blue, Appl Surf Sci, 348 (2015) 66-70. [16] M.H. Li, X.Y. Guo, H. Wang, Y. Wen, H.F. Yang, Rapid and label-free Raman detection of azodicarbonamide with asthma risk, Sensor Actuat B-Chem, 216 (2015) 535-541. [17] S. Emamian, A. Eshkeiti, B.B. Narakathu, S.G.R. Avuthu, M.Z. Atashbar, Gravure printed flexible surface enhanced Raman spectroscopy (SERS) substrate for detection of 2,4-dinitrotoluene (DNT) vapor, Sensor Actuat B-Chem, 217 (2015) 129-135. [18] W. Ruan, T. Zhou, Y. Cui, Y. Dong, Z. Liu, F. Dong, H. Wang, X. Luan, X. Wang, W. Song, B. Zhao, Submicron patterns obtained by thermal-induced reconstruction of self-assembled monolayer of Ag nanoparticles and their application in SERS, Appl Surf Sci, 309 (2014) 295-299. [19] A.Y. Li, Z. Baird, S. Bag, D. Sarkar, A. Prabhath, T. Pradeep, R.G. Cooks, Using Ambient Ion Beams to Write Nanostructured Patterns for Surface Enhanced Raman Spectroscopy, Angew Chem Int Edit, 53 (2014) 12528-12531. [20] S. Schlucker, Surface-enhanced Raman spectroscopy: concepts and chemical applications, Angewandte Chemie, 53 (2014) 4756-4795. [21] S. Bai, W.P. Zhou, C. Tao, K.D. Oakes, A.M. Hu, Laser-processed Nanostructures of Metallic Substrates for Surface-Enhanced Raman Spectroscopy, Curr Nanosci, 10 (2014) 486-496. [22] B. Sharma, R.R. Frontiera, A.-I. Henry, E. Ringe, R.P. Van Duyne, SERS: Materials, applications, and the future, Mater Today, 15 (2012) 16-25. [23] S.L. Kleinman, E. Ringe, N. Valley, K.L. Wustholz, E. Phillips, K.A. Scheidt, G.C. Schatz, R.P. Van Duyne, Single-Molecule Surface-Enhanced Raman Spectroscopy of Crystal Violet Isotopologues: Theory and Experiment, J. Am. Chem. Soc., 133 (2011) 4115-4122. [24] M. Fan, G.F.S. Andrade, A.G. Brolo, A review on the fabrication of substrates for surface enhanced Raman spectroscopy and their applications in analytical chemistry, Anal Chim Acta, 693 (2011) 7-25. [25] D.-K. Lim, K.-S. Jeon, J.-H. Hwang, H. Kim, S. Kwon, Y.D. Suh, J.-M. Nam, Highly uniform and reproducible surface-enhanced Raman scattering from DNA-tailorable nanoparticles with 1-nm interior gap, Nat Nanotechnol, 6 (2011) 452-460. [26] V. Torres, M. Popa, D. Crespo, J.M. Calderón Moreno, Silver nanoprism coatings on optical glass substrates, Microelectron Eng, 84 (2007) 1665-1668. [27] C. Xue, X. Chen, S.J. Hurst, C.A. Mirkin, Self-Assembled Monolayer Mediated Silica Coating of Silver Triangular Nanoprisms, Adv. Mater., 19 (2007) 4071-4074. [28] C. Xue, Z. Li, C.A. Mirkin, Large-Scale Assembly of Single-Crystal Silver Nanoprism Monolayers, Small, 1 (2005) 513-516. [29] M. Zhu, M.Z. Lerum, W. Chen, How To Prepare Reproducible, Homogeneous, and Hydrolytically Stable Aminosilane-Derived Layers on Silica, Langmuir, 28 (2011) 416-423. [30] J. Pilipavicius, A. Chodosovskaja, A. Beganskiene, A. Kareiva, Silver nanoprisms self-assembly on differently functionalized silica surface, IOP Conf Ser: Mater Sci Eng, 77 (2015) 012006. [31] D. Aherne, D.M. Ledwith, M. Gara, J.M. Kelly, Optical Properties and Growth Aspects of Silver Nanoprisms Produced by a Highly Reproducible and Rapid Synthesis at Room Temperature, Adv Funct Mater, 18 (2008) 2005-2016. [32] M. Kotkowiak, B. Grześkiewicz, E. Robak, E. Wolarz, Interaction between Nanoprisms with Different Coupling Strength, J Phys Chem C, 119 (2015) 6195-6203. [33] A.F. Scarpettini, A.V. Bragas, Coverage and Aggregation of Gold Nanoparticles on Silanized Glasses, Langmuir, 26 (2010) 15948-15953.

[34] S. Yang, Z. Zhang, J. Zhao, Z. Yu, H. Jiang, Raman spectroscopy study of the effect of urea, uric acid and creatinine on steric configuration of bovine hemoglobin using SERS-active BN nanosheets/Ag nanoparticles hybrids, Mater Lett, 131 (2014) 78-81. [35] L. Zhao, J. Blackburn, C.L. Brosseau, Quantitative Detection of Uric Acid by Electrochemical-Surface Enhanced Raman Spectroscopy Using a Multilayered Au/Ag Substrate, Anal Chem, 87 (2015) 441-447. [36] E.C. Le Ru, E. Blackie, M. Meyer, P.G. Etchegoin, Surface Enhanced Raman Scattering Enhancement Factors:  A Comprehensive Study, The Journal of Physical Chemistry C, 111 (2007) 13794-13803. [37] W.J. Cho, Y. Kim, J.K. Kim, Ultrahigh-Density Array of Silver Nanoclusters for SERS Substrate with High Sensitivity and Excellent Reproducibility, Acs Nano, 6 (2012) 249-255. [38] F. Shao, Z.C. Lu, C. Liu, H.Y. Han, K. Chen, W.T. Li, Q.G. He, H. Peng, J.N. Chen, Hierarchical Nanogaps within Bioscaffold Arrays as a High-Performance SERS Substrate for Animal Virus Biosensing, Acs Appl Mater Inter, 6 (2014) 6281-6289. [39] Q.Q. Su, X.Y. Ma, J. Dong, C.Y. Jiang, W.P. Qian, A Reproducible SERS Substrate Based on Electrostatically Assisted APTES-Functionalized Surface-Assembly of Gold Nanostars, Acs Appl Mater Inter, 3 (2011) 1873-1879.

FIGURES

Fig 1 Water contact angle (black line) and thickness (blue line) of the prepared coatings versus APTES molar fraction Fig 2 FTIR spectra of the prepared coatings by varying APTES molar fraction Fig 3 Excitation spectra of synthesized AgNPR colloids Fig 4 AgNPRs deposition rate curves of silica-APTES coatings prepared by using various APTES molar fractions Fig 5 AgNPRs deposition curves, displaying: a) deposition rate b) LSPR peak position shift against deposition time. Fig 6 SEM images of deposited various size AgNPRs against different deposition time (large scale bar 1 µm, small scale bar 100 nm) Fig 7 SERS spectra of AgNPR03 deposited layer (a) and SERS integral intensity at 500 cm-1 (CN bending/in-plane ring deformation) against deposition time of all AgNPR layers

This paper is considered to be published online, no printed version is requested.

Table 1 Experimental details of the prepared AgNPRs colloids Sample name AgNPR01 AgNPR02 AgNPR03

Added amount of seed solution, µL 2000 1000 200

LSPR peak position, nm 664 756 1000

Average tipedge length, nm 34 ±8 50 ±11 120 ±25

Table 1 Experimental details of the prepared AgNPRs colloids

Sample name AgNPR01 AgNPR02 AgNPR03

Added amount of seed solution, µL 2000 1000 200

LSPR peak position, nm 664 756 1000

Average tipedge length, nm 34 ±8 50 ±11 120 ±25

graphical_abs1 .