Review
Recent developments in optofluidic-assisted Raman spectroscopy

https://doi.org/10.1016/j.pquantelec.2012.11.001Get rights and content

Abstract

This paper reviews and compares the different optofluidic techniques for enhancing the retrieved Raman signal in liquids with a focus on aqueous solutions. Recent progress in characterizing different nanostructures and biological molecules utilizing optofluidic fibers such as photonic crystal fibers (PCFs) in Raman spectroscopy are discussed. Techniques and applications to combine surface enhanced Raman spectroscopy (SERS) with optofluidic-assisted Raman spectroscopy are further reviewed. Finally, challenges and future opportunities to advance Raman spectroscopy combined with optofluidics are presented.

Introduction

Raman spectroscopy has witnessed significant advances in its applicability since the discovery of the Raman effect by Sir C. V. Raman in 1928, and has established itself as a valuable analytical technique in a large number of different fields. It is currently being applied but not limited to the fields of art and archeology [1], [2], [3], [4], [5], pharmaceuticals [6], [7], [8], chemistry [9], [10], [11], [12], biology [8], [12], [13], [14], diagnostics [14], [15], [16], material science [17], forensics [18], [19], [20], and environmental monitoring [21], [22], [23]. Raman spectroscopy is known for its ability to rapidly and non-destructively provide information on the molecular vibrations of materials with exceptional specificity. Since molecular vibrations are specific to the molecular bonds and their symmetries, they are unique to every type of material at the molecular level. Therefore, spectra obtained from Raman spectroscopy, namely the Raman scattering, can also be described as the ‘fingerprint’ of materials. The collection of Raman spectra can be used to identify materials, as well as their properties such as stress/strain, doping, and crystallinity. However, the Raman effect is also an extremely weak process in which only one in 10 million photons are Raman scattered [24]. This low probability of the Raman scattering effect is responsible for the low sensitivity of Raman spectroscopy. As a result, this technique suffers from the background superimposed on the Raman modes by other stronger optical phenomena such as fluorescence. In the past two decades, advances in optical technologies and instrumentation have greatly improved the sensitivity of Raman spectroscopy. Until this present day, the sensitivity of Raman spectroscopy remains to be one of the challenges that is being improved through research in both academia and industry.

Moreover, owing to the weak sensitivity of Raman spectroscopy, its applications have been, and still are, largely limited to solid samples. In the liquid and gas state of materials, the density of the molecules is much lower than that in the solid state. In Raman spectroscopy, this means that there are fewer molecules to interact with the pump laser or excitation source to generate the Raman scattered signal. Thus, the Raman scattering effect is weaker in liquids, and it is even lower in weakly concentrated solutions and gases.

There are many variations of Raman spectroscopy developed with the aim of improving its sensitivity for samples in the liquid and gaseous state. For example, resonant Raman (RR) spectroscopy can be used to enhance the intensity of the Raman signal by tuning the pump laser frequency to match the excited electronic state of the molecules in question. This technique can enhance the Raman intensity by 2–6 orders of magnitude [25]. However, only a small number of Raman modes can be enhanced at any given time due to the frequency-matching condition. Moreover, a tunable laser with high power is usually required which is costly and often leads to photo-degradation of the sample.

Surface enhanced Raman spectroscopy (SERS) and tip-enhanced Raman spectroscopy (TERS) are two other techniques in which metallic nanostructures in close proximity, or adsorbed, to the analyte can improve the Raman signal through electromagnetic and chemical enhancement. In particular, SERS has demonstrated enhancement factors up to 9–10 orders of magnitude which enabled the detection of a single molecule [26], [27], [28], [29]. However, metallic nanostructures are required to be mixed with the analyte to achieve this enhancement, which could alter the structure, chemical environment, or chemical properties of the analyte from its as-synthesized state due to conformational changes when interacting with metallic nanostructure. For example, hotspots created from metallic nanostructures could exert physical stress on the analyte, causing molecular reorientations [30]. Consequently, SERS often gives different Raman modes than normal Raman spectroscopy [31]. The positions and full width at half maximums (FWHMs) of the Raman modes, as well as the relative intensities between Raman modes could also be different between SERS spectra and normal Raman spectra. These differences further vary depending on the metallic structure used in the SERS measurement which makes analysis and interpretation of the spectra more difficult [32]. In addition, SERS only enhances some Raman modes of the analyte while suppressing the others depending on the orientation of the analyte relative to the metallic nanostructure [33], [34]. Enhancements are also often not reproducible due to instability of the hotspots created [35]. In addition, the maximal enhancement factor could only be obtained through careful design of the metallic nanoparticles and tailored for a specific target analyte [36]. Furthermore, the metallic nanostructure might create additional background signal in the Raman spectrum, which would further complicate the analysis and interpretation of the Raman spectra [32].

For the aforementioned techniques, the efficiency in the process of generating and collecting Raman scatters in the practical scheme is not addressed, especially for liquid and gas samples. In conventional Raman spectroscopy, RR and SERS, the pump laser or excitation source is focused directly into the analyte to generate Raman scattering (Fig. 1). Most of the Raman signal retrieved at the detector is scattered from the beam waist of the pump laser at which the pump laser is most intense and the power density is the highest. In this case, the pump laser and the analytes are only interacting in the volume limited by the spot size of the pump laser in the lateral direction and the depth of field in the axial direction. Since Raman signals are scattered in omni-directions, only a fraction of the scattered signal is collected by the objective and detected by the detector due to the limited numerical aperture (NA) of the objective lens. In the case of a solution, the amount of Raman scattering detected from the small quantity of analyte dispersed in the buffer solution is minimal.

An alternative to increasing the sensitivity of Raman spectroscopy (not just for a few Raman modes of the analyte, but for all the Raman modes generated) while not altering the native state of the analyte is to incorporate optofluidic devices into the Raman spectroscopy setup as the interaction medium. Optofluidic devices enable the confinement of both the laser light and the liquid of interest into the same cavity, enabling a strong and efficient process of light–matter interaction. Fig. 2 shows an example of a Raman spectroscopy setup utilizing an optofluidic fiber as the light–matter interaction medium in the backscattering configuration. In this example, the analyte is filled into the central core of the optofluidic fiber from the bottom end through capillary action. At the other end the pump laser light is focused into the same central core through the objective to generate Raman scattering signals from the analyte. Due to the strong confinement of the pump laser light in the optofluidic fiber, the pump laser light maintains a strong power density in the central core. As both the pump laser light and the analyte are confined within the same central core of the optofluidic fiber, the strongly confined pump laser can interact with the analyte throughout the entire fiber length. Thus, Raman signals are scattered throughout the entire length of the fiber as opposed to just the depth of field of the objective in the conventional scheme. Since Raman signals are mostly shifted by less than several tens of nanometers from the wavelength of the pump laser, they will also be confined inside the liquid core and collected throughout the fiber. As a result, the output signal from the fiber end will be collected more efficiently by the objective for detection; thus an enhanced Raman scattering signal could be retrieved at the detector.

An advantage to this technique is that prior knowledge of the analyte's electronic structure and surface chemistry is not required to enhance the retrieved Raman signal assuming that the pump laser light is within the guidance band of the optofluidic fiber after the liquid is injected. In RR and SERS, prior knowledge of the electronic structure and surface chemistry of the analyte is required to achieve Raman enhancement, respectively, which is not possible in applications where the exact composition of the sample is not known (e.g. forensic analysis).

In addition, most optofluidic fibers have core diameters on the order of a few microns. This means that only nano-liters to micro-liters of sample volumes are required for analysis. Furthermore, the advantage of long interaction length in optofluidic fibers can be extended through simply increasing the length of the optofluidic fiber. A longer interaction length can further enhance the retrieved Raman signal up to a certain length at which the enhancement factor will saturate due to large propagation losses. The relationship between the enhancement factor and the length of the optofluidic fiber will be discussed in detail in Section 2.5.

It is also important to note that the pump laser is confined inside the liquid core of the optofluidic fiber; thus, interactions between the pump laser light and the sidewall of the fiber are minimal. Therefore, the resulting interference due to the fiber material is limited in the presence of stronger Raman scattering signal from the analyte and any background. Furthermore, the use of optofluidic fiber does not increase the scattering cross-section, since optofluidic-assisted Raman spectroscopy enhances the ‘retrieved Raman signal’ (i.e. the Raman scattering signal retrieved at the detector). This is in contrast to techniques such as SERS or TERS where the Raman scattering cross-section is increased to achieve the enhancement in the Raman signal. Optofluidic devices described here are not limited to the fiber form. However, in this review, we will focus on optofluidic devices in the fiber form as the benefit of increased light–matter interaction could be obtained with fibers with centimeters in length. More importantly, further enhancement can be easily achieved with fiber length extended to meters without significant increases in complexity.

In this review, we will first describe and compare the different optofluidic devices used to enhance the retrieved Raman signal of liquid samples. Then, applications and recent progress in utilizing optofluidic-assisted Raman spectroscopy for characterizing nanostructures and biological molecules are reviewed. Finally, different techniques to combine SERS with optofluidic devices to achieve an ultimate sensitivity in Raman spectroscopy are discussed, followed by a discussion of the outstanding challenges and opportunities for this platform.

Section snippets

Capillary tubes

A capillary tube is the simplest form of optofluidic devices that provides a long path-length (i.e. centimeters to meters) for light–matter interaction. Capillary tubes are usually composed of a cladding made out of a solid material and a hollow core in which the liquid of interest is filled into for Raman analysis. When a liquid with a refractive index that is higher than that of the cladding is filled into the core of the capillary tube, the pump laser light can be guided through and interact

Characterizing colloidal nanostructures

In the last two decades, there has been a tremendous interest in synthesizing nanometer-sized semiconductor and metallic structures suspended in solution [85], [86], [87], [88], [89]. These novel colloidal nanostructures possess intriguing optical properties of quantum-confined particles as well as practical advantages of solution-based processing. Advancements in synthesis techniques have enabled simple and cost-effective ways of synthesizing small, monodisperse and water-soluble

Characterizing biological molecules

Raman spectroscopy is routinely used to identify and distinguish between different biological macromolecules. In conventional Raman spectroscopy (i.e. without the integration of optofluidic platforms or SERS enhancements), relatively large laser powers are typically employed to obtain a spectrum with high signal-to-background ratio [129], [130], [131], [132]. However, these high laser powers may cause thermally induced destruction of the secondary or tertiary structures adopted by these

Combining surface enhanced Raman spectroscopy with optofluidic devices

When optofluidic devices are combined with SERS, Raman analysis with sensitivity higher than that of each technique alone could potentially be achieved. This is ideal for identifying analytes with small Raman cross-sections, as well as interactions between analytes, in which extremely high sensitivity is required to obtain a strong and clear Raman spectrum. Since detailed analysis of the Raman modes is typically not required in material identification; changes in Raman modes’ positions, FWHMs,

Summary

The integration of different optofluidic fibers with Raman spectroscopy to achieve enhancement in the retrieved Raman signal was reviewed. Recent progress in utilizing this novel enhancement technique for various applications in examining nanomaterials and biological samples was presented. Furthermore, recent efforts and applications in combining SERS with optofluidic fibers to achieve superior sensitivity in Raman spectroscopy were also reviewed.

Despite the advancements in this novel field of

Acknowledgments

This work was supported by Ontario Centers of Excellence and the Natural Sciences and Engineering Research Council of Canada.

References (178)

  • M. Pérez-Alonso et al.

    Analytica Chimica Acta

    (2006)
  • T. Vankeirsbilck et al.

    TrAC Trends in Analytical Chemistry

    (2002)
  • E.V. Efremov et al.

    Analytica Chimica Acta

    (2008)
  • M. Li et al.

    Current Opinion in Biotechnology

    (2012)
  • A.G. Brolo et al.

    Journal of Electroanalytical Chemistry

    (1996)
  • R.M. Abu-Ghazalah et al.

    Biophysical Chemistry

    (2010)
  • J. Liu et al.

    Journal of Colloid and Interface Science

    (2010)
  • M.S. Abd El-sadek et al.

    Current Applied Physics

    (2010)
  • S. Kim et al.

    Nature Biotechnology

    (2004)
  • P. Baraldi et al.

    Journal of Raman Spectroscopy

    (2008)
  • P. Vandenabeele et al.

    Journal of Raman Spectroscopy

    (2001)
  • P. Vandenabeele et al.

    Analytical and Bioanalytical chemistry

    (2007)
  • P. Vandenabeele et al.

    Analytical and Bioanalytical chemistry

    (2004)
  • B.D. Patel et al.

    Current Pharmaceutical Analysis

    (2010)
  • P. Matoušek et al.

    Emerging Raman Applications and Techniques in Biomedical and Pharmaceutical Fields

    (2010)
  • K. Nakamoto

    Infrared and Raman Spectra of Inorganic and Coordination Compounds

    (2009)
  • A.M. Kelley

    Annual Review of Physical Chemistry

    (2010)
  • K. Kneipp et al.

    Journal of Physics: Condensed Matter

    (2002)
  • A. Downes et al.

    Sensors

    (2010)
  • A. Kumar et al.

    Journal of Nanomaterials

    (2011)
  • L.M. Almond et al.

    Journal of Biophotonics

    (2011)
  • M. Tanaka et al.

    Journal of Materials Science

    (2006)
  • J.M. Chalmers et al.

    Infrared and Raman Spectroscopy in Forensic Science

    (2012)
  • V. Sikirzhytski et al.

    Applied Spectroscopy

    (2011)
  • D.S. Moore et al.

    Analytical and Bioanalytical Chemistry

    (2009)
  • A. Sarmiento et al.

    Journal of Raman Spectroscopy

    (2008)
  • U. Villanueva et al.

    Journal of Raman Spectroscopy

    (2008)
  • D. Larroumet et al.

    Journal of Raman Spectroscopy

    (2008)
  • C.F. Bohren et al.

    Absorption and Scattering of Light by Small Particles

    (1983)
  • H.A. Szymanski

    Raman Spectroscopy: Theory and Practice

    (1967)
  • K. Kneipp et al.

    Physical Review Letters

    (1997)
  • E.C.L. Ru et al.

    Annual Review of Physical Chemistry

    (2012)
  • E.C.L. Ru et al.

    Journal of Physical Chemistry C

    (2007)
  • N.P. Pieczonka et al.

    Chemical Society Reviews

    (2008)
  • T. Chen et al.

    ACS Nano

    (2010)
  • J. Guicheteau et al.

    Proceedings of SPIE

    (2006)
  • W. Xua et al.

    Proceedings of the National Academy of Science USA

    (2012)
  • P.L. Stiles et al.

    Annual Review of Analytical Chemistry

    (2008)
  • G. Chen et al.

    Journal of the American Chemical Society

    (2010)
  • S.L. Kleinman, R.R. Frontiera, A.-I. Henry, J.A. Dieringer, R.P.V. Duyne, Phys. Chem. Chem. Phys. 15 (2012)...
  • G.E. Walfaren et al.

    Spectroscopy

    (1972)
  • H.B. Ross et al.

    Applied Spectroscopy

    (1981)
  • G.E. Walfaren

    Applied Spectroscopy

    (1975)
  • W.J. Schmid et al.

    Journal of Raman Spectroscopy

    (1981)
  • Z. Li et al.

    Chinese Physics Letters

    (1993)
  • R. Altkorn

    Applied Optics

    (1997)
  • L. Song et al.

    Applied Spectroscopy

    (1998)
  • D. Qi et al.

    Journal of Biomedical Optics

    (2005)
  • M. Holtz et al.

    Analytical Chemistry

    (1999)
  • Y.-J. Tian et al.

    Applied Physics B

    (2007)
  • Cited by (33)

    View all citing articles on Scopus
    View full text