Analytical and Bioanalytical Chemistry

, Volume 407, Issue 27, pp 8177–8195 | Cite as

Tip-enhanced Raman spectroscopy: tip-related issues

Review
Part of the following topical collections:
  1. Nanospectroscopy

Abstract

After over 15 years of development, tip-enhanced Raman spectroscopy (TERS) is now facing a very important stage in its history. TERS offers high detection sensitivity down to single molecules and a high spatial resolution down to sub-nanometers, which make it an unprecedented nanoscale analytical technique offering molecular fingerprint information. The tip is the core element in TERS, as it is the only source through which to support the enhancement effect and provide the high spatial resolution. However, TERS suffers and will continue to suffer from the limited availability of TERS tips with a high enhancement, good stability, and high reproducibility. This review focuses on the tip-related issues in TERS. We first discuss the parameters that influence the enhancement and spatial resolution of TERS and the possibility to optimize the performance of a TERS system via an in-depth understanding of the enhancement mechanism. We then analyze the methods that have been developed for producing TERS tips, including vacuum-based deposition, electrochemical etching, electrodeposition, electroless deposition, and microfabrication, with discussion on the advantages and weaknesses of some important methods. We also tackle the issue of lifetime and protection protocols of TERS tips which are very important for the stability of a tip. Last, some fundamental problems and challenges are proposed, which should be addressed before this promising nanoscale characterization tool can exert its full potential.

Graphical Abstract

Keywords

TERS Tip Electrochemical etching Coating Protection 

Introduction

Raman spectroscopy is a vibrational spectroscopic technique capable of providing molecular fingerprint information for chemical identification. As an analytical method, it has been widely used in various fields such as chemical analysis [1], material sciences [2], and biomedical applications [3]. However, the sensitivity of Raman spectroscopy is low because of the inherent low efficiency of the inelastic scattering [4] compared with other optical techniques. Fortunately, the discovery and development of surface-enhanced Raman spectroscopy (SERS) [5,6] have enormously improved the sensitivity of Raman spectroscopy. By exploiting the giant enhancement effect of Raman scattering from molecules in close proximity to plasmonic metal nanostructures, SERS has achieved sensitivity down to monolayer molecules or even single molecules [7, 8, 9, 10]. Nevertheless, the ideal spatial resolution of normal Raman spectroscopy and SERS achieved on a confocal system is in the range of half a wavelength of the excitation laser, determined by the optical diffraction limit. In 2000, tip-enhanced Raman spectroscopy (TERS) was first demonstrated [11, 12, 13, 14]; in the following 15 years, TERS has been developed into a promising and powerful tool for chemical identification at the nanometer scale. TERS combines scanning probe microscopy (SPM) with Raman spectroscopy and is able to provide not only a high detection sensitivity down to single-molecule level but also a high spatial resolution down to sub-nanometers, offering chemical fingerprint information and morphologic information about the surface simultaneously. TERS has found increasing applications in areas including catalysis [15, 16, 17], materials [18, 19, 20, 21, 22, 23, 24, 25, 26], biology [27, 28, 29], molecular electronics [30], and surface science [31, 32, 33, 34, 35, 36, 37, 38, 39].

In TERS, a sharp metallic or metallized tip (gold or silver) is brought in close proximity to the surface of a sample controlled by SPM. Three SPM techniques, i.e., scanning tunneling microscopy (STM), atomic force microscopy (AFM), and shear force microscopy (SFM), have been adopted in TERS setups. Then, a beam of laser light with an appropriate wavelength and polarization is focused onto the tip apex to excite the TERS signal. In TERS, the tip is the dominant enhancing source of Raman signals: it determines, to a large extent, not only the enhancement of the signal but also the spatial resolution of TERS and SPM imaging. In addition, the reliability of TERS results also relies on the stability of the TERS tip. As a decisive component of the TERS technique, the development of reliable and highly enhanced TERS tips is the basis for further development and application of TERS. However, how to reproducibly fabricate the TERS tip with a high enhancement is still one of the bottleneck problems in TERS. The low reproducibility is mainly due to the lack of control over the morphology and geometry of the tip, especially the apex, which in turn prevents the comprehensive understanding and thus optimization of TERS measurements. Therefore, the first key issue related to the TERS tip is to figure out the influence of various tip parameters, including materials, radius, cone angle, and surface roughness, on the enhancement and spatial resolution, which will be very helpful for the design and selection of suitable tips for TERS studies. The second issue is to develop methods to fabricate TERS tips (including that to be used with STM, AFM, and SFM), with optimal enhancement and high reproducibility on the basis of the above understanding, which will improve the performance and expand the applications of TERS. Last but not least is the issue of the lifetime of the TERS tip, i.e., how long can a tip retain the enhancement? It requires methods to prevent the tip from contamination and improve the stability of the tip.

In this review, we will mainly focus on these tip-related issues. We will first discuss the principle of enhancement mechanism of TERS, especially the role of the tip in the TERS enhancement. Then we will introduce various methods for fabrication of tips together with our opinions on their advantages and challenges. Finally we will deal with the methods to extend the lifetime of TERS tips. The review aims to provide a guideline to choose the right tips for those who are interested in or working on TERS.

Principles of TERS tip enhancement

Today’s consensus is that TERS involves two enhancement mechanisms, the so-called lightning-rod effect and localized surface plasmon resonance (LSPR). The lightning-rod effect is a purely shape-induced effect, whereby charge accumulates at sharp points in metal objects [40]; whereas, the LSPR is a collective oscillation of free electrons of metals driven by light of an appropriate wavelength, which induces the charge accumulation at the tip apex. As a result of the charge accumulation, both effects generate a strongly localized electromagnetic (EM) field at the tip apex and subsequently enhance the TERS signal. The overall average enhancement contributed by the lightning-rod effect and LSPR may reach 104 for AFM-based TERS and ca. 105–106 for STM-based TERS, by comparing TERS enhancement factors in more than 20 papers [41]. However, many variables governing the LSPR can dramatically change the field enhancement at the tip apex. As given in Fig. 1, incident laser (including the wavelength, polarization, and incident angle θ), tip (including the material, radius r, cone angle α, and morphology), substrate (including the material, roughness, and distance d between tip and substrate), and even surrounding medium (air or water) will significantly affect the overall TERS enhancement.
Fig. 1

Schematic illustration of parameters affecting the TERS enhancement in a TERS system

It is generally accepted that when the wavelength of the incident laser matches the LSPR peak of the TERS system, a strong EM field can be produced at the tip apex [42,43]. Since only the electric field component along the tip axis plays a dominant role in the TERS enhancement, a p-polarized incident beam can induce a stronger near-field enhancement than that with an s-polarized incident beam [42,44, 45, 46]. Recently, the radially polarized beam has been increasingly used in TERS [47,48], because it has a large component of the longitudinal field that can efficiently excite the LSPR at the tip apex to create a strong EM enhancement. It can increase the TERS signals by about eightfold compared to that excited by the linear polarized laser. In addition, the incidence angle-dependent TERS enhancement has been studied theoretically, which reveals the maximum field enhancement at an incidence angle in the range of ca. 40–60° [49,50].

In a practical TERS measurement, the tip apex has to approach the substrate sufficiently close, because the EM field significantly increases with the decrease of the distance between the tip and the substrate [49,51, 52, 53, 54, 55]. The typical working distance is ca. 1–2 nm. In addition, both experimental results and theoretical simulations have demonstrated that the properties of substrate can dramatically influence the field enhancement [49,56, 57, 58]. According to the substrate materials, the TERS working mode can be categorized into two types: uncoupled and coupled. The uncoupled-type TERS usually involves the use of substrates of dielectric materials (SiO2, Al2O3, etc.), which gives a weak enhancement [49,56]. It will not lead to an obvious shift of the LSPR peak of the tip compared with that of an isolated tip [59]. The coupled-type TERS, also called gap-mode TERS, employs a substrate of noble metals (Au, Ag, and Cu), which will induce a strong electromagnetic coupling between the tip and substrate and result in a huge field enhancement and an obvious red shift (by tens of nanometer) of the LSPR peak [59]. It has been found that the TERS enhancement of the coupled type is hundreds of times larger than that of the uncoupled type [49,56]. Therefore, most of the works with single-molecule sensitivity were achieved by using the coupled-type TERS. In comparison with noble metals, the use of transition metals (Pd, Pt, Ni, etc.) as substrates gives a weaker TERS enhancement due to the lower density of the surface conduction electrons [33,49,56]. Therefore it is challenging to obtain TERS signals from the transition metals that are important in surface science and heterogeneous catalysis. Anyhow, the enhancement from transition metals is still about 100 times higher than the uncoupled type, pointing to the promising future of this direction. Furthermore, it was found that the nanoscale roughness on the noble metal substrates can increase TERS enhancement by an order of magnitude [60].

It is well known that the dielectric permittivity of the surrounding medium can affect the LSPR position of metal nanostructures. We would expect a sizable red shift of the LSPR peaks when the tip is in water compared with that in air [59] without much change in the enhancement [61]. However, more delicate calculation is still necessary to obtain the exact peak position. For the sake of comparison, there is about a 30-nm red shift of the LSPR peak when a Ag nanoparticle is immersed in water compared with that in air [43].

In addition to the aforementioned three variables (laser, substrate, and surrounding medium), the TERS enhancement also depends strongly on the tip itself. In the following section, we focus on the influence of the tip parameters on the TERS enhancement.

Tip material

A metal tip or metal-coated tip is generally used to produce the enhanced field at its apex. The tip material determines the wavelength of the incident laser that can be used and the enhancement that can be achieved. Gold and silver are the two most commonly used materials to fabricate TERS tips owing to their high free electron density, strong LSPR effect in the visible region, and relatively high stability compared with alkali metals. It is well recognized that the high free electron density of the metal leads to a negative real part of the dielectric constants and a large polarizability, resulting in a strong field enhancement; whereas, the imaginary part presents the absorption process of light, leading to a loss of the energy. Therefore, the imaginary part should be as small as possible for an effective SPR process and thus a strong surface electric field. As shown in Fig. 2b, gold has a small imaginary part only in the red visible light region and thus can be excited by the red excitation light to produce a high TERS enhancement. Silver has low imaginary values throughout the whole visible light region and possesses a stronger electric field enhancement than that of gold. The optimal TERS enhancement with a silver tip can be obtained by the excitation of green light. The advantage of gold is the chemical inertness in air or an aqueous environment. When a silver tip works with noble metal substrates, a red light is commonly used. For some special applications in the UV region, aluminum is an ideal material owing to its small imaginary part in the UV region, as shown in Fig. 2b. UV-TERS has been successfully demonstrated with aluminum tips; this provides an additional opportunity for TERS to study catalytic materials, bioscience, and energy science [62,63].
Fig. 2

Optical properties of gold, silver, and aluminum in the UV–visible region. Real (a) and imaginary (b) parts of permittivity ε as a function of wavelength. (Plot using the data from [64])

The key of TERS is the excitation of LSPR at the tip apex by the incident laser. Therefore, it is crucial for TERS to match the LSPR wavelength of the tip with that of the excitation laser. There are ways to achieve the matching by changing either the laser wavelength or the LSPR position of the tip. It seems impractical to scan the wavelength of the laser because such a laser is not affordable in normal TERS labs and, more importantly, it is challenging to design an optical path to allow the convenient change of several wavelengths. Conversely, it is more feasible to tune the LSPR frequency of the tip to match the excitation laser to attain a high field enhancement. In 2007, the Zenobi group found by simulation that the LSPR wavelength of the silver-coated AFM tips can be red shifted by increasing the refractive index of the dielectric layer coating on the AFM tip (Fig. 3b) [65]. Moreover, the LSPR wavelength of the tip is modified, but the magnitude of the electric field enhancement is comparable. This result has been verified by experiments [66,67]. Unfortunately, the precoated dielectric layer will increase the tip radius, resulting in a lower spatial resolution. Alternatively, the Kawata group developed a method to grow a SiO2 layer on Si AFM tips by direct thermal oxidation before the deposition of the silver layer. In this way, the LSPR wavelength can be gradually blue shifted by the increasing thickness of the SiO2 layer without increasing the tip radius (Fig. 3d) [63].
Fig. 3

a Simulation model of a metal-coated dielectric tip. b Spectral responses of the field enhancement for Ag-coated Si3N4, SiO2, and AlF3 tips. The refractive index n of the three materials is nSi3N4 = 2.05, nSiO2 = 1.5, nAlF3 = 1.4. (Redrawn with permission from [65], copyright 2007 The Optical Society.) c FDTD model representing the Ag-coated oxidized Si probe that consists of Si, SiO2, and Ag layers. d Calculated near-field spectra of the Ag-coated oxidized Si probes with different SiO2 thicknesses. (Reprinted with permission from [63], copyright 2009 The Optical Society)

Tip radius

Another dominant factor influencing the field enhancement and spatial resolution of TERS is the tip radius. When a dielectric substrate (e.g., glass) is used, the tip plays the dominant role in the field enhancement because of the lack of coupling between the tip and substrate. The field enhancement will increase with the decreased tip radius (Fig. 4a) [68]. When the tip is coupled with a noble metal substrate (e.g. Au), the trend is different from the uncoupled case. As shown in Fig. 4b, when a gold tip is coupled with a gold substrate, the field enhancement increases slowly as the tip radius increases from 15 to 50 nm [49]. Further increase of the tip radius does not provide further enhancement. It is interesting to find that when the tip radius is decreased from 15 to 5 nm, the field enhancement rapidly increases, possibly owing to the lightning-rod effect as a result of the induced strong surface charge density at the tip apex. More data points are needed to reach a convincing conclusion. The spatial resolution is normally smaller than the tip radius, and it will increase with the decrease of the tip radius regardless of being in a coupled or uncoupled mode.
Fig. 4

Effect of the tip radius on the field enhancement and spatial resolution between tip and substrate. a Silver tips couple with a glass substrate. (Reprinted with permission from [68], copyright 2005 American Institute of Physics.) b Gold tips couple with gold substrate. (Reprinted with permission from [49], copyright 2009 John Wiley & Sons, Ltd.)

A similar trend was observed when a gold-coated Si tip approaches a gold substrate (Fig. 5a, b) [69]. From Fig. 5b, we find that the electric field enhancement increases sharply with the increased thickness of Au coating from 5 to 80 nm as a result of the increased size of the tip. Meanwhile, the LSPR peak shifts from 528 to 613 nm. Further increase of the thickness will lead to a decrease in the peak intensity. The maximum enhancement for the thickness from 50 and 100 nm is in the same order of magnitude. The theoretical prediction is in good agreement with the experimental result, which demonstrates the maximum TERS enhancement at a thickness of ca. 55–75 nm (Fig. 5d) (Yang et al. 2015, in review). Furthermore, the spatial resolution of TERS increases with the decrease of the thickness and is much smaller than the tip radius, as shown in Figs. 4b and 5c. This tendency is similar to that of metal tips coupled with a dielectric substrate (Fig. 4a). However, the spatial resolution is higher on the gold substrate than that on the dielectric substrate for a given radius.
Fig. 5

a Calculation model for a gold-coated silicon AFM tip over a Au substrate. b Dependence of the electric field enhancement on the thickness (h) of the gold coating. c Dependence of the spatial resolution (SR) (using 632.8 nm as the excitation wavelength) on the thickness. (Reprinted with permission from [69] copyright 2015 The Optical Society.) d Experimental results of the dependence of the TERS enhancement on the thickness. The peak of MGITC at 1176 cm−1 was used

The conclusion of this part is that a higher spatial resolution can be achieved at a smaller tip radius both for the uncoupled and coupled-type TERS. However, the enhancement follows different trends depending on the substrate properties. For the uncoupled type, the trend is that the smaller the tip, the higher the enhancement. However, for the coupled type, the maximum TERS enhancement can be achieved when the tip radius is in the range of 50 to 80 nm.

In fact, there are still some open questions concerning the influence of the tip radius. First, it is still necessary to perform comprehensive theoretical calculations so that the differences in tip materials (also the dielectric layer), substrate materials, and the coating properties can all be included [68,70,71]. Second, the lightning-rod effect needs to be carefully considered, especially when the tip radius is below 10 nm [49]. Lastly, the effect of the tip radius for all different configurations should be carefully verified by elegantly designed experiments.

Tip angle

There are far fewer studies on the effect of the tip angle compared with those of the tip radius. Zhang et al. [71] studied the cone angle in a range of ca. 15–25° for the side illumination and found a minor effect of the tip angle on the field enhancement. Thereafter, Angulo et al. systematically studied the dependence of the cone angle on the EM field enhancement for wedge-shaped metal nanostructures, which have a similar geometry to the TERS tip [72]. The conclusion obtained in this study may be utilized by TERS. Figure 6 shows the EM field enhancement as a function of the wavelength and cone angle of the wedge in different media. The cone angle with the EM field intensity maxima exponentially increases with the decrease of the incident wavelength. For a given wavelength, there is an optimal cone angle to obtain the maximum EM field enhancement. The optimal angle increases with the increase of the refraction index of the surrounding medium. For a sharp tip with a cone angle smaller than 30°, there is a broad continuous LSPR band, which becomes broader with decreasing angle and increasing refractive index of the surrounding medium. On using these conclusions for TERS, we have to realize the following points: First, the wedge-type structure is still a little bit different from a real TERS tip, so we would expect some minor difference. Second, the simulation was done without substrate. We would expect a change of the optimal cone angle when the tip is coupled with a noble metal substrate. Third, in a real TERS study, the incident angle of the excitation laser may affect the optimal cone angle of tip at a given wavelength. Lastly, a very sharp tip with a small cone angle may result in severe mechanical drift and poor SPM imaging.
Fig. 6

EM field intensities as a function of cone angel α and wavelength at 1 nm from the tip of the wedge. White circles show the strongest field. (Reprinted with permission from [72], copyright 2015 American Chemical Society)

Morphology of the tip apex

The LSPR effect is the main source of enhancement for SERS/TERS and it strongly depends on the morphology of the metal nanostructures. It is widely accepted that the morphology of metal nanoparticles including the size, shape, and roughness significantly influences the surface enhancement effect in SERS [73, 74, 75, 76]. In TERS, as the tip is the only enhancement source, we would expect the morphology of the tip to play a more crucial role. At present, there are four types of tips generally used in practical TERS experiments: the tip with attached single metal nanoparticle (NP tip) [77], the tip covered with discrete metal nanoparticles (NPs tip) [78], the tip covered with a continuous layer of metal nanoparticles (NPs-layer tip) [70], and the full metal tip [79], as shown in the inset of Fig. 7. The first three are usually used in AFM-based TERS, and the last one is commonly used in STM-based TERS.
Fig. 7

a A tip attached with a single metal nanoparticle (NP tip). (Redrawn with permission from [77], copyright 2008 American Chemical Society.) b A tip covered with discrete metal nanoparticles (NPs tip). (Redrawn with permission from [78], copyright 2015 The Royal Society of Chemistry.) c A tip coated with a continuous layer of metal nanoparticles (NPs-layer tip). d A full metal tip (Redrawn with permission from [79], copyright 2007 American Institute of Physics)

The NP tip can be treated as a single metal nanoparticle. Therefore, it will have the same optical response as a metal nanoparticle. However, it is time consuming to attach an individual nanoparticle to the tip. In comparison, NPs tips and NPs-layer tips are more widely used. For the NPs tip, the distance between the neighboring nanoparticles is too large to significantly affect the optical response of the nanoparticle at the apex. Therefore, the NPs tip can be also treated as a single nanoparticle. However, different from the NP tip, the nanoparticles other than the one at the apex of the NPs tip will also be excited by the laser; this gives a continuum background to the TERS spectra and reduces the signal to noise ratio. It is still very challenging to control the surface morphology, including the size and shape of evaporated nanoparticles, during the fabrication of the NPs tip and NPs-layer tip. It was recognized recently that the dipole direction of the tip significantly influences the optical response and the imaging quality of TERS. Therefore, a defocused imaging method has been developed to evaluate the direction of the tip dipole of NPs-layer tips and to correlate the morphology and optical response of the NPs-layer tip [80].

The field enhancement of the full metal tip arises from the lightning-rod effect and SPR effect. When the gold/silver tip is excited by a light polarized along the tip axis, the surface charge will oscillate at the same frequency as the exciting field. The surface charges form an oscillating standing wave and support the surface plasmon. Therefore, the induced surface charges have a large chance to accumulate at the apex of the tip and give a strongly enhanced electric field [81,82]. The NPs-layer tip shares a similar optical process with the full metal tip, but the scattering from the rough surface of the NPs-layer tip will reduce the efficiency of surface plasmon and leads to a dramatic decrease of the field enhancement. Benefitting from the lighting-rod effect and the effective accumulation of surface charges, the full metal tip shows much higher enhancement than that of the NP tip, NPs tip, and NPs-layer tip [83]. This may partially explain why STM-based TERS (using full metal tip) provides stronger signals and better spatial resolution than those of AFM-based TERS (especially using NPs-layer tip) [84].

Methods for fabricating TERS tips

STM, AFM, and SFM are commonly employed to build TERS systems. Accordingly, there are three types of tips: STM TERS tips, AFM TERS tips, and SFM TERS tips. In the STM TERS, a tip with a high conductivity is required for the flow of the tunneling current. In AFM TERS, the feedback relies on either the changed reflection of a laser beam or a change in the resonance frequency of the tip, and it requires a cantilever-based tip with different resonance frequencies in the range of ca. 10–400 kHz. In SFM TERS, almost all kinds of tips can be attached to a quartz crystal tuning fork to work in the SFM feedback. Up to now, considerable efforts have been devoted to the fabrication of TERS tips with an improved enhancement and reproducibility. In this section, we provide an overview of the existing methods for fabrication of TERS tips.

STM TERS tips

Sharp tips are essential for the high resolution STM imaging. Although cutting with scissors may produce very nice STM tips, it does not produce reliable tips for TERS because of the irregular tip apex. A more common and reproducible way to produce STM TERS tips is electrochemical (EC) etching, which is to anodically dissolve a metal wire in an etching electrolyte. Sharp tips are produced through the “drop off” mechanism [85, 86]. As shown in Fig. 8, during the etching processes, the part near the air–solution interface will be etched and the thinnest neck will break down until it is too thin to support the lower metal wire. Although two sharp tips are produced near the air–electrolyte interface, the lower part is usually discarded because it is too thin to be used as an STM tip. Therefore, only the upper part is used in most cases. Since the upper part may still be immersed in the solution and become blunt as a result of overetching, it is critical to cut off the electronic circuit immediately when the lower part drops off.
Fig. 8

Schematic of the EC etching setup: the tip is produced by etching the center metal wire with an anodic voltage and a large inert metal ring is used as the counter electrode, usually placed at the air–solution interface

The quality of the tip fabricated by the EC etching depends on the composition of the etchant, the etching voltage, and the method used to cut off the circuit, etc. However, the detailed etching mechanism and determining parameters are still too complicated to discuss, because it was found that the etching temperature and the concentration of the original chemicals also play key roles in the reproducibility.

Fabrication of gold STM TERS tips

Up to now the most established method for preparing TERS tips is to fabricate gold tips by the EC etching method. Several groups have made significant efforts towards this goal. Our group [87] has developed an EC etching method for fabricating gold tips for TERS. We developed a recipe consisting of fuming HCl and anhydrous ethanol (1:1, v/v), by which we can now routinely fabricate gold tips with a radius below 30 nm and smooth surface as shown in Fig. 9b by applying a DC voltage of 2.2–2.4 V. The as-etched tip shows a good TERS enhancement (Fig. 9c). During the etching process, electrochemical current oscillation occurs (Fig. 9a) as a result of the following reactions [87,88]:
Fig. 9

a Current–time curves of Au wires etched in a mixture of HCl and ethanol (1:1, v/v). b SEM image of the Au tips etched at 2.2 V. c TERS spectra for the 4,4-bipyridine SAM layer on the Au (111) surface (top spectrum). The bottom spectrum is for the tip-retracted case for comparison. Acquisition time 100 s, laser power 1 mW. (Redrawn with permission from [79], copyright 2007 American Institute of Physics)

$$ \begin{array}{l}\mathrm{A}\mathrm{u}+4{\mathrm{Cl}}^{-}\to {\mathrm{AuCl}}_4^{-}+3{e}^{-}\\ {}\mathrm{A}\mathrm{u}+2{\mathrm{Cl}}^{-}\to {\mathrm{AuCl}}_2^{-}+{e}^{-}\\ {}{\mathrm{AuCl}}_2^{-}+2{\mathrm{Cl}}^{-}\to {\mathrm{AuCl}}_4^{-}+2{e}^{-}\end{array} $$

The depletion of Cl ions during the dissolution of gold leads to the formation of gold oxides which may act as a passivation layer to inhibit the electrochemical reaction and thus result in a decrease of the reaction current. The Cl ion is then replenished by diffusion from solution to dissolve the surface oxide and the fresh gold surface is exposed again resulting in an increase of the reaction current. The process is repeated during the etching process, leading to the oscillation of current [87,89,90]. On the gold ring cathode, some reduction reactions take place, such as the hydrogen evolution and Au deposition, resulting in rigorous bubbling. It was found that the generated bubbles are driven to the gold tip on etching. If the bubbles are stuck on the Au wire surface, the surface will be rough. A cell with a jacket structure [91,92] has been developed to avoid the bubbling effect and stabilize the etching process. To overcome the volatile and corrosive nature of fuming HCl, some groups have used less toxic chloride salts like CaCl2 [93,94] and NaCl [95] instead of HCl as etchants. In fact, if Cl ion is used, it is unavoidable to produce Cl2, which may be an irritant to eyes and lungs. Since the mixture of fuming HCl and anhydrous ethanol is still the best recipe for fabrication of gold tips concerning the high reproducibility and sharp apex, it is currently the most widely used recipe in TERS laboratories and companies.

Etching voltage is an important parameter to produce sharp gold tips with a smooth surface and an appropriate aspect ratio. For different etchants the optimal voltages are different. Both DC [87, 88, 89,91,96] and AC (including square DC-pulsed voltage) [92,94, 95, 96, 97] voltage have been used for etching. Generally, DC etching is simple and the etching process generally finishes in minutes; whereas, the AC method can further accelerate the mass transfer process and thus the supplement of the etchants, yielding a faster etching rate [94]. For the AC method etching parameters such as high limit voltage, low limit voltage, period time, and duty cycle can also be controlled to fabricate gold tips with designed geometries [98].

In addition to the etchant and voltage, one should also pay attention to the crystallinity of the gold wire to be etched, which may also affect the shape and roughness of a tip [96]. Usually a soft gold wire (Premion type) has less defects and grain boundaries than the hard wires, which may have a better reproducibility and yield of sharp and smooth gold tips. Annealing the wire prior to etching may be a good practice to reduce the defects and grain boundaries [99].

Fabrication of silver STM TERS tips

In general, silver tips have a higher enhancement than that of the gold tips. It is not surprising that a large amount of effort has been made to fabricate silver TERS tips. The studied etchants contain citric acid [100], ammonia [101], perchloric acid [102], nitric acid [103], and concentrated sulfuric acid [104]. Although different kinds of etchants have been developed, a short cutoff time to avoid serious overetching is found to be more critical for making sharp silver tips. Up to now, three main cutoff methods have been used: an electric control circuit, a relay, and an optical method. The electric control circuit may have a response time of less than microseconds by detecting a sudden potential change [105]. It is the most widely used method to fabricate silver tips [101,103,104,106,107]. Sharp tips with radii less than 50 nm can be fabricated routinely [106], as shown in Fig. 10a. However, this method has inherent problems. For example, it is difficult to set a proper reference voltage for cutoff. A high reference voltage leads to the premature cutoff, whereas a low one results in serious overetching. In practical applications, the real period of overetching is longer than the response time of the circuit, so the fabricated tips are not as sharp as expected. The relay-based method is simple. The potentiostat is able to cut off the circuit automatically through the built-in relay when the measured current is smaller than a preset threshold [79]. The cutoff time is usually on the time scale of several milliseconds. The radius of the silver tip prepared by this method is about 100 nm [108]. Figure 10b shows the typical result. The third method is an optical method-based machine vision (Li et al. 2015, in review), which is to detect the drop off of the lower part of tips by tracking the vertical position change instead of the current or potential change. Therefore the crosstalk between cutoff method and EC etching can be eliminated. The demonstrated response time of this method has reached tens of milliseconds and it can be further reduced by using short pulses for EC etching. In this way, sharp silver tips with a mean radius of curvature of about 58 nm are fabricated (Fig. 10c). The common problem of these three methods is that the shape of the fabricated tip is uncontrollable. In contrast, sharp silver tips with smooth surface can be fabricated (Fig. 10d) by a two-step approach involving a process of EC polishing after etching [109]. However the setup is somewhat complicated and the method is time consuming.
Fig. 10

SEM images of silver tips fabricated by the following methods: a Method using an electric control circuit. (Reprinted with permission from [106], copyright 2010 American Chemical Society.) b Method using a potentiostat. (Reprinted with permission from [108], copyright 2011 American Chemical Society.) c Optical cutoff method. d Two-step approach. (Reprinted with permission from [109], copyright 2013 AIP Publishing LLC). Etchants containing perchloric acid and ammonia are used by the first three methods and the last method, respectively

Although various methods have been developed to fabricate silver tips, the EC etched silver tips are still relatively blunt and rough and the reproducibility is still low compared with gold tips. Therefore, there is still plenty of room to improve the reproducibility and sharpness, and more statistical data to show the reproducibility are still needed [103,109].

Other fabrication methods and specialized STM TERS tips

The EC etching method is a common way to fabricate STM TERS tips. However, as a result of the reproducibility issue of the EC etched Ag tips and the requirement for tips with specialized structures for special purposes, other methods for fabricating STM TERS tips have been reported. In order to exploit the very sharp apex of tungsten tips, Yi et al. [110] coated a Ag film on tungsten tips through sputtering; however, the tip apex becomes blunt after Ag sputtering. On the other hand, You et al. [111] and Fujita et al. [112] attached Ag nanowires onto the tungsten tips with an AC dielectrophoresis method. It is still time consuming to finish the attachment processes. Focused ion beam (FIB) milling has been used to further sharpen the tip [51]. More interestingly, the Raschke group [113] used FIB milling to fabricate gratings on the tip shaft (Fig. 11). When laser is focused onto the shaft, surface plasmon polaritons (SPP) can be excited owing to the coupling effect of the grating, which can subsequently propagate to the tip apex for the nanoscale excitation. Compared with the normal tip by direct illumination at the apex, this grating coupling scheme could eliminate the far-field background. Although the FIB method is good for obtaining special types of tips, it is not affordable in ordinary laboratories.
Fig. 11

a Schematic of principle for the grating coupled tip. Incident light is focused onto the grating, and the Raman scattered light excited by the nanofocused SPP at the apex is detected at a 90° angle to the incident plane. b SEM image of a FIB milled EC etched Au tip superimposed with optical image of grating illumination and apex emission. (Reprinted with permission from [113], copyright 2010 American Chemical Society)

AFM TERS tips

Different from the STM case, most AFM studies use the commercial tips for measurement. Therefore, it is also common practice to fabricate AFM TERS tips by deposition of metal onto a commercially available silicon cantilever. In addition, different kinds of specialized tips have been developed and used in AFM TERS studies.

Vacuum evaporation

The most prevalent method of preparing the AFM TERS tip is to use vacuum evaporation to deposit metal onto the surface of a commercial silicon tip. The vacuum evaporation method allows one to controllably transfer atoms from a heated source to a substrate located at a distance away from the source to grow a film in the vacuum environment. Many factors may affect the quality of the as-deposited film [114,115]. The morphology of the metal film could be controlled by the evaporation rate [114,116]. The metal film is rough with distinct particles at a slow evaporation rate and becomes a smooth continuous structure with the increase of the evaporation rate. The shapes and sizes of the metal film depend on the migration rate of the metal atoms on the substrate surface. The metal atoms migrate over the substrate surface until they are adsorbed at the active sites to form a strong chemical bond with the substrate, or these migrated metal atoms may collide with each other to form an agglomerate. The active sites and agglomerates act as the nuclei for the growth of the metal film. The arrival of new atoms will disrupt the migration of the atoms on the surface to increase the probability of nucleation. Consequently, high nucleation density is obtained at a higher deposition rate and leads to relatively smooth metal film. Temperature is another important factor that interplays with the deposition rate [117,118]. In general, the temperature of the substrate is slightly higher than the room temperature, considering the long separation distance between the source and the substrate (e.g., 12 cm). When the distance becomes shorter, e.g., 5 cm, a large variation in temperature across the substrate may affect the uniformity of the surface morphology of the metal film. Furthermore, vapor pressure also affects the rate and moving direction of the vapor atom and thus changes the crystalline texture of the film. In practice, the dependence of the vapor pressure at a given temperature for varied materials is different. In addition, other factors, such as the placement of the heater, source, and substrate may also affect the structure of the deposited films.

Up to now, it is still an art to control all the technical parameters of the vacuum evaporation method. Consequently, tips with quite different morphologies have been reported from different groups (shown in Fig. 12) [63,66,78,119]. Furthermore, it has been reported that the mechanical stability of the tip fabricated by this method is poor. It was found that the wear of the metal film occurred during AFM scanning and the peeling off of the coating layer happened after the tip was immersed in water for 1 h [120, 121, 122]. Hence, an extra auxiliary layer or protection layer is needed to protect the coating layer from peeling off. Most importantly, the enhancement factors of the tips fabricated by the vacuum evaporation are known to vary a lot [123]. This variation may be due to the random deposition of the metal nanoparticles onto the tip apex during the physical deposition process and therefore the probability of obtaining a “hot spot” on the tip apex is low [123].
Fig. 12

Vacuum evaporation of silver-coated AFM TERS tips by different groups: a Zenobi’s group. (Reprinted with permission from [66], copyright 2006 Society for Applied Spectroscopy.) b Kawata’s group. (Redrawn with permission from [63], copyright 2009 The Optical Society.) c Deckert’s group. (Reprinted with permission from [78], copyright 2015 The Royal Society of Chemistry.) d McNaughton’s group. (Reprinted with permission from [119], copyright 2011 John Wiley & Sons, Ltd.)

Chemical deposition

Compared with the vacuum deposition processes, chemical deposition has emerged as a more simple, environmentally friendly, and cost-effective micro/nanofabrication method. The quality of the coating layer is mainly controlled by the reductant (or deposition potential or current) and deposition time. In addition, as a benchtop method, chemical deposition is convenient for the mass production of TERS tips in the lab. Two kinds of chemical deposition methods have been used to fabricate the metal-coated AFM tip: electrodeposition method and electroless deposition method.

Electrodeposition is widely employed in applications ranging from surface finishing, corrosion inhibition, to nanoscale feature fabrication for the ultra-large-scale integration (ULSI) [124,125]. The deposition thickness can range from a few angstroms to several millimeters by controlling the deposition current, potential, or time. The surface roughness depends on the nucleation density on the substrate. Similar to the vacumm evaporation methods, the surface roughness is high at low overpotentials with a low deposition rate and becomes low at high overpotentials with a high deposition rate. In addition, the stronger the complexing ability of the ligand used in the electrodeposition bath is, the more negative the overpotential that is required, which is beneficial for the formation of a smooth metal layer. Several methods have been developed to obtain a uniform coating, such as direct-current, direct-potential, and pulsed eletrodeposition. Recently, we developed a new method to conveniently fabricate Au-coated AFM TERS tips with a high reproducibility by pulsed electrodeposition [70].

Figure 13a, b shows the SEM images of two Au-coated AFM tips prepared by pulsed electrodeposition with different deposition times. With the increase of the deposition time, the radius of the tip increases. Thus, Au-coated AFM tips of different sizes can be conveniently prepared by controlling the deposition times. The tip produced by the electrodeposition method is highly reproducible, as reflected by the dependence of TERS enhancement factor on the tip radius shown previously in Fig. 5d. Most importantly, the tip can be prepared in minutes in a normal wet chemistry lab.
Fig. 13

SEM images of metal-coated tips prepared by the chemical method. Au-coated AFM tip produced by pulsed electrodeposition with a deposition time of 35 ms (a) and 50 ms (b). c Ag-coated tip fabricated by mirror reaction. (Redrawn with permission from [126], copyright 2005 The Chemical Society of Japan.) d Ag-coated tip fabricated by galvanic reaction. (Redrawn with permission from [127], copyright 2010 Society for Applied Spectroscopy)

Unlike the electrodeposition method, electroless deposition allows metals to be deposited on the surface without controlling the tip potential. The latter is an important method to coat a metal film on a nonconductive surface, such as SiO2. In particular, the uniform coating on substrates of complicated shapes can be obtained with the electroless deposition [128]. Some efforts have been devoted to the fabrication of metal-coated tips by electroless deposition [126,127,129,130]. The mirror reaction is one of the most commonly used electroless deposition techniques to produce silver on Si AFM tips [126]. Figure 12c shows a tip that was decorated with isolated Ag nano islands owing to the sparse nucleation. The low density of nucleation sites may be a result of the smooth Si surface with few defects for the formation of the nuclei and the low concentration of the reactants. To achieve a continuous film, the nucleation density may be improved by increasing the concentration of the reaction species and self-assembly of a layer of Au or Ag nanoparticles on the AFM tip surface to act as the nucleation sites. Galvanic reaction of Ag(I) ions with elemental silicon was developed by Brejna and Griffiths [127] to electrolessly deposit silver film on silicon. As a result of the nature of the galvanic reaction, part of the silicon acted as the cathode so that the silver can be deposited on it; the other part of the silicon was dissolved. As a result, the tip morphology is rough. Surprisingly, the particles do not form nuclei on the apex (Fig. 13d), and the silver coating is an extension of the coating from the shaft resulting in a large tip radius.

Fabrication of specialized AFM TERS tips

In addition to the metal-coated tips, considerable efforts have been devoted to produce non-standard AFM TERS tips. In general, non-standard tips have special metal nanostructures at the tip apex to confine the laser energy and produce a strong field enhancement. The tip can be obtained by the attachment of nanoparticles to an AFM tip, or by the microfabrication method, or bending of the electrochemical etched metal tip in the lab.

Attachment of nanoparticles

Attaching nanoparticles to the AFM tip apex is one of the most commonly used method to fabricated specialized tips [77,131, 132]. The in situ pick-up technique is the most primitive method to attach a nanoparticle to the tip apex. However, the efficiency is low because of the random attachment of the nanoparticle. In 2006, Vakarelski and Higashitani developed a full wet surface-assembly technique to attach a gold nanoparticle onto the AFM tip apex [131]. An AFM tip was silanized with a phenethyltrichlorosilane (PETS) passivation layer and then the tip was scanned against a silica wafer to remove the passivation layer at the end of the tip. The exposed end was then functionalized with 3-methacryloxypropyltrimethoxysilane (3-MPTS). Finally, the tip was dipped into the Au nanoparticle suspension to allow a single nanoparticle to covalently bond to the tip apex. As clearly shown in Fig. 14, the tip was terminated with a single gold nanoparticle. Compared with the in situ pick-up technique, this method has a higher success rate for attaching a nanoparticle to the end of the tip, which is as high as 30–50 %. In additional, the tip fabricated with this method has the advantage that only one nanoparticle is located at the apex, which is different from the one with multiple nanoparticles distributed on the tip fabricated by the inverse self-assembly technique [132].
Fig. 14

Schematic of the procedure of the wet surface-assembly technique to attach a nanoparticle to an AFM tip apex. Inset shows the SEM image of the tip terminated with a single gold nanoparticle with a size of 25 nm. (Redrawn with permission from [133], copyright 2006 American Chemical Society)

Selective deposition of nanoparticles

To overcome the low efficiency of the nanoparticles attaching method, Wang et al. developed a method to selectively deposit nanoparticles onto the AFM tip apex by sequential electrochemical oxidation followed by site-selective growth of the nanoparticle onto the AFM tip apex [77]. The preparation procedure is shown in Fig. 15a. The method can be used to prepare AFM TERS tips modified with silver, copper, and gold nanoparticles. Another different approach was developed by the Verma group [133] by photoreduction of the Ag+ ion to Ag atom under light irradiation. As a result of the lightning-rod effect and high surface energy of the sharp corner, silver was selectively deposited at the tip apex and edges of the cantilever (Fig. 15b). Interestingly, the authors found that such a tip provided an enhancement an order of magnitude higher in comparison with the fully metallized standard tips. As a result of the irregular growth of silver on the tip apex, a “multiple-tips” effect may result in poor spatial resolution. Okamoto and Yamaguchi followed a similar approach and developed a photocatalytic deposition method to selectively grow a gold nanoparticle onto the tip apex [134]. Instead of using direct illumination, they illuminated the tip with an evanescent wave generated by the attenuated total reflection of a He–Cd beam at the prism–electrolyte interface.
Fig. 15

a Schematic of the preparation a single nanoparticle-terminated AFM tip by the sequential electrochemical oxidation and site-selective growth. Inset is the SEM image of an AFM tip terminated with one silver nanoparticle with a diameter of about 130 nm. (Redrawn with permission from [77], copyright 2008 American Chemical Society.) b Schematic of the photoreduction deposition of silver onto the AFM tip apex and the SEM image of an AFM tip after photoreduction. (Reprinted with permission from [135], copyright 2012 The Japan Society of Applied Physics)

Microfabrication method

Although photoinduced deposition has the advantage of easy operation, it is still challenging to control the shape of the deposited particle. Microfabrication appears to be a very promising way to fabricate TERS tips with a high enhancement and reproducibility because it provides a precise control of the size and shape of the nanomaterials as well as a much higher reproducibility. To date, microfabrication methods have been used to mill [135, 136, 137], sculpt [138,139], and grow [140,141] specialized AFM TERS tips with various nanostructures on the tip apex. Fleischer et al. [140] combined the thin-film metallization, electron beam-induced deposition of etching masks, and Ar+ ion milling to fabricate single individually engineered gold cones at an AFM tip apex (Fig. 16a). The gold cone acts as an optical antenna with a tip radius on the order of 10 nm and an adjustable plasmon resonance frequency. The sharp gold cone offers a strong near-field enhancement and a high spatial resolution. Small nanostructures have also been made at the AFM tip apex to confine and enhance the EM field. In 2005, Farahani et al. [135] designed a bowtie antenna at the AFM tip apex by FIB milling (Fig. 16b). The coupling of the two arms of the antenna produced strong field enhancement. However, this specialized tip has poor AFM imaging properties due to the “double-tips” effect of the two arms. A similar idea but with different geometry was developed by the Schuck group [136]. They fabricated AFM tips with coaxial apertures as shown in Fig. 16c. This kind of tip has a strong TERS enhancement as well as a good AFM imaging capability owing to the sharp apex. They further developed a campanile tip, which had an ultra-large field enhancement extended over an enormous bandwidth [142]. This kind of structure can be integrated into the apex of AFM and near-field scanning optical microscopy (NSOM). Another specialized tip with a different geometry was presented by the Di Fabrizio group [138,143] where a photonic crystal cavity was sculpted on the cantilever of an AFM tip (Fig. 16d). The photonic crystal cavity had an efficient coupling between the excitation laser and the tapered waveguide. The SPPs were excited by the incident laser coming from the photonic crystal cavity and confined at the apex of the tapered waveguide, leading to the drastic decrease of the background intensity at the sample plane. In addition, the sharp tip apex provided good AFM imaging properties.
Fig. 16

Specialized AFM TERS tips fabricated by the FIB technique. a Overview over the gold cone scanning probe. (Redrawn with permission from [140], copyright 2011 American Chemical Society.) b Bowtie antenna on an AFM tip apex. (Reprinted with permission from [135], copyright 2005 The American Physical Society.) c Coaxial optical antenna on Au-coated SiN tip. (Reprinted with permission from [136], copyright 2011 American Chemical Society.) d Photonic crystal cavity on a Si3N4 cantilever with attached tapered waveguide. (Redrawn with permission from [138], copyright 2010 Macmillan Publishers Limited)

Owing to the precise control of the size and shape of the nanomaterials, micromanufacturing technology is a promising tool for fabricating special tips to study the tip effect on the near-field enhancement. By using the FIB milling, metal-coated AFM tips with different lengths have been fabricated [137,144]. It was found that the resonant wavelength red shifts with increasing tip length and a high TERS enhancement could be obtained when the laser wavelengths were near the plasmonic resonance of the tip. Benefitting from the precise control of FIB milling, the plasmon resonance of the tip can be tuned to obtain better TERS enhancement and imaging resolution. More recently, a “tip-on-tip” TERS tip has been fabricated by combining a simple Ar+ ion sputtering route to obtain silver nanoneedle arrays on the tip apex [145]. It should be noted that this specialized TERS tip exhibits a higher enhancement than that of a Ag-coated AFM tip because of the intense electromagnetic field around the apexes of the Ag nanoneedles.

Full metal AFM TERS tip

Microfabrication, such as FIB, is not accessible for normal TERS labs. Another specialized TERS tip is made in the lab from the full EC etched metal AFM tip [93,146]. The probe consists of an AFM chip and etched metal wire (Fig. 17) and it has both electrochemical and force-sensing capabilities. A metal tip was first fabricated by EC etching of the metal microwire, and the cantilever component of tip was then achieved by pressing the unetched part of the microwire between two stainless steel disks using pliers. Finally, the tip was glued onto an AFM chip. This tip has all the advantages of the EC etched Au or Ag tip, including excellent TERS activity. However, it is tedious and time consuming to make.
Fig. 17

a Schematic of AFM tip made from Pt wire. (Reprinted with permission from [146], copyright 2000 American Chemical Society.) b SEM image of the full Au AFM tip

SFM TERS tips

There are several advantages of the tuning fork-based SFM [147]. First, it is not necessary to use a feedback laser to control the tip–sample distance. Therefore, the structure of the SFM head can be significantly simplified compared with AFM. Second, SFM can be applied to all different types of samples as in AFM. Third, almost all kinds of tips can be attached to the quartz tuning forks for later use.

The most widely used tip for SFM-based TERS was made by gluing a silver or gold tip prepared by EC etching to the quartz tuning fork [51,148, 149, 150]. The mass of the glued tip and the amount of glue have a great effect on the quality factor of the quartz tuning fork. They should be as small as possible to maintain a high quality factor to achieve sufficiently high force sensitivity. SFM TERS tips can also be made by coating a metal layer or attaching nanoparticles [151, 152, 153] to the sharp end fiber apex, which appears to have a low reproducibility and efficiency. In 2012, Johnson et al. [154] proposed a promising method to fabricate over one million gold pyramids in one wafer by using a template-stripping fabrication method. These pyramids are highly reproducible and can be stored for a long time. However, the pyramid should be glued to a thin wire and then glued to a tuning fork for TERS studies. This method is very promising for commercialization if the procedure can be further simplified.

Cleaning, protection, and lifetime

A clean tip with a high stability guarantees the reliability and reproducibility of the TERS measurement. Generally, some organic contaminants adsorbed on the tip surface may interfere with the TERS measurement when the tip is exposed to air. Similarly, when TERS is performed in liquids, analyte molecules and contaminants in the aqueous solution may be adsorbed on the tip and thus interfere with TERS signals [122]. Thus, the contamination may prevent the re-use of a TERS tip or even shorten the lifetime of the tip. A simple way to inhibit the adsorption of contaminants including carbon and analyte molecules is to coat a SAM (self-assembly monolayer) of ethanethiolate (EtS) on the tip [122]. The SAM may protect the tip from contaminants in both the aqueous solution and air. Furthermore, it is possible to remove the contaminants when the gold tip is contaminated. Our group [35] proposed a method to clean gold tips by dipping them in concentrated sulfuric acid. The tip is then soaked in ultrapure water to remove sulfuric acid and in ethanol for quick drying. With this method, the stored gold tip can be regenerated with a high TERS activity.

On the other hand, the stability of TERS tips, including chemical stability and mechanical stability, is a critical issue for their use. As is known, silver shows low chemical stability compared with that of gold. As a result of the oxidation and tarnish of silver, the lifetimes of silver tips reported in the literature are different, ranging from several hours [155] to several days [121]. Recently, we found that metal silver tips kept a high activity without obvious degradation of enhancement after storing in air (T approx. 22 °C, RH approx. 45 %) for more than 2 months (Huang et al. 2015, in review). The main reason is the formation of a compact oxide layer (only about 1.7 nm thick) when silver is exposed to air, which prevents the further oxidation of silver. More interestingly, such an ultrathin layer of silver oxide protects the silver tip from contamination.

In addition to the chemical stability, the mechanical stability must also be taken into consideration. For the AFM TERS tip, it has been found that the coating may be easily worn after scanning in AFM experiments [121,156], especially in the contact mode. Two methods have been proposed to improve the mechanical stability of the coating. One is to add a buffer layer, such as SiOx, to improve the adhesion of the coating and reduce the peeling off of the coating layer when performing TERS in liquids [122]. The other is to deposit an additional layer to protect the metal coating. Foster et al. [121,156] added an ultrathin coating of Al2O3 or SiOx on silver-coated AFM tips (Figs. 18 and 19). The coated hard oxide layer improves the wear resistance, prevents chemical attack, and inhibits degradation. Therefore, the lifetime can be extended. However, the protective layer increases the gap between the tip and sample, and thus the TERS activity will be decreased. Therefore, it is vitally important to find a delicate balance between an effective protection and a high enhancement.
Fig. 18

a, b SEM images of two unprotected Ag-coated tips; c, d SEM images of two protected Ag-coated tips after six scans on a standard silicon grid. In all cases, the apex is circled. All the scale bars are 1 μm. (Redrawn with permission from [156], copyright 2013 John Wiley & Sons, Ltd.)

Fig. 19

SEM images of AlF3/Ag-coated tip a before and b after soaking in water for 1 h, and SiOx/Ag-coated tip c before and d after soaking in water for 1 h. The scale bars represent 100 nm. The dotted circle highlights the detachment of a few individual Ag particles under the action of water from the side wall of the tip. (Redrawn with permission from [122], copyright 2009 John Wiley & Sons, Ltd.

Conclusion and perspectives

The tip plays dual roles in TERS. It is the most important part in TERS to provide not only the highly localized enhanced electric field to boost Raman signal of the sample so that the molecular information can be achieved with a nanometer spatial resolution but also simultaneously the morphological and other information that SPM can provide. The further development of TERS highly depends on the development of reliable methods for fabricating TERS tips and the in-depth understanding of the role of the tip in the TERS enhancement.

In this review, we discussed the role of the tip including the materials, radius, cone angle, and morphology in the TERS enhancement. It is clear that gold and silver are still the best materials to achieve a high enhancement in the visible light region. The tip apex with a smaller radius gives a higher spatial resolution. The choice of the tip radius depends on the application because there is a compromise between the enhancement and spatial resolution as a result of their different dependencies on the radius. An optimized TERS experiment not only involves the optimization of the tip parameters but also requires a good match between the incident laser and the LSPR of TERS systems.

We also introduced methods to fabricate TERS tips, including physical deposition (vacuum evaporation), chemical methods (EC etching, electrodeposition, electroless deposition, etc.), and microfabrication techniques (FIB, Ar+ ion sputtering, etc.). Vacuum evaporation is the most prevalent method of fabricating AFM TERS tips. Further optimization is still required to improve the reproducibility of the obtained TERS tip. Chemical methods are suitable for the fabrication of all kinds of TERS tips. Among them, electrochemical etching is particularly useful for preparing STM TERS tips. Microfabrication methods are a powerful but expensive tool, which can be used for preparing specialized TERS tips. In addition to all the above methods, the tip can be further refined by sharpening with FIB or modifying with high-voltage pulses to increase the TERS enhancement.

We further tackled the issues of lifetime and protection protocols of TERS tip, which have seldom been discussed in the TERS literature. Cleaning with concentrated sulfuric acid is an effective way to regenerate the Au tip with retained TERS activity. An ultrathin film of silver oxide naturally grown on the silver tip can provide good protection from contamination and further oxidation. The Al2O3 or SiOx can be used to improve the mechanical stability of metal-coated AFM TERS tips and extend the lifetime of TERS tips.

It should be noted that in spite of the progress in the understanding and fabrication of tips mentioned above, there are still some fundamental problems and challenges that need to be addressed before TERS, a promising nanoscale characterization tool, can exert its full potential.
  1. 1.

    Although ultrahigh sub-nanometer resolution has been achieved, the mechanism is still under hot debate [37,157,158]. The challenges lie in how to characterize the atomic structure of the very end of the tip and how to reliably simulate the electromagnetic response at atomic resolution precision without sacrificing the role of the rest of the bulky part of the tip [157].

     
  2. 2.

    Furthermore, when the tip is coupled with a substrate, the challenges are how to precisely simulate the distribution of the electromagnetic field in the narrow gap, how the molecular species (including the dipole orientation [158] and conductivity) will influence the enhancement effect, and how the far-field and near-field strengths contribute to the obtained TERS signal.

     
  3. 3.

    Most experimentalists are now doing TERS experiments by choosing tips from a batch of tips and directly measuring the TERS signal. This is a time-consuming approach. If certain physical properties of a tip can be related to the TERS enhancement by normal optical methods, this will accelerate the quality control of TERS and the TERS measurement. This requires a good understanding of the optical response of the tips and TERS system.

     
  4. 4.

    Concerning the fabrication of STM tips, it is still a great challenge to obtain a highly enhanced silver tip with a nice tip shape in a reproducible way. Furthermore, if the tip end of both silver and gold can be engineered to a specific shape and/or LSPR wavelength in a convenient way, it will also be very important to TERS.

     
  5. 5.

    Concerning the commercialization of TERS tips, there are some reports on the batch production of AFM-based TERS tips using special protocols and recently Bruker is even commercializing TERS tips [159]. However, the real commercialization of TERS tip requires the tip to have a good reproducibility of the signal enhancement to resist them wearing off during measurement. Most importantly, if the price of the commercial TERS tips can be lowered to that close to AFM tips (instead of 10 times higher), then true wide application of TERS may eventually come as the operator need not worry about breaking the tip.

     

Notes

Acknowledgement

We acknowledge the support from the Ministry of Science and Technology (MOST, 2011YQ03012406 and 2013CB933703), National Natural Science Foundation of China (NSFC; 21227004, 21321062, and J1310024), and Ministry of Education of the People’s Republic of China (MOE, IRT13036).

Conflict of interest

The authors declare no competing financial interest.

References

  1. 1.
    McCreery RL (2001) Meas Sci Technol 12(5):663CrossRefGoogle Scholar
  2. 2.
    Weber WH, Merlin R (2013) Raman scattering in materials science. Springer Science & Business Media, BerlinGoogle Scholar
  3. 3.
    Lawson EE, Barry BW, Williams AC, Edwards HGM (1997) J Raman Spectrosc 28(2–3):111–117CrossRefGoogle Scholar
  4. 4.
    Barron LD, Buckingham AD (1971) Mol Phys 20(6):1111–1119CrossRefGoogle Scholar
  5. 5.
    Fleischmann M, Hendra PJ, McQuillan AJ (1974) Chem Phys Lett 26(2):163–166CrossRefGoogle Scholar
  6. 6.
    Jeanmaire DL, Van Duyne RP (1977) J Electroanal Chem 84(77):1–20CrossRefGoogle Scholar
  7. 7.
    Nie S, Emory SR (1997) Science 275(5303):1102–1106CrossRefGoogle Scholar
  8. 8.
    Michaels AM, Nirmal M, Brus LE (1999) J Am Chem Soc 121(43):9932–9939CrossRefGoogle Scholar
  9. 9.
    Le Ru EC, Meyer M, Etchegoin PG (2006) J Phys Chem B 110(4):1944–1948CrossRefGoogle Scholar
  10. 10.
    Dieringer JA, Lettan RB, Scheidt KA, Van Duyne RP (2007) J Am Chem Soc 129(51):16249–16256CrossRefGoogle Scholar
  11. 11.
    Stockle RM, Suh YD, Deckert V, Zenobi R (2000) Chem Phys Lett 318(1–3):131–136CrossRefGoogle Scholar
  12. 12.
    Hayazawa N, Inouye Y, Sekkat Z, Kawata S (2000) Opt Commun 183(1–4):333–336CrossRefGoogle Scholar
  13. 13.
    Anderson MS (2000) Appl Phys Lett 76(21):3130–3132CrossRefGoogle Scholar
  14. 14.
    Pettinger B, Picardi G, Schuster R, Ertl G (2000) Electrochem Jpn 68:942–949Google Scholar
  15. 15.
    Zhang Z, Chen L, Sun M, Ruan P, Zheng H, Xu H (2013) Nanoscale 5(8):3249–3252CrossRefGoogle Scholar
  16. 16.
    Lantman EMV, Deckert-Gaudig T, Mank AJG, Deckert V, Weckhuysen BM (2012) Nat Nanotechnol 7(9):583–586CrossRefGoogle Scholar
  17. 17.
    Kumar N, Stephanidis B, Zenobi R, Wain AJ, Roy D (2015) Nanoscale 7(16):7133–7137CrossRefGoogle Scholar
  18. 18.
    Stadler J, Schmid T, Zenobi R (2011) ACS Nano 5(10):8442–8448CrossRefGoogle Scholar
  19. 19.
    Domke KF, Pettinger B (2009) J Raman Spectrosc 40(10):1427–1433CrossRefGoogle Scholar
  20. 20.
    Marquestaut N, Talaga D, Servant L, Yang P, Pauzauskie P, Lagugne-Labarthet F (2009) J Raman Spectrosc 40(10):1441–1445CrossRefGoogle Scholar
  21. 21.
    Ogawa Y, Toizumi T, Minami F, Baranov AV (2011) Phys Rev B 83(8):081302CrossRefGoogle Scholar
  22. 22.
    Lee N, Hartschuh RD, Mehtani D, Kisliuk A, Maguire JF, Green M, Foster MD, Sokolov AP (2007) J Raman Spectrosc 38(6):789–796CrossRefGoogle Scholar
  23. 23.
    Gucciardi PG, Valmalette JC (2010) Appl Phys Lett 97(26):263104CrossRefGoogle Scholar
  24. 24.
    Zhu L, Atesang J, Dudek P, Hecker M (2007) Mater Sci-Pol 25(1):19–31Google Scholar
  25. 25.
    Yeo BS, Amstad E, Schmid T, Stadler J, Zenobi R (2009) Small 5(8):952–960CrossRefGoogle Scholar
  26. 26.
    Xue L, Li W, Hoffmann GG, Goossens JGP, Loos J, de With G (2011) Macromolecules 44(8):2852–2858CrossRefGoogle Scholar
  27. 27.
    Domke KF, Zhang D, Pettinger B (2007) J Am Chem Soc 129(21):6708–6709CrossRefGoogle Scholar
  28. 28.
    Bailo E, Deckert V (2008) Angew Chem Int Ed 47(9):1658–1661CrossRefGoogle Scholar
  29. 29.
    Snopok B, Snitka V, Naumenko D, Bruzaite I, Serviene E (2013) Analyst 138(18):5371–5383CrossRefGoogle Scholar
  30. 30.
    Liu Z, Ding SY, Chen ZB, Wang X, Tian JH, Anema JR, Zhou XS, Wu DY, Ma BW, Xu X, Ren B, Tian ZQ (2011) Nat Commun 2:305CrossRefGoogle Scholar
  31. 31.
    Pettinger B, Ren B, Picardi G, Schuster R, Ertl G (2004) Phys Rev Lett 92(9):096101CrossRefGoogle Scholar
  32. 32.
    Pettinger B, Ren B, Picardi G, Schuster R, Ertl G (2005) J Raman Spectrosc 36(6–7):541–550CrossRefGoogle Scholar
  33. 33.
    Ren B, Picardi G, Pettinger B, Schuster R, Ertl G (2005) Angew Chem Int Ed 44(1):139–142CrossRefGoogle Scholar
  34. 34.
    Domke KF, Pettinger B (2009) ChemPhysChem 10(11):1794–1798CrossRefGoogle Scholar
  35. 35.
    Liu Z, Wang X, Dai K, Jin S, Zeng ZC, Zhuang MD, Yang ZL, Wu DY, Ren B, Tian ZQ (2009) J Raman Spectrosc 40(10):1400–1406CrossRefGoogle Scholar
  36. 36.
    Horimoto NN, Tomizawa S, Fujita Y, Kajimoto S, Fukumura H (2014) Chem Commun 50(69):9862–9864CrossRefGoogle Scholar
  37. 37.
    Zhang R, Zhang Y, Dong ZC, Jiang S, Zhang C, Chen LG, Zhang L, Liao Y, Aizpurua J, Luo Y, Yang JL, Hou JG (2013) Nature 498(7452):82–86CrossRefGoogle Scholar
  38. 38.
    Steidtner J, Pettinger B (2008) Phys Rev Lett 100(23):236101CrossRefGoogle Scholar
  39. 39.
    Zhang WH, Yeo BS, Schmid T, Zenobi R (2007) J Phys Chem C 111(4):1733–1738CrossRefGoogle Scholar
  40. 40.
    Krug JT, Sanchez EJ, Xie XS (2002) J Chem Phys 116(24):10895–10901CrossRefGoogle Scholar
  41. 41.
    Pettinger B, Schambach P, Villagómez CJ, Scott N (2012) Annu Rev Phys Chem 63(1):379–399CrossRefGoogle Scholar
  42. 42.
    Sun WX, Shen ZX (2003) J Opt Soc Am A 20(12):2254–2259CrossRefGoogle Scholar
  43. 43.
    Notingher I, Elfick A (2005) J Phys Chem B 109(33):15699–15706CrossRefGoogle Scholar
  44. 44.
    Demming AL, Festy F, Huang F, Richards D (2005) J Korean Phys Soc 47:S1–S4Google Scholar
  45. 45.
    Neacsu CC, Steudle GA, Raschko MB (2005) Appl Phys B Lasers Opt 80(3):295–300CrossRefGoogle Scholar
  46. 46.
    Picardi G, Nguyen Q, Ossikovski R, Schreiber J (2007) Appl Spectrosc 61(12):1301–1305CrossRefGoogle Scholar
  47. 47.
    Hayazawa N, Saito Y, Kawata S (2004) Appl Phys Lett 85(25):6239–6241CrossRefGoogle Scholar
  48. 48.
    Kazemi-Zanjani N, Vedraine S, Lagugné-Labarthet F (2013) Opt Express 21(21):25271–25276CrossRefGoogle Scholar
  49. 49.
    Yang ZL, Aizpurua J, Xu HX (2009) J Raman Spectrosc 40(10):1343–1348CrossRefGoogle Scholar
  50. 50.
    Roth RM, Panoiu NC, Adams MM, Osgood RM, Neacsu CC, Raschke MB (2006) Opt Express 14(7):2921–2931CrossRefGoogle Scholar
  51. 51.
    Hartschuh A, Sánchez EJ, Xie XS, Novotny L (2003) Phys Rev Lett 90(9):095503CrossRefGoogle Scholar
  52. 52.
    Pettinger B, Domke KF, Zhang D, Schuster R, Ertl G (2007) Phys Rev B 76(11):113409CrossRefGoogle Scholar
  53. 53.
    Pettinger B, Domke KF, Zhang D, Picardi G, Schuster R (2009) Surf Sci 603(10–12):1335–1341CrossRefGoogle Scholar
  54. 54.
    Yano TA, Ichimura T, Taguchi A, Hayazawa N, Verma P, Inouye Y, Kawata S (2007) Appl Phys Lett 91(12):121101CrossRefGoogle Scholar
  55. 55.
    Sun MT, Zhang ZL, Chen L, Xu HX (2013) Adv Opt Mater 1(6):449–455CrossRefGoogle Scholar
  56. 56.
    Stadler J, Oswald B, Schmid T, Zenobi R (2013) J Raman Spectrosc 44(2):227–233CrossRefGoogle Scholar
  57. 57.
    Uetsuki K, Verma P, Nordlander P, Kawata S (2012) Nanoscale 4(19):5931–5935CrossRefGoogle Scholar
  58. 58.
    Sun MT, Fang YR, Yang ZL, Xu HX (2009) Phys Chem Chem Phys 11(41):9412–9419CrossRefGoogle Scholar
  59. 59.
    Downes A, Salter D, Elfick A (2006) J Phys Chem B 110(13):6692–6698CrossRefGoogle Scholar
  60. 60.
    Zhang WH, Cui XD, Yeo BS, Schmid T, Hafner C, Zenobi R (2007) Nano Lett 7(5):1401–1405CrossRefGoogle Scholar
  61. 61.
    Downes A, Salter D, Elfick A (2006) Opt Express 14(12):5216–5222CrossRefGoogle Scholar
  62. 62.
    Poborchii V, Tada T, Kanayama T, Geshev P (2009) J Raman Spectrosc 40(10):1377–1385CrossRefGoogle Scholar
  63. 63.
    Taguchi A, Hayazawa N, Saito Y, Ishitobi H, Tarun A, Kawata S (2009) Opt Express 17(8):6509–6518CrossRefGoogle Scholar
  64. 64.
    Palik ED (1998) Handbook of optical constants of solids III. Academic, San DiegoGoogle Scholar
  65. 65.
    Cui XD, Zhang WH, Yeo BS, Zenobi R, Hafner C, Erni D (2007) Opt Express 15(13):8309–8316CrossRefGoogle Scholar
  66. 66.
    Yeo BS, Zhang WH, Vannier C, Zenobi R (2006) Appl Spectrosc 60(10):1142–1147CrossRefGoogle Scholar
  67. 67.
    Yeo BS, Schmid T, Zhang WH, Zenobi R (2007) Anal Bioanal Chem 387(8):2655–2662CrossRefGoogle Scholar
  68. 68.
    Demming AL, Festy F, Richards D (2005) J Chem Phys 122(18)Google Scholar
  69. 69.
    Meng LY, Huang TX, Wang X, Chen S, Yang ZL, Ren B (2015) Opt Express 23(11):13804–13813CrossRefGoogle Scholar
  70. 70.
    Downes A, Salter D, Elfick A (2006) Opt Express 14(23):11324–11329CrossRefGoogle Scholar
  71. 71.
    Zhang WH, Cui XD, Martin OJF (2009) J Raman Spectrosc 40(10):1338–1342CrossRefGoogle Scholar
  72. 72.
    Angulo AM, Noguez C, Schatz GC (2011) J Phys Chem Lett 2(16):1978–1983CrossRefGoogle Scholar
  73. 73.
    Le Ru EC, Etchegoin PG (2009) Principles of surface-enhanced Raman spectroscopy and related plasmonic effects. Elsevier, AmsterdamGoogle Scholar
  74. 74.
    Lin HX, Li JM, Liu BJ, Liu DY, Liu J, Terfort A, Xie ZX, Tian ZQ, Ren B (2013) Phys Chem Chem Phys 15(12):4130–4135CrossRefGoogle Scholar
  75. 75.
    Stiles PL, Dieringer JA, Shah NC, Van Duyne RP (2008) Annu Rev Phys Chem 1(1):601–626CrossRefGoogle Scholar
  76. 76.
    Schlücker S (2014) Angew Chem Int Ed 53(19):4756–4795CrossRefGoogle Scholar
  77. 77.
    Wang HT, Tian T, Zhang Y, Pan ZQ, Wang Y, Xiao ZD (2008) Langmuir 24(16):8918–8922CrossRefGoogle Scholar
  78. 78.
    Deckert V, Deckert-Gaudig T, Diegel M, Gotz I, Langeluddecke L, Schneidewind H, Sharma G, Singh P, Singh P, Trautmann S, Zeisberger M, Zhang Z (2015) Faraday Discuss 177:9–20CrossRefGoogle Scholar
  79. 79.
    Wang X, Liu Z, Zhuang MD, Zhang HM, Wang X, Xie ZX, Wu DY, Ren B, Tian ZQ (2007) Appl Phys Lett 91(10):101105CrossRefGoogle Scholar
  80. 80.
    Mino T, Saito Y, Verma P (2014) ACS Nano 8(10):10187–10195CrossRefGoogle Scholar
  81. 81.
    Novotny L, Bian RX, Xie XS (1997) Phys Rev Lett 79(4):645–648CrossRefGoogle Scholar
  82. 82.
    Novotny L, Stranick SJ (2006) Annu Rev Phys Chem 57(1):303–331CrossRefGoogle Scholar
  83. 83.
    Mishra N, Kumar GVP (2012) Plasmonics 7(2):359–367CrossRefGoogle Scholar
  84. 84.
    Blum C, Opilik L, Atkin JM, Braun K, Kämmer SB, Kravtsov V, Kumar N, Lemeshko S, Li JF, Luszcz K, Maleki T, Meixner AJ, Minne S, Raschke MB, Ren B, Rogalski J, Roy D, Stephanidis B, Wang X, Zhang D, Zhong JH, Zenobi R (2014) J Raman Spectrosc 45(1):22–31CrossRefGoogle Scholar
  85. 85.
    Bryant PJ, Kim HS, Zheng YC, Yang R (1987) Rev Sci Instrum 58(6):1115CrossRefGoogle Scholar
  86. 86.
    Melmed AJ (1991) J Vac Sci Technol B 9(2):601–608CrossRefGoogle Scholar
  87. 87.
    Ren B, Picardi G, Pettinger B (2004) Rev Sci Instrum 75(4):837–841CrossRefGoogle Scholar
  88. 88.
    Billot L, Berguiga L, de la Chapelle ML, Gilbert Y, Bachelot R (2005) Eur Phys J Appl Phys 31(2):139–145CrossRefGoogle Scholar
  89. 89.
    Wang X, Cui Y, Ren B (2007) Chem J Chinese Univ 28(3):522–525Google Scholar
  90. 90.
    Li ZL, Wu TH, Niu ZJ, Huang W, Nie HD (2004) Electrochem Commun 6(1):44–48CrossRefGoogle Scholar
  91. 91.
    Eligal L, Culfaz F, McCaughan V, Cade NI, Richards D (2009) Rev Sci Instrum 80(3):033701CrossRefGoogle Scholar
  92. 92.
    Kharintsev SS, Noskov AI, Hoffmann GG, Loos J (2011) Nanotechnology 22(2):025202CrossRefGoogle Scholar
  93. 93.
    Snitka V, Rodrigues RD, Lendraitis V (2011) Microelectron Eng 88(8):2759–2762CrossRefGoogle Scholar
  94. 94.
    Boyle MG, Feng L, Dawson P (2008) Ultramicroscopy 108(6):558–566CrossRefGoogle Scholar
  95. 95.
    Gingery D, Buehlmann P (2007) Rev Sci Instrum 78(11):113703CrossRefGoogle Scholar
  96. 96.
    Lopes M, Toury T, de La Chapelle ML, Bonaccorso F, Gucciardi PG (2013) Rev Sci Instrum 84(7):073702CrossRefGoogle Scholar
  97. 97.
    Park J, Hong T, Lee N, Kim K, Seo Y (2011) Curr Appl Phys 11(6):1332–1336CrossRefGoogle Scholar
  98. 98.
    Kharintsev S, Hoffmann G, Fishman A, Salakhov MK (2013) J Phys D Appl Phys 46(14):145501CrossRefGoogle Scholar
  99. 99.
    Roy D, Williams CM, Mingard K (2010) J Vac Sci Technol B 28(3):631–634CrossRefGoogle Scholar
  100. 100.
    Gorbunov AA, Wolf B, Edelmann J (1993) Rev Sci Instrum 64(8):2393CrossRefGoogle Scholar
  101. 101.
    Dickmann K, Demming F, Jersch J (1996) Rev Sci Instrum 67(3):845–846CrossRefGoogle Scholar
  102. 102.
    Iwami M, Uehara Y, Ushioda S (1998) Rev Sci Instrum 69(11):4010–4011CrossRefGoogle Scholar
  103. 103.
    Lloyd JS, Williams A, Rickman RH, McCowen A, Dunstan PR (2011) Appl Phys Lett 99(14):143108CrossRefGoogle Scholar
  104. 104.
    Hodgson PA, Wang Y, Mohammad AA, Kruse P (2013) Rev Sci Instrum 84(2):026109CrossRefGoogle Scholar
  105. 105.
    Ibe JP, Bey PP Jr, Brandow SL, Brizzolara RA, Burnham NA, Dilella DP, Lee KP, Marrian CRK, Colton RJ (1990) J Vac Sci Technol A 8(4):3570–3575CrossRefGoogle Scholar
  106. 106.
    Stadler J, Schmid T, Zenobi R (2010) Nano Lett 10(11):4514–4520CrossRefGoogle Scholar
  107. 107.
    Zhang C, Gao B, Chen LG, Meng QS, Yang H, Zhang R, Tao X, Gao HY, Liao Y, Dong ZC (2011) Rev Sci Instrum 82(8):083101CrossRefGoogle Scholar
  108. 108.
    Jiang N, Foley ET, Klingsporn JM, Sonntag MD, Valley NA, Dieringer JA, Seideman T, Schatz GC, Hersam MC, Van Duyne RP (2012) Nano Lett 12(10):5061–5067CrossRefGoogle Scholar
  109. 109.
    Sasaki SS, Perdue SM, Perez AR, Tallarida N, Majors JH, Apkarian VA, Lee J (2013) Rev Sci Instrum 84(9):096109CrossRefGoogle Scholar
  110. 110.
    Yi KJ, He XN, Zhou YS, Xiong W, Lu YF (2008) Rev Sci Instrum 79(7):073706CrossRefGoogle Scholar
  111. 111.
    You YM, Purnawirman NA, Hu HL, Kasim J, Yang HP, Du CL, Yu T, Shen ZX (2010) J Raman Spectrosc 41(10):1156–1162CrossRefGoogle Scholar
  112. 112.
    Fujita Y, Chiba R, Lu G, Horimoto NN, Kajimoto S, Fukumura H, Uji-i H (2014) Chem Commun 50(69):9839–9841CrossRefGoogle Scholar
  113. 113.
    Berweger S, Atkin JM, Olmon RL, Raschke MB (2010) J Phys Chem Lett 1(24):3427–3432CrossRefGoogle Scholar
  114. 114.
    Schlegel VL, Cotton TM (1991) Anal Chem 63(3):241–247CrossRefGoogle Scholar
  115. 115.
    Golan Y, Margulis L, Rubinstein I (1992) Surf Sci 264(3):312–326CrossRefGoogle Scholar
  116. 116.
    Zhang J, Matveeva E, Gryczynski I, Leonenko Z, Lakowicz JR (2005) J Phys Chem B 109(16):7969–7975CrossRefGoogle Scholar
  117. 117.
    Lüssem B, Karthäuser S, Haselier H, Waser R (2005) Appl Surf Sci 249(1–4):197–202CrossRefGoogle Scholar
  118. 118.
    DeRose JA, Thundat T, Nagahara LA, Lindsay SM (1991) Surf Sci 256(1–2):102–108CrossRefGoogle Scholar
  119. 119.
    Asghari-Khiavi M, Wood BR, Hojati-Talemi P, Downes A, McNaughton D, Mechler A (2012) J Raman Spectrosc 43(2):173–180CrossRefGoogle Scholar
  120. 120.
    Peng L, Lee H, Teizer W, Liang H (2009) Wear 267(5–8):1177–1180CrossRefGoogle Scholar
  121. 121.
    Barrios CA, Malkovskiy AV, Kisliuk AM, Sokolov AP, Foster MD (2009) J Phys Chem C 113(19):8158–8161CrossRefGoogle Scholar
  122. 122.
    Schmid T, Yeo BS, Leong G, Stadler J, Zenobi R (2009) J Raman Spectrosc 40(10):1392–1399CrossRefGoogle Scholar
  123. 123.
    Kawata S, Shalaev VM (2011) Tip enhancement. Elsevier, AmsterdamGoogle Scholar
  124. 124.
    Youssef KMS, Koch CC, Fedkiw PS (2004) Corros Sci 46(1):51–64CrossRefGoogle Scholar
  125. 125.
    Shacham–Diamand Y, Osaka T, Datta M, Ohba T (2009) Advanced nanoscale ULSI interconnects fundamentals and applications. Springer, New YorkCrossRefGoogle Scholar
  126. 126.
    Saito Y, Murakami T, Inouye Y, Kawata S (2005) Chem Lett 34(7):920–921CrossRefGoogle Scholar
  127. 127.
    Brejna PR, Griffiths PR (2010) Appl Spectrosc 64(5):493–499CrossRefGoogle Scholar
  128. 128.
    Zhang Q, Wu M, Zhao W (2005) Surf Coat Tech 192(2–3):213–219CrossRefGoogle Scholar
  129. 129.
    Saito Y, Wang JJ, Batchelder DN, Smith DA (2003) Langmuir 19(17):6857–6861CrossRefGoogle Scholar
  130. 130.
    Wang JJ, Saito Y, Batchelder DN, Kirkham J, Robinson C, Smith DA (2005) Appl Phys Lett 86(26):263111CrossRefGoogle Scholar
  131. 131.
    Vakarelski IU, Higashitani K (2006) Langmuir 22(7):2931–2934Google Scholar
  132. 132.
    Sqalli O, Bernal MP, Hoffmann P, Marquis-Weible F (2000) Appl Phys Lett 76(15):2134–2136Google Scholar
  133. 133.
    Umakoshi T, Yano TA, Saito Y, Verma P (2012) Appl Phys Express 5(5):052001–052003Google Scholar
  134. 134.
    Okamoto T, Yamaguchi I (2001) J Microsc-oxford 202(1):100–103Google Scholar
  135. 135.
    Farahani JN, Pohl DW, Eisler HJ, Hecht B (2005) Phys Rev Lett 95(1):017402CrossRefGoogle Scholar
  136. 136.
    Weber-Bargioni A, Schwartzberg A, Cornaglia M, Ismach A, Urban JJ, Pang YJ, Gordon R, Bokor J, Salmeron MB, Ogletree DF, Ashby P, Cabrini S, Schuck PJ (2011) Nano Lett 11(3):1201–1207CrossRefGoogle Scholar
  137. 137.
    Zou Y, Steinvurzel P, Yang T, Crozier KB (2009) Appl Phys Lett 94(17):171107CrossRefGoogle Scholar
  138. 138.
    De Angelis F, Das G, Candeloro P, Patrini M, Galli M, Bek A, Lazzarino M, Maksymov I, Liberale C, Andreani LC, Di Fabrizio E (2010) Nat Nanotechnol 5(1):67–72CrossRefGoogle Scholar
  139. 139.
    Comstock DJ, Elam JW, Pellin MJ, Hersam MC (2012) Rev Sci Instrum 83(11):113704CrossRefGoogle Scholar
  140. 140.
    Fleischer M, Weber-Bargioni A, Altoe MVP, Schwartzberg AM, Schuck PJ, Cabrini S, Kern DP (2011) ACS Nano 5(4):2570–2579CrossRefGoogle Scholar
  141. 141.
    Christiansen SH, Becker M, Fahlbusch S, Michler J, Sivakov V, Andra G, Geiger R (2007) Nanotechnology 18(3):035503CrossRefGoogle Scholar
  142. 142.
    Bao W, Melli M, Caselli N, Riboli F, Wiersma DS, Staffaroni M, Choo H, Ogletree DF, Aloni S, Bokor J, Cabrini S, Intonti F, Salmeron MB, Yablonovitch E, Schuck PJ, Weber-Bargioni A (2012) Science 338(6112):1317–1321CrossRefGoogle Scholar
  143. 143.
    Bek A, De Angelis F, Das G, Di Fabrizio E, Lazzarino M (2011) Micron 42(4):313–317CrossRefGoogle Scholar
  144. 144.
    Imad M, Atsushi T, Yuika S, Satoshi K, Verma P (2015) Appl Phys Express 8(3):032401CrossRefGoogle Scholar
  145. 145.
    Yang Y, Li ZY, Nogami M, Tanemura M, Huang Z (2014) RSC Adv 4(9):4718–4722CrossRefGoogle Scholar
  146. 146.
    Macpherson JV, Unwin PR (2000) Anal Chem 72(2):276–285CrossRefGoogle Scholar
  147. 147.
    Karral K, Grober RD (1995) Appl Phys Lett 66(14):1842–1844CrossRefGoogle Scholar
  148. 148.
    Zhang D, Wang X, Braun K, Egelhaaf HJ, Fleischer M, Hennemann L, Hintz H, Stanciu C, Brabec CJ, Kern DP, Meixner AJ (2009) J Raman Spectrosc 40(10):1371–1376CrossRefGoogle Scholar
  149. 149.
    Roy D, Wang J, Welland ME (2006) Faraday Discuss 132:215–225CrossRefGoogle Scholar
  150. 150.
    Neacsu CC, Dreyer J, Behr N, Raschke MB (2006) Phys Rev B 73(19):193406CrossRefGoogle Scholar
  151. 151.
    Kalkbrenner T, Ramstein M, Mlynek J, Sandoghdar V (2001) J Microsc-Oxford 202:72–76CrossRefGoogle Scholar
  152. 152.
    Christiane H, Lukas N (2008) Nanotechnology 19(38):384012CrossRefGoogle Scholar
  153. 153.
    Le NV, Mevellec JY, Minea T, Louarn G (2012) Int J Opt 2012:591083–591088Google Scholar
  154. 154.
    Johnson TW, Lapin ZJ, Beams R, Lindquist NC, Rodrigo SG, Novotny L, Oh SH (2012) ACS Nano 6(10):9168–9174CrossRefGoogle Scholar
  155. 155.
    Stadler J, Schmid T, Opilik L, Kuhn P, Dittrich PS, Zenobi R (2011) Beilstein J Nanotech 2(1):509–515CrossRefGoogle Scholar
  156. 156.
    Agapov RL, Sokolov AP, Foster MD (2013) J Raman Spectrosc 44(5):710–716CrossRefGoogle Scholar
  157. 157.
    Barbry M, Koval P, Marchesin F, Esteban R, Borisov AG, Aizpurua J, Sánchez-Portal D (2015) Nano Lett 15(5):3410–3419CrossRefGoogle Scholar
  158. 158.
    Zhang C, Chen BQ, Li ZY (2015) J Phys Chem C 119(21):11858–11871CrossRefGoogle Scholar
  159. 159.
    (2015) IRIS TERS probes. http://www.brukerafmprobes.com/t-IRIS-TERS-Probes.aspx. Accessed 14 Jan 2015

Copyright information

© Springer-Verlag Berlin Heidelberg 2015

Authors and Affiliations

  1. 1.State Key Laboratory of Physical Chemistry of Solid Surface, Collaborative Innovation Center of Chemistry for Energy Materials (iChEM), Key Laboratory of Analytical Sciences, Department of Chemistry, College of Chemistry and Chemical EngineeringXiamen UniversityXiamenChina

Personalised recommendations