Open Access
Add to BibSonomyAbstract
We report a novel fiber probe based Raman detection system on a microfluidic platform where a split Raman probe is directly embedded into a polydimethylsiloxane (PDMS) chip. In contrast to previous Raman detection schemes in microfluidics, probe based detection offers reduced background and portability. Compared to conventional backscattering probe designs, the split fiber probe we used in this system, results in a reduced size and offers flexibility to modify the collection geometry to minimize the background generated by the fibers. Also our microfluidic chip design enables us to obtain an alignment free system. As a proof of concept we demonstrate the sensitivity of the device for urea detection at relevant human physiological levels with a low acquisition time. The development of this system on a microfluidic platform means portable, lab on a chip devices for biological analyte detection and environmental sensing using Raman spectroscopy are now within reach.
©2010 Optical Society of America
1. Introduction
Raman spectroscopy has emerged as a powerful and effective tool for analytical studies of biological and chemical samples. Raman scattering refers to inelastic light scattering from a sample that may yield a molecular fingerprint of the constituent molecules. An inherent limitation of this spectroscopic technique is the low Raman cross section of bio-molecules and hence long integration times are required to obtain good signal to noise ratio. Nevertheless, Raman spectra have rich information content and a single Raman spectrum, owing to its high chemical specificity, can provide information about all the molecular constituents of the sample [1]. In the field of chemical analyses, the main driving factors have always been achieving fast yet sensitive measurements on miniaturized “Lab on a chip” devices [2]. The integration of Raman spectroscopic techniques with microfluidics devices opens up new possibilities in the field of bio-chemical detection.
In previous studies, Raman spectroscopy was combined with microfluidic systems for online and offline monitoring of chemical processes [3]. In order to overcome the limitation of the inherently low Raman cross section, Surface Enhanced Raman Spectroscopy (SERS) based detection schemes have been employed in microfluidic systems [4,5]. Other experiments used confocal Raman microscopy for online monitoring of chemical reactions [6,7]. Crucially, however, in all these applications, the monitoring was performed through a combination of a bulk Raman microscope and the microfluidic chip. The microfluidic chip essentially served only as a sample platform for a bulk optics system. One of the main problems in using microscope based systems to collect Raman data from such microfluidic chips is that the signal is acquired through a substrate which has its own background signal [3]. This limits the detection efficiency of the system. The microscope based approaches also preclude the miniaturization towards a true portable microfluidic-Raman system for on-chip monitoring and detection.
We report a wholly fiber probe based Raman detection device on a microfluidic platform, which is the first of its kind to the best of our knowledge. The probe based approach allows us to overcome the limitations of Raman spectroscopy to be used as a potential detection scheme in microfluidics. Since the probe is directly inserted into the microfluidic channel, the collected Raman spectra are free from any background signal coming from the substrate of microfluidic chip. Also, this fiber based system is completely alignment free making it more appealing for practical applications. In contrast to traditional back-scattering probe design, this device uses a split Raman probe, where the excitation and collection part of the probe is decoupled. The split probe heads serve to miniaturize the system and also allow the geometry where the fiber background signal is minimized. Their incorporation in a polydimethylsiloxane (PDMS) based microfluidic chip (Fig. 1 ) results in a novel alignment free robust system that can be used for quantitative analysis of bioanalytes in clinically relevant data acquisition time.
Fig. 1 [a] Design of the microfluidic chip. The head of the fiber probe is inserted to the chip and the analyte to be detected is injected to the chip through inlet and goes out through outlet [b] Photograph of the PDMS based chip where collection and excitation probes are inserted
To date, Raman fiber probe designs have been mainly restricted to process monitoring probes (InPhotonics Inc., MA, USA) and endoscopic detection and diagnosis probes [8–12]. In such designs, the Raman excitation and collection fibers are bundled together and lack flexibility of inspection of samples at different angles between the collection probe and the excitation probe. Our split Raman probe opens up new areas of application for Raman probe. This concept can lead to the development of Raman probe based, portable “lab on a chip” biochemical sensors for online and offline monitoring of analytes.
2. Experimental
2.1 Design of probe
In contrast to the existing fiber based Raman probes where the excitation and collection parts are integrated into a single console, in our probe, the excitation and collection parts are split into two which provides enhanced flexibility to the system. It is the design of the chip that decides the Raman detection geometry, offering the flexibility for inspection at several angles without modifying the probe design. The pre-aligned fiber channels in the chip which have identical dimensions as that of the probe head, allows insertion of the probe into the chip as shown in Fig. 1b. Thus resulting in a completely alignment free system.
In our split Raman fiber probe design, the excitation probe head contains a bandpass filter (3 nm bandwidth, Semrock, Inc. USA), which allows only light with wavelength 785nm to pass, placed between a pair of achromatic doublets (Comar Optics, UK). The collection probe head contains a long pass filter (cut off wavelength 795.2 nm, Semrock, Inc. USA), which blocks 785 nm light, also placed between a pair of achromatic doublets. The achromatic doublets in the probe heads are used for collimation of the light before the filters and subsequent focusing. This ensures that the filters work at their maximum efficiency. To minimise the fiber background, low OH multimode fibers (Polymicro Technologies, Arizona, USA) of core diameter 200 μm are used.
The main design constraint for the current probe is the diameter of the filters used. We use filters of diameter 2.8mm in order to keep the system cost effective, as obtaining filters of smaller sizes is prohibitively expensive. In order to match the filter diameter, we use fused silica achromatic doublets of 2.5mm diameter and 2.5mm focal length.
The throughput (Etendue) of an optical system is limited by the throughput of the most restrictive element and for an optical element the throughput is calculated by the following equation [8]
where, A is the collection area and Ω’ is its projected solid angle, given by,where NA is the numerical aperture of the element. In our present design, the throughput restricting element is the use of a collection probe with a single multimode fiber. The numerical aperture (NA) of the collection fiber we are using is 0.22, and the core diameter is 200μm. The throughput of the fiber is evaluated to be 0.0048 mm2.sr. The F number of our spectrometer is 4 and the slit height is 8mm. For a slit width of 200 μm, the throughput of this spectrometer would have been 0.0785 mm2.sr. However, since we need to couple our collection fiber to the spectrometer, the throughput of the spectrometer is matched with that of the fiber using a pair of lenses which acts as an F-number matcher. Thus the resulting throughput of the spectrometer is also 0.0048 mm2sr. As mentioned before, in order to satisfy the design constraint introduced by the size of the filter, we use a 2.5mm diameter achromatic doublet for collection, which has ~0.5NA. Therefore the throughput of the collecting lens is higher compared to that of the collection fiber, which introduces a mismatch. However, the throughput of the whole system could be improved in future by using multiple fibers for collection in contrast to the present single fiber design.The overlap volume of both excitation and collection probes can be used to estimate the overall collection volume for this device. The lenses at the tip of the probes are positioned in such a way that at orthogonal collection geometry, the focal points of the excitation and collection probes overlap. The beam diameter at the focal spot is ~200 μm. Hence at orthogonal geometry, collection volume could be approximated as a cube of edge 200 μm. The collection volume, from where the Raman spectrum of the sample is collected is estimated to be ~8 nanolitres.
2.2 Design and fabrication of chip
The design of the chip with orthogonal collection geometry is given in Fig. 1a. The chip contains two sets of channels, the fiber probe channels and the fluidic channel. Both channels are fabricated with same dimension as that of the fiber probe head which has an outer diameter of 3mm in the current design. Size of the channel is solely dictated by the outer diameter of the probe heads. The excitation probe and the collection probe are inserted into the chip through the fiber channels (Fig. 1b) and the detection analyte is flown through the fluidic channel. The physical dimension of the chip with inserted probe heads is 25mm × 30mm.
The microfluidic chip is fabricated with PDMS using conventional soft lithographic techniques [13]. The mould for the chip is fabricated by adhering 3mm metallic sleeves onto a glass substrate using a medical epoxy (Loctite M-31CL Hysol), which defines the probe entrance channel and fluidic channel. This makes it possible to insert the probe into the chip conveniently and the channel design ensures the relative alignment of excitation probe to the collection probe. In principle, with this same approach, it is possible to reduce the channel size to the order of 100μm. Hence the same fabrication method may readily be extended to fabricate chips of smaller footprints.
2.3 Detection device
Figure 2 illustrates the portable microfluidic Raman detection system. We use upto 200 mW power from a 785nm diode laser (Laser2000 (UK) Ltd.) for Raman excitation at the excitation probe head. The Raman excitation and the collection probes are coupled to the laser and the spectrometer (Shamrock SR-303i, Andor Technology) respectively using SMA connectors (Thorlabs Inc., USA). The spectrometer employs a 400 lines/mm grating, blazed at 850 nm and is equipped with a deep depletion, back-illuminated and thermoelectrically cooled CCD camera (Newton, Andor Technology) for the detection of Raman signal. The resolution of the Raman system is measured by the FWHM of the Silicon Raman peak at 520 cm−1 and is found to be better than 6 cm−1.
The physical dimension of the whole system including the PDMS chip, Raman probe, laser source, spectrometer and CCD is 60cm × 35cm × 25cm. The physical dimensions and also the alignment free nature of this fiber probe based system make it a portable device. Our probe based approach gives flexibility to the system such that it can reach otherwise inaccessible areas, a feature which has potential applications in online monitoring of the samples.
3. Results and Discussion
3.1Optimum collection geometry
The split Raman probe design offers the flexibility to implement various Raman detection geometries. As a proof of principle, we explore two geometries based on the relative orientation of the excitation and collection probes. In the first geometry, the excitation probe and the collection probe are aligned collinearly, resulting in forward collection of Raman signal. The second geometry aligns the excitation and collection probes perpendicular to each other in such a way that the Raman signal is collected orthogonally as shown in Fig. 1a. The performance characteristics of both of these geometries are compared by taking Raman spectra of ethanol. Figure 3 shows a comparison of the Raman spectra collected in these two geometries. Each Raman spectrum is normalized with respect to its maximum intensity value. It is evident that the spectra recorded with the 90° (orthogonal) collection geometry have substantially lower background signal than the spectra collected with the 180° geometry. This background is primarily due to the fluorescence excited in the collection probe by forward scattering 785nm light, which leaks through the imperfections in the periphery of the collection probe head.
Fig. 3 Raman spectra of ethanol recorded for an acquisition time of 5s for probes at 90° and 180° orientations
In order to verify the source of enhanced background at 180° geometry, the Raman spectra was collected in the above specified geometries for de-ionized water and air (when there is no sample inside the chip). It was observed that the background signal is high for 180° geometry, which confirms that the major contribution for the fluorescent background is from the fiber. The experiment with air showed no Raman peaks corresponding to PDMS, which ensures that the obtained Raman spectra does not have any cross talk signals from the material of the chip, unlike the other approaches where detection is performed externally.
3.2 Detection limit of the integrated probe system
The detection sensitivity of the system for measuring bio-analytes is demonstrated by detecting urea (Sigma Aldrich Inc.) close to physiological level of human urine. The concentration of urea in human fluids is an important indicator of proper kidney function in mammals [14]. The sample solution of urea is gravimetrically prepared in deionised water for various concentrations ranging from 0.05M to 1M. The Raman spectra of the samples are acquired in the probe based microfluidic chip with an acquisition time of 5s each.
To calculate the detection limit of a Raman spectroscopy based system, limiting uncertainty in concentration detection, denoted by Δc [15], is estimated as below,
where, s is the signal strength of the analyte of interest at unit concentration, σ is the measurement noise and olf is the overlap factor that indicates the amount of non-orthogonality between the analyte of interest and the spectral interferents. The overlap factor can range from 1 (no overlap with interfering agents) to ∞ (complete overlap). Since we are analysing a single component analyte, olf = 1. In the absence of modelling, the measurement noise σ is evaluated by calculating the standard deviation of 10 spectra at the spectral band corresponding to 999 cm−1 in Raman spectra of urea at 1M concentration. For our study we estimated σ = 18 (photon counts) and sk = 118 (photon counts/M), giving Δck = 0.152M.We also calculated the minimum detection limit of our system by estimating the Noise Equivalent Concentration (NEC). NEC refers to the limiting value of concentration of the analyte of interest when signal level from the analyte of interest is just equal to the measurement noise, i.e. SNR = 1. For the SNR calculation, the signal is calculated as the norm of mean of 10 measurements at a particular concentration, over the spectral bands where the signature of the analyte of interest appears. The noise is calculated as the norm of the standard deviation of the 10 measurements over the same spectral band. Here we used the well known characteristic peak of urea in the Raman fingerprint region at 999 cm−1 for the above mentioned estimation. The SNR vs. concentration plot is given in Fig. 4 . The value of NEC from our measurements is 0.144M. It has to be noted that, in the absence of any modelling, the value of Δc approaches the value of NEC and our calculations confirm this.
Fig. 4 Plot of concentration vs. SNR for Raman spectra of Urea. The ‘ + ’ symbol represents the SNR measured at a particular concentration from 10 spectra recorded with 5s acquisition time each. The solid line represents the linear fit for the evaluated SNR data. The dotted line represent the limit of detection where SNR = 1. The Noise Equivalent Concentration (NEC) is evaluated as 0.144 from the plot.
Thus the results show that the minimum detection limit of our current system to detect urea is ~0.15M for an acquisition time of 5s. Figure 4 shows that, with the estimated system performance characteristics, our system is capable of detecting concentration of urea at physiological level of human urine [14]. The detection limit of our system is two orders lower compared to the state of the art bulk optics based Raman detection systems [16]. However compared to the state of the art bulk optics systems, our system offers portability and alignment free detection of analyte. The detection limit of the present system can be further improved by improving the collection efficiency of the Raman probe and increasing the throughput of the system. Currently work is in progress in this regard.
3.3 Effect of flow
In order to ensure the mechanical robustness of this device in flow based systems, it is necessary to evaluate the sensitivity under flow conditions. Ethanol is flown through the microfluidic chip at different flow rates and Raman spectra are recorded for an acquisition time of 5s each. The intensity, averaged over 10 spectra corresponding to one flow rate, of the strongest Raman peak at 884 cm−1 is plotted against different flow rates as shown in Fig. 5 . From Fig. 5 it is clear that the sensitivity of the system is not affected from the changing flow rates. The result shows that the system is leak proof under pressure and there is no misalignment for the collection and excitation fiber due to the pressure introduced by the flow.
3.4 Residual free nature of the chip
One chip can be used for detecting different samples, with proper rinsing procedure after flowing each chemical through the chip. The rinsing procedure with de-ionized water keeps the system contamination free. In order to verify the residual free nature of the chip, ethanol is flown through the chip, followed by de-ionized water for rinsing. Subsequently 0.4M urea is flown through the chip. The Raman spectra of urea are recorded with an acquisition time of 5s. We observe no peaks from ethanol in the recorded Raman spectra of urea, which proves that our rinsing procedure makes the chip 100% residual free and reusuable.
4. Conclusions
We have demonstrated a novel fiber probe based microfluidic Raman sensor. The capability of the system to be used for qualitative and quantitative analysis has been demonstrated. As a potential application, the capability of the system to detect urea at physiologically relevant levels has been shown. The minimum detection limit of our system to detect urea is estimated to be ~0.15M. The sensitivity of the system may also be enhanced by incorporating latest modulation techniques to the Raman excitation signal [17]. The mechanical robustness of the system under various flow rates is verified, enabling the device to be used for online monitoring in flow systems. The direct combination of a split Raman probe design and microfluidic platform enables the development of powerful, portable bio-chemical sensing devices. For some applications, the sensitivity can be orders of magnitude better if resonance Raman spectroscopy or SERS is utilised. The prospect of a Raman detection system built on PDMS based microfluidic platform makes this technology inexpensive and biocompatible. This approach will help to develop point of care, environmental microfluidic monitoring, biotechnological and forensic portable detection devices utilizing Raman spectroscopy.
Acknowledgments
We thank CTA Brown and TF Krauss for critical reading of the manuscript and useful discussions. We acknowledge the UK EPSRC and the EU FP7 STREP programme “Araknes” for funding. KD is a Royal Society-Wolfson Merit Award Holder. PCA and GS contributed equally to the work presented.
References and links
1. K. Kneipp, H. Kneipp, I. Itzkan, R. R. Dasari, and M. S. Feld, “Ultrasensitive chemical analysis by Raman spectroscopy,” Chem. Rev. 99(10), 2957–2976 (1999). [CrossRef]
2. R. Daw and J. Finkelstein, “Lab on a chip,” Nature 442(7101), 367–367 (2006). [CrossRef]
3. P. J. Viskari and J. P. Landers, “Unconventional detection methods for microfluidic devices,” Electrophoresis 27(9), 1797–1810 (2006). [CrossRef] [PubMed]
4. R. M. Connatser, L. A. Riddle, and M. J. Sepaniak, “Metal-polymer nanocomposites for integrated microfluidic separations and surface enhanced Raman spectroscopic detection,” J. Sep. Sci. 27(17-18), 1545–1550 (2004). [CrossRef]
5. L. X. Quang, C. Lim, G. H. Seong, J. Choo, K. J. Do, and S. K. Yoo, “A portable surface-enhanced Raman scattering sensor integrated with a lab-on-a-chip for field analysis,” Lab Chip 8(12), 2214–2219 (2008). [CrossRef] [PubMed]
6. K. Ramser, J. Enger, M. Goksör, D. Hanstorp, K. Logg, and M. Käll, “A microfluidic system enabling Raman measurements of the oxygenation cycle in single optically trapped red blood cells,” Lab Chip 5(4), 431–436 (2005). [CrossRef] [PubMed]
7. S. A. Leung, R. F. Winkle, R. C. R. Wootton, and A. J. deMello, “A method for rapid reaction optimisation in continuous-flow microfluidic reactors using online Raman spectroscopic detection,” Analyst (Lond.) 130(1), 46–51 (2005). [CrossRef]
8. J. T. Motz, M. Hunter, L. H. Galindo, J. A. Gardecki, J. R. Kramer, R. R. Dasari, and M. S. Feld, “Optical fiber probe for biomedical Raman spectroscopy,” Appl. Opt. 43(3), 542–554 (2004). [CrossRef] [PubMed]
9. A. Mahadevan-Jansen, M. F. Mitchell, N. Ramanujam, U. Utzinger, and R. Richards-Kortum, “Development of a fiber optic probe to measure NIR Raman spectra of cervical tissue in vivo,” Photochem. Photobiol. 68(3), 427–431 (1998). [CrossRef] [PubMed]
10. H. P. Buschman, E. T. Marple, M. L. Wach, B. Bennett, T. C. Schut, H. A. Bruining, A. V. Bruschke, A. van der Laarse, and G. J. Puppels, “In vivo determination of the molecular composition of artery wall by intravascular Raman spectroscopy,” Anal. Chem. 72(16), 3771–3775 (2000). [CrossRef] [PubMed]
11. U. Utzinger and R. R. Richards-Kortum, “Fiber optic probes for biomedical optical spectroscopy,” J. Biomed. Opt. 8(1), 121–147 (2003). [CrossRef] [PubMed]
12. Y. Komachi, T. Katagiri, H. Sato, and H. Tashiro, “Improvement and analysis of a micro Raman probe,” Appl. Opt. 48(9), 1683–1696 (2009). [CrossRef] [PubMed]
13. J. C. McDonald, D. C. Duffy, J. R. Anderson, D. T. Chiu, H. Wu, O. J. A. Schueller, and G. M. Whitesides, “Fabrication of microfluidic systems in poly(dimethylsiloxane),” Electrophoresis 21(1), 27–40 (2000). [CrossRef] [PubMed]
14. B. X. Yang and L. Bankir, ““Urea and urine concentrating ability: new insights from studies in mice,” Am,” J. Physiol-Renal 288(5), F881–F896 (2005). [CrossRef]
15. I. Barman, G. P. Singh, R. R. Dasari, and M. S. Feld, “Turbidity-corrected Raman spectroscopy for blood analyte detection,” Anal. Chem. 81(11), 4233–4240 (2009). [CrossRef] [PubMed]
16. J. W. McMurdy 3rd and A. J. Berger, “Raman spectroscopy-based creatinine measurement in urine samples from a multipatient population,” Appl. Spectrosc. 57(5), 522–525 (2003). [CrossRef] [PubMed]
17. A. C. De Luca, M. Mazilu, A. Riches, C. S. Herrington, and K. Dholakia, “Online fluorescence suppression in modulated Raman spectroscopy,” Anal. Chem. 82(2), 738–745 (2010). [CrossRef]
