SERS Detection Dictated By Substrate Fabrication

Surface-enhanced Raman scattering (SERS) has emerged as a powerful analytical tool that extends the possibilities of vibrational spectroscopy to solve a vast array of chemical and biochemical problems. SERS is an extension and variation of standard Raman spectroscopy, a vibrational spectroscopic technique that provides detailed information about the materials under investigation at the molecular level. Since the discovery of the SERS effect in the 1970s, SERS has been applied to a wide variety of analytical applications, including biochemistry and the life sciences. As has been reviewed in detail elsewhere, two primary mechanisms are believed to be responsible for SERS enhancement: a long-range classical electromagnetic (EM) effect and a short-range chemical (CHEM) effect. These two mechanisms contribute simultaneously to the overall enhancement; EM is thought to contribute the most (10–10) to the observed intensity enhancement, while CHEM is thought to contribute a lesser amount (10–10).

Since SERS is useful for determining molecular structural information and also provides ultrasensitive detection limits, including single molecule sensitivity, it has been used to detect pathogens that include bacteria and viruses. There are two principle SERS configurations that have been used in biosensing, intrinsic or extrinsic, as shown in fig 1. In intrinsic detection (Fig 1a), the analyte can be directly applied to the nanostructured surfaces and the inherent Raman spectrum of the biomolecule directly measured to identify the specimen. To allow for capture and to aid specificity of detection, antibodies, aptamers, or related molecules can be immobilized onto nanostructured surfaces as shown in fig 1b and the Raman spectral differences before and after capture of the specimen can be used to identify the species. In extrinsic detection, a Raman reporter molecule is used to generate a signal for detection. For example, a Au nanoparticle may be used as the SERS-active substrate to which a Raman reporter molecule is immobilized (fig 1c ). By coating this structure with another layer of dielectrics such as SiO₂, TiO₂, or a polymer, a core-shell complex is formed in which the outer-shell may be decorated with capture molecules such as antibodies. Thus, specimens may be captured and detected via a sandwich structure as shown in fig 1c . This extrinsic SERS detection method has been successfully used for in vivo SERS imaging of unique or rare cancer cells.

Fig. 1. Different SERS detection configurations: (a) Direct intrinsic detection; (b) indirect intrinsic detection; and (c) extrinsic detection.

The remarkable analytical sensitivity of SERS has yet to be translated into the development of widely accepted, commercially viable diagnostic applications, due in large part to the difficulty in preparing robust, metal-coated substrates of the correct surface morphology that provide maximum SERS enhancements. Some of the important requirements for an ideal SERS substrate in practical diagnostic applications are that the substrate produces a high enhancement, generates a reproducible and uniform response, has a stable shelf-life, and is simple to fabricate.

Many substrate preparation techniques exist that can form roughened metal surfaces of the types required for ideal SERS enhancements. These methods include roughening of a surface by oxidation-reduction cycles (ORC), metal colloid hydrosols, laser ablation of metals by high-power laser pulses, chemical etching, roughened films prepared by Tollen's reagent, photodeposited Ag films on TiO₂, and vapor-deposited Ag metal films. While the majority of substrate preparation techniques reported to date focus on the problem of achieving large SERS enhancements, the other requirements listed above for the production of a practical SERS sensing substrate are seldom addressed. Currently, there are five fabrication techniques that could potentially produce the desired SERS substrates to meet these requirements: electron beam lithography, nanosphere lithography, the template method, the hybrid method, and an oblique angle vapor deposition method.

The electron beam lithography (EBL) method is an ideal method for producing uniform and reproducible SERS substrates. Unfortunately, it is very expensive to produce large area substrates using EBL, unless the technique is combined with a nanoimprint lithography method. The nanosphere lithography (NSL) method pioneered by Van Duyne and coworkers involves evaporating Ag onto preformed arrays of nanopore masks by colloid particles, which are subsequently removed, leaving behind the Ag metal deposited in the interstices to form a regular Ag nanoparticle array. The template method utilizes a nanotube-like array, such as anodized Al₂O₃, as a template to deposit Ag or Au nanorods directly into the channels via an electrochemical plating method. Hybrid methods fabricate SERS substrates by depositing metal particles onto nanoporous scaffolds such as porous silicon, nanorod arrays, etc. The oblique angle deposition (OAD) method is based on a conventional physical vapor deposition principle and can be used to fabricate aligned and tilted Ag nanorod arrays on large substrate areas. For OAD fabrication, the surface normal of the substrate in a vacuum chamber is positioned at a very large angle with respect to the incoming vapor direction (> 75^o), as shown in fig 2a This deposition configuration results in a so-called geometric shadowing effect that leads to a preferential growth of nanorods on the substrate in the direction of deposition. The nanorods grow aligned but tilted on the substrate as shown in fig 2b. The benefits and limitations of the different methods for preparation of SERS-active substrates are summarized in table 1

Fig. 2. (a) Schematic of oblique angle deposition. (b) Top view, and (c) cross-sectional view scanning electron micrographs of the Ag nanorod array SERS substrates.

Table 1.

Comparison of different SERS substrate fabrication techniques that could potentially generate uniform and large area SERS substrates

Fabrication Method	EBL	NSL	Template method	Hybrid Method	OAD
Enhancement Factor	–	107–10 9	106–10 7 (50,52,53)	106–10 8 (59)	>108 (64)
Substrate Area (cm2)	Typically <>	1 x 1	> 2.5 x 2.5	> 2.5 x 5.0	> 2.5 x 7.5
Uniformity (%)	–	–	<15>	–	<>
Reproducibility	<20>	–	–	–	<>
Shelf time (days)	–	–	–	> 40 (for Au)	7 (for Ag)
Fabrication Steps	3	3	3	> 2	1–2
Cost	Expensive	Inexpensive	Inexpensive	Inexpensive	Moderate