WO2008128521A1

WO2008128521A1 - Method for the production of substrates for surface-enhanced raman spectroscopy

Info

Publication number: WO2008128521A1
Application number: PCT/DE2008/000650
Authority: WO
Inventors: Mohammed Es-Souni
Original assignee: Fachhochschule Kiel
Priority date: 2007-04-20
Filing date: 2008-04-17
Publication date: 2008-10-30
Also published as: EP2140248A1; DE102007019166A1

Abstract

Disclosed is a method for producing a substrate for surface-enhanced Raman spectroscopy. Said method comprises the following steps: a sol is produced from a stabilized solution containing precious metal ions and a precursor solution for a titanium oxide; a heat-resistant support is coated by applying the sol by means of a sol-gel process; the layer is pyrolyzed and sintered without using light; and the layer produced without using light is illuminated at least in subareas while at least the illuminated subareas of the layer are simultaneously heated.

Description

Process for the preparation of substrates for surface-enhanced Raman spectroscopy

The invention relates to a method for producing substrates for surface-enhanced Raman spectroscopy.

Raman spectroscopy plays an important role among the known methods for investigating material properties. Raman scattering refers to the inelastic scattering of light to matter (eg crystals or molecules), in which a portion of the incident photon energy leads to the excitation of lattice or molecular vibrations ("Stokes scattering", if the sample is in the ground state before light incidence) Also, the reverse energy transfer and thus a blue shift of the scattered light ("anti-Stokes") occurs - albeit less frequently. The energy transfers are characteristic of the material and lead to defined

Wavelength shifts of the Raman-scattered light.

On the basis of Raman spectra in particular the identification of molecules is possible, which occur in low concentration in a solution. Thus, Raman spectroscopy is fundamentally a very robust process for chemical analysis, for example in biotechnology, pharmacy ("drug discovery"), medical diagnosis or in environmental monitoring.

However, the classical Raman signal is very weak, because only less than 0.001% of the incident light is Raman scattered. But as early as 1974, Fleischmann and McQuillan (Chem. Phys. Lett 26 (1974) 123) described the dramatic increase in signal yield in the analysis of pyridine on a rough silver electrode as a substrate. In later work, it was confirmed that even the presence of a metal surface in the vicinity of a molecule to be excited to vibrate is sufficient to amplify the Raman signal by several orders of magnitude. The reason lies in the possibility of resonant excitation of plasmon in the metal, which then leads via electric field interaction (dipole-dipole) in turn to the enhanced excitation of molecular vibrations. It has proven to be advantageous if local electric field peaks can be formed ("hot spots"), so that no smooth metal surfaces are used, but preferably rough surfaces.Theoretically, signal amplifications up to 10 ^{12 can be} achieved in this way (Surface Enhanced Raman - SER) Reinforcements of about 10 ^{6 are} common on today commercially available SER substrates.

In addition to rough surfaces, smooth non-metallic surfaces with sporadic metal tips (e.g., adsorbed clusters) are also likely to cause SER amplification. Another prerequisite for the SER effect is that the metal tips show plasmon resonance in the wavelength range from visible light (VIS) to infrared (IR), which can be excited with common lasers. This is the case in particular for precious metals such as gold and silver, which are preferably also used as nanoparticle material to produce SER-active substrates.

Many known processes for the preparation of SER substrates pursue the application of metal colloids to non-metallic substrates, e.g. Semiconductor or glass, such as by dropping and drying of colloid suspensions, by laser deposition or by vapor deposition of metal under vacuum on a masked or shadowed substrate. The aim is to produce evenly spaced, over the surface of the substrate uniformly distributed metal particles in a regular arrangement. It is known from the literature (Lidong Qi et al., PNAS 103 (6), 2006, pp. 13300-13303) that particle diameters around 120 nm and mean particle spacings around 30 nm provide the best SER gains.

As an example of a masking which seems particularly suitable for SER substrates, mention should be made of nanosphere lithography (NSL). In this case, spheres with a diameter of several 100 nanometers are preferably arranged on the substrate in a monolayer with dense sphere packing. Then metal is evaporated and the

Bullets are removed. In the interstices coarsely triangular, regularly arranged metal islands have formed thereafter, which can act as metal tips for SER.

Of course, the use of templates with structure sizes on the nanometer scale is always particularly complex. Specifically, at the NSL is z. B. the problem of preventing the formation of multiple layers of the balls.

In order to further develop SER spectroscopy into a widely used standard of trace analysis, substrates with the following properties should be available: Economically producible in large numbers;

Excellent reproducibility in terms of SER signal amplification; Chemically inert or robust against aggressive chemicals - arrays of SER substrates should be producible with little effort.

The requirement of low-cost manufacturing currently makes all template methods, at least for mass production, not promising. Rather, one will be content with the production of irregularly arranged, but at least statistically evenly distributed metal particles on a non-metal in order to limit the effort.

The document WO 2006/060734 A2 pursues this approach and proposes that the metal particles first grow in situ on the substrate - namely from metal ions which are located in the matrix forming the substrate from the outset. To this end, the metal ions (e.g., gold) are added to a precursor solution for a suitable oxide (e.g., silica), which is then divided into equal portions and dried at ambient temperature for 12 to 48 hours. These pellets are treated with a reducing agent in two steps, first reducing the metal ions rapidly to form particle growth nuclei on the pellet surface, so that in the subsequent weaker reduction phase (duration: up to 12 hours), particles on these nuclei gradually become to grow. WO 2006/060734 A2 attaches great importance to the fact that approximately equal-sized ("monodisperse-sized") metal particles are formed which are located on the surface of the matrix substrate and are not embedded therein.

The described method of WO 2006/060734 A2 is clearly too time-consuming, not least because of the relatively slow treatment with the reducing agent, which additionally has to be disposed of.

It is an object of the invention to provide a method which enables a fast, reproducible and cost-effective production of SER substrates.

The object is achieved by a method having the features of claim 1. The subclaims indicate advantageous embodiments of the invention. The process according to the invention is based firstly on the teaching of WO 2006/060734 A2, in that here too an additional cursor containing precursor for the production of a metal oxide film via a sol-gel process is used. However, for the realization of the invention, the precursor must necessarily contain titanium in order to be able to form a titanium oxide matrix, preferably TiO ₂ .

As noble metal ions, very particular preference is given to using silver ions. Although other precious metals should not be excluded, there are no studies available yet.

The invention now unexpectedly continues the known process by first applying the silver ion-containing titanium precursor to a heat-resistant substrate (eg, glass, semiconductor, metal) by a sol-gel method (especially spinning, spraying, dipping) where it is immediately pyrolyzed and sintered with exclusion of light. After the heat treatment, the layer is dry and hard and largely resistant to chemical attack. It has virtually no silver nanoparticles on the surface and is therefore not suitable as a SER substrate. Further processing according to the teaching of WO 2006/060734 A2 is also not suitable for improving this. The layer can be stored very well when stored in the absence of light.

If the silver-containing titanium oxide layer is now intensively illuminated with sufficient heating, electron-hole pairs are generated in the matrix according to FIG

Present silver ions have a strong tendency to absorb the released electrons.

Ag ⁺ + Ie ^" → Ag - metal

It form smallest particles of silver, which have a certain mobility in the same heated matrix. You can also connect by diffusion to larger particles. Silver in the immediate vicinity of the layer surface passes through it and forms silver nanoparticles on the surface. When the resulting silver-titania nanocomposite cools back to ambient temperature, the silver particle distribution is virtually "frozen" with light and heat treated, originally SER-inactive layer is then suitable as a SER substrate, showing an amplification factor that can compete with commercially available substrates.

The invention will be explained in more detail below with reference to an exemplary embodiment and to the associated figures. Showing:

Figure 1 is a scanning electron micrograph of the titanium oxide layer, which has the highest measured SER activity, with the silver particles thereon;

2 shows the results of two Raman measurements of a known dye (R6G), once on the substrate according to the invention (upper curve) and on the not yet exposed, under the exclusion of light pyrolyzed and sintered substrate.

There follows an embodiment which refers to the figures.

First, a TiO ₂ -Ag precursor solution (Sol) is prepared. The silver content should be between 10% and 60% by weight based on the total mass of the (after pyrolysis and sintering) dried layer. Preferably, the Ag mass fraction is established between 30% and 60%. In the following example, it is about 50%, which is considered to be particularly advantageous.

For the preparation of 100 ml of an approximately 0.6 molar solution, first 10 ml of 2-

Methoxyethanol and acetylacetone presented in a beaker. Then the titanium isopropoxide is added, followed by stirring for 30 minutes. As a second solution, mix 10 ml of 2-methoxyethanol with water. After stirring for 30 minutes, the hydrous solution is added to the Ti-acetylacetone complex. Let it stir again for 30 minutes. For the silver solution, 10 ml of 2-methoxyethanol are placed in a beaker and AgNO ₃ and pyridine (as a stabilizer) are added. This complex must also be stirred for 30 minutes. Thereafter, the silver solution may be added to the stabilized and hydrolyzed titanium solution. After stirring again for 30 minutes, 2 g of polyethylene glycol 400 are added to the solution, made up to 100 ml with 2-methoxyethanol and then filtered. The polyethylene glycol serves the crack-free Layer formation. The exact weights, including mass, volume or molar information can be found in the following table.

Ti-iso. = Ti isopropoxide

Hacac = acetylacetone

Stoichiometries used: - Ti isoprop. : Hacac: H ₂ O = 1: 0.5: 4 (mol) - AgNO ₃ : pyridine = 1: 15 (mol)

The Tiθ ₂ -Ag layers are produced by spin coating. The carrier used is exemplified oxidized silicon. The pyrolysis of the layers is then carried out at 250 ° C. The final treatment temperature (sintering step) is between 450 and 550 ° C. The thickness of the layers can be between 50 and 1000 nm. In the case of thermal treatment, it is essential to ensure that it is carried out in the absence of light (UV to the end of VIS). This is the only way to avoid any uncontrolled silver precipitation / reduction. After production, the substrates can be exposed immediately or, if necessary, stored in the dark. The exposure can be done with virtually all conventional lamps that emit light in the spectral range of 280 to 800 nm. The TiO ₂ matrix absorbs very well over the entire visible range, which is why TiO _{2 is} also known as a solar absorber. If the lamp emits heat at the same time which heats the layer, SER activation can already be initiated by forming the silver particles on the layer surface. Alternatively, lower power light sources (eg, laser or light emitting diodes) may be used in combination with a heat source that allows a temperature of up to 250 ° C.

The adjustment of the particle size and distribution takes place through the interaction of exposure and heat. The exposure leads to the reduction of the silver from silver oxide to elemental silver, the heat to the coarsening of the particles by diffusion.

The heat supply should be set up for this purpose so that the layer has at least temperatures above 80 ° C during illumination. Temperatures between about 150 ° C and 250 ° C are preferred. However, it has proved to be unfavorable to use temperatures above 250 ⁰ C during illumination, because then the mobility of the silver particles would be too large. There may be quite uneven particle distributions on the surface of the layer that interfere with SER activity.

Typical parameters for obtaining a substrate suitable for SER spectroscopy are e.g.

Green light (550 nm) at 150 ⁰ C. Exposure time 1 hour UV light (300 to 350 nm) at 200 ° C. Exposure time 30 minutes.

In the illuminated area of the layer, finely distributed silver particles are formed

Sizes and (mean) distances to each other depend on the treatment time. Fig. 1 shows the scanning electron microscope Aufhahme a layer which is illuminated for one hour with a halogen lamp, power consumption 1000 W, from close range. The surface temperature achieves up to 250 ⁰ C. It is NEN good to recognize that a relatively wide particle size distribution (diameter about 100 ± 40 nm) is established, which however occurs anywhere on the substrate in about the same way. E- benso the mean distances between the particles are relatively small and almost the same across the area. It should be emphasized that the SER substrates produced here are anything but "monodisperse-sized" within the meaning of WO 2006/060734 A2, which is not surprising given the substantial differences in the methods.

The SER activity of the layer of Fig. 1 can be evaluated with an argon Raman spectrometer at a wavelength of 514.5 nm. A color molecule (R6G, 10.sup.- ⁶ mol) whose Raman lines are known is analyzed and as a result gives FIG. 2 in which the Raman shift as a change of wavenumber versus incident radiation (k = 2π / λ = 122122 cm ^"1 ) is indicated. The dashed lines show the known line positions. The lower trace is recorded on the still unexposed, silver-ion-containing titanium oxide and shows virtually no Raman signal. The upper curve is formed on the layer of FIG. 1 and clearly shows the Raman spectrum of the sample. The amplification factors are in the range of 5 × 10 ⁶ for the substrate of FIG. 1 and thus reach a maximum of

Achievable values for large area Raman spectroscopy.

Further, because the unexposed, non-photodetected substrate with the film thereon exhibits no appreciable SER activity, it is proposed to generate an array of SER-active spots on this film by using the light and heat treatment of the present invention performed by means of an optical mask. In principle, any light-absorbing or reflective masking is suitable (depending on the light source). In the simplest case, you can put a simple, mechanically made template with recesses on the substrate, which is removed after exposure. When very small feature sizes are desired, known masking techniques, e.g. from the semiconductor technology, which can be provided with a microstructure. The masking layer may, after processing, be e.g. be removed chemically. However, it may also remain on the substrate if it does not contribute to the SER gain itself. The processing of all separate spots (non-masked subareas of the substrate) preferably takes place simultaneously (if necessary fan out light by means of lenses), so that the processing parameters are identical.

As expected, an array made in this way shows very good agreement of the properties of different spots. Of course, each array can also be cut into individual sample carriers. Since the sol-gel technique is used anyway for the larger flat coating, mass production of SER substrates is likely to result in making relatively large SER spot arrays and cutting them as needed.

In summary, the method according to the invention has the following advantages over the prior art:

• Only standard industrial processes (sol-gel coating, pyrolysis, masking, exposure) are used, which can be carried out at correspondingly high speeds. • Titanium oxide layers are widely used and are known to be chemically stable. Disposal channels for used SER substrates are thus already available.

Apart from the inevitable burnout of the organics in pyrolysis, no further chemicals are used and released, i. E. there is no new disposal problem.

• The reproducibility of the SER substrates is - as always - a matter of precise process control. The processes used here are all mastered by the industry and do not require any new developments.

Claims

1. A method for producing a substrate for surface enhanced (Raman) spectroscopy comprising the steps of: i) preparing a sol from a stabilized solution containing noble metal ions and a precursor solution for a titanium oxide, ii) coating a heat-resistant support by applying the sol with a sol-gel method, iii) pyrolyzing and sintering the layer in the absence of light, and iv) illuminating the exposed-to-light layer at least on partial areas while simultaneously heating at least the illuminated partial areas of the layer ,

2. The method according to claim 1, characterized in that the illumination of the layer produced under the exclusion of light takes place simultaneously on a plurality of spatially separated partial surfaces.

3. The method according to any one of the preceding claims, characterized in that the layer produced under the exclusion of light is partially covered before lighting with an opaque stencil having recesses.

4. The method according to any one of the preceding claims, characterized in that between the pyrolysis and sintering in the absence of light and illuminating the layer produced under the exclusion of light a storage time is provided with the exclusion of light.

5. The method according to any one of the preceding claims, characterized in that the illumination of the layer produced under the exclusion of light takes place while heating at least the illuminated partial surfaces to temperatures between 80 ° C and 250 ⁰ C.

6. A process for the preparation of a photoexpositable intermediate for the manufacture of a substrate for surface enhanced Raman spectroscopy, characterized by the steps of: i) preparing a sol from a stabilized solution containing noble metal ions and ii) coating a heat-resistant support by applying the sol with a sol-gel method, and iii) pyrolyzing and sintering the layer in the absence of light.

7. The method according to any one of the preceding claims, characterized in that silver ions are used as noble metal ions, wherein the mass proportion of silver based on the mass of the dried layer between 10% and 60% is established.

8. The method according to any one of the preceding claims, characterized in that the solution containing silver ions is stabilized by the addition of pyridine.

9. The method according to any one of the preceding claims, characterized in that the pyrolyzing under exclusion of light at temperatures around 250 ⁰ C and the sintering in the absence of light at temperatures between 450 ° C and 550 ⁰ C.