rapid communication
Total Internal Reflection Fluorescence Spectrometry Using a Dual Optical Fiber Sample Cell Cole A. Guyer, Shiquan Tao* Department of Mathematics, Chemistry and Physics, West Texas A&M University, WTAMU Box 60787, Canyon, TX 79016 USA
A novel total internal reflection fluorescence (TIRF) spectrometric technique using a dual optical fiber sample cell was developed. A conventional silica optical fiber was used for exciting fluorescence compounds in its evanescent wave field. A liquid core waveguide (LCW) was used to collect the fluorescence photons emitted from fluorescence compounds existing in an excitation fiber’s evanescent wave field. The collected fluorescence photons were guided through the LCW and sent to a fluorescence spectrometer for detection. The spatial separation of excitation light and fluorescence light reduces the excitation-light-related optical noise signal, which is the major factor limiting fluorescence techniques from achieving lower detection limit. Preliminary results obtained from this work indicate that the optical fiber TIRF system of this work can detect 4.6*1018 mole rhodamine 6G (2.7*106 molecules) existing in the evanescent wave field of the excitation optical fiber. Index Headings: Total internal reflection fluorescence; TRIF; Evanescent wave excited fluorescence; Liquid core waveguide; Dual optical fiber sample cell; Rhodamine 6G; R6G.
INTRODUCTION Fluorescence spectrometry is a critical chemical– biochemical detection method and has found a broad spectrum of applications in biology, medicine, environmental science, and agriculture. In fluorescence spectrometry, optical noise signal is a major factor prohibiting the technique to achieve a lower detection limit. The major optical noise sources include scattering-caused optical noise signals (Rayleigh scattering, Raman scattering), diffuse reflected excitation light, and fluorescence photons emitted from sample matrix compounds. Usually both the fluorescence intensity of a targeted fluorescence compound and noise signal caused by scattering–diffuse reflection increase with the increase of excitation light intensity. Therefore, in fluorescence spectrometry, increasing excitation light intensity is needed, but not enough to achieve lower detection limit. Approaches must also be developed to reduce scattering reflection-caused optical noise intensity. Received 22 September 2014; accepted 5 November 2014. * Author to whom correspondence should be sent. E-mail: stao@ wtamu.edu. DOI: 10.1366/14-07744
608
Volume 69, Number 5, 2015
Fluorescence technologies developed for reducing excitation light-caused optical noise intensity include time-resolved fluorescence spectroscopy,1 confocal fluorescence microscopy,2–4 laser induced fluorescence of micro sample channel,5,6 and total internal reflection fluorescence (TIRF) spectroscopy.7,8 In time-resolved fluorescence spectroscopy, a pulsed light source is used to excite the fluorescence molecules in a sample.1 The fluorescence photons are collected and detected after the excitation pulse. The timely separation of excitation and fluorescence photon collection makes it possible to eliminate excitation light interference. This technique has significantly improved the detection capability of fluorescence spectroscopy. However, expensive equipment, such as a nanosecond pulsed laser, fast response photon detector with nanosecond resolution, and data acquisition systems, are needed in order to separate fluorescence excitation from fluorescence photon detection. Both confocal microscopy and microsample channel fluorescence spectroscopy reduce the excitation light interference through reducing the volume from which fluorescence photons are collected. In a confocal microscope, the photons (fluorescence photons, scattering-generated interference photons, and diffuse reflection-generated interference photons) from a volume of 1018 L range are collected.2 The number of interference photons is limited from such a small volume and can be eliminated by using a filter before sending the collected photons to a detector. Confocal microscopy and microsample channel fluorescence techniques have been reported achieving the ultimate limit of chemical detections: detecting a single fluorescence molecule. However, the confocal microscopy and microsample channel fluorescence spectroscopy need precise optics systems and are very difficult to operate. This is especially true if such systems have to be used outside of scientific research labs. Total internal reflection fluorescence (TRIF) is a different technique in which fluorescence molecules on the surface of an optical waveguide are excited by the evanescent wave of excitation light guided inside the optical waveguide. The excitation light and the fluorescence photons are spatially separated. Therefore, a lens used to collect the fluorescence photons will not pick up
0003-7028/15/6905-0608/0 Q 2015 Society for Applied Spectroscopy
APPLIED SPECTROSCOPY
FIG. 1. A diagrammatic graph of the optical fiber TIRF system of this work. The green lines show the excitation light that was guided through the excitation optical fiber, emerging at the end of the fiber that is outside of the LCW. The red arrows show the fluorescence photons emitted from the excitation optical fiber surface and guided within the LCW through total internal reflections.
scattered excitation light, which is the major noise signal when an intensive light source, such as a laser, is used for excitation. In TIRF, the diffuse reflection-caused noise is negligible as long as a high-quality waveguide (a prism or a microscope slide) is used as an excitation element. Using an optical filter can eliminate the limited Raman scattering noise. Due to its extremely low noise level, TIRF has also achieved single molecular detection (SMD). Planar waveguides and prisms have been used in TIRF as optical waveguides.7,8 In addition to SMD, TIRF can also be used to image the two-dimensional distribution of a limited number of fluorescence molecules absorbed on the waveguide’s surface if a highresolution optics system and a sensitive array detector are used to detect the fluorescence photon distribution.8 In TIRF, a lens has to be used to collect fluorescence photons from a large waveguide element. The noise originating from an instrument’s inner surface diffuse reflection is the factor limiting its detection capability. An optical fiber is a cylindrical optical waveguide. Light is guided inside an optical fiber through total internal reflection. If a fluorescence molecule exists on the surface of an optical fiber core, the molecule can be excited by the evanescent wave of light guided within the fiber core to emit fluorescence photons.9 Theoretically, this optical fiber TIRF phenomenon can be used for detecting the fluorescence compound. However, with traditional fluorescence detection methods, such as using a large lens to collect the optical fiber TIRF fluorescence photons, the lens will collect fluorescence photons from only a small area where the fiber is located, but also pick up noise photons from a large area where there is no optical fiber. In this paper, we present a special method for efficiently collecting the optical fiber TIRF photons. A liquid core waveguide (LCW) is a special optical fiber which has a liquid (water or an organic solvent) filled inside a capillary made from a special material as a fiber core for guiding light. The Teflon AF (DuPont) polymer has a refractive index between 1.29 and 1.31, which is lower than the refractive index of water (1.33) and most organic solvents. Therefore, a water-filled Teflon AF capillary can function as an optical fiber with water as the fiber core. Presently, a Teflon AF capillary filled with water or an organic solvent is the most-reported LCW for applications in optical spectrometry.10–14 In the reported works, the LCWs were used as long path-length sample cells for the purpose of increasing the sensitivity of optical spectrometric methods. However, when fluorescence molecules
in an LCW are excited with a transversely illuminating excitation method, the fluorescence signal-to-noise ratio does not have much improvement15 because the optical noise signal also increases with the increase of sample cell path-length. In this work, the LCW is used to collect fluorescence photons, which is different from the reported LCW fluorescence techniques.
EXPERIMENTAL The sample cell of our optical fiber TIRF is shown in Fig. 1. The sample cell consists of two optical fibers: a conventional silica optical fiber for exciting the fluorescence molecules and an LCW for collecting the fluorescence photons emitted from the silica optical fiber surface. The LCW used in this work is a 7 cm Teflon AF 2400 capillary (Random Technologies LLC, San Francisco, CA), which has an inner diameter of 800 lm. The Teflon AF 2400 polymer has a refractive index of 1.29 and is optically transparent in broad wavelength range from deep ultraviolet (UV) to near infrared. Deionized (DI) water is filled into the Teflon AF 2400 capillary to form an LCW. The optical fiber used for exciting the fluorescence compound is a silica optical fiber (AFS50/125Y, Thorlabs Inc.). This fiber has a 50 lm core diameter and 125 lm cladding diameter and has an acrylate polymer jacket (250 lm diameter). A 5 cm acrylate jacket of the silica optical fiber from one end of the fiber was removed by using a flame. The jacket-removed fiber was then inserted in 50% hydrofluoric acid for 20 min to etch off the silica cladding and part of the fiber core. The etched fiber was rinsed with DI water and was then inserted into the LCW (Teflon AF 2400 capillary). The diameter of the etched optical fiber was measured with an optical microscope (Micromaster, Fisher Scientific) by comparing the diameter of the etched silica fiber with the originally cladded fiber, which has a claimed cladding diameter of 125 lm. An optical microscopic imaging of the etched fiber and original jacket-removed fiber is shown in Fig. 2. The diameter of the etched fiber was measured to be 43 lm. The two ends of the LCW were connected to two barbed T-connectors as shown in Fig. 1. The etched fiber core was pushed through the T-connectors, and about 1 cm of the fiber core is outside of the T-connector in the far end. The excitation light will only be guided through the excitation silica fiber (shown as a green line in Fig. 1) by pushing the excitation optical fiber outside of the T-connector, and the LCW will collect only the evanescent wave excited
APPLIED SPECTROSCOPY
609
FIG. 2. An optical microscopic image of an etched optical fiber (left) and the original AFS50/125Y optical fiber (jacket removed). The diameter of the etched fiber was measured by comparing with the original fiber that has a cladding diameter equal to 125 lm.
fluorescence photons (shown as red arrows in Fig. 1). A second silica optical fiber (BHF48-600, Thorlabs Inc.), which has a 600 lm core diameter and 630 lm cladding diameter, was inserted into the T-connector as shown in Fig. 1. This silica optical fiber was inserted into a black tube in order to avoid coupling the excitation light from the excitation optical fiber end into the optical fiber. The distal end of the BHF48-600 optical fiber was connected to an optical fiber-compatible mini monochromator (MC1-03, Optometrics Corp.). The exit wavelength of the small monochromator was set to the peak fluorescence emission wavelength (556 nm) of rhodamine 6G (R6G). This mini monochromator was used as a filter to exclude excitation light leaked to the LCW. The exit port of the monochromator was connected with an optical fiber-compatible fluorescence spectrometer (QE Pro, OceanOptics Inc.) by using another piece of BHF48-600 optical fiber, which was also inserted inside a black tube in order to reduce the chance of the optical fiber picking up diffuse reflected photons in the working environment. A continuous wave 532 nm laser (1500 mW, 532 nm DPSS laser system, Opto Engine LLC) was used as an excitation light source. The light beam from the laser was focused into the excitation silica optical fiber by using a fiber port collimator–coupler (PAF-SMA-5-B. Thorlabs Inc.). Rhodamine 6G solutions were used in this work as samples. The R6G aqueous solutions were prepared by dissolving the fluorescence compound (rhodamine 6G, C28H31N2O3Cl, 95%, Sigma-Aldrich Co.) in DI water and appropriate diluting prepared solutions with DI water. During a test, the laser was operated at its maximum power output (1500 mW). A sample solution was injected into the LCW via the vertical port of a T-connector by using a syringe. During the test, the integrated light intensity from 554 to 558 nm was continuously monitored by using a data acquisition program (OceanView, OceanOptics Inc.) provided by the optical fiber compatible
610
Volume 69, Number 5, 2015
FIG. 3. Recorded spectra after injecting R6G sample solutions into the CLW. A low laser intensity at 532 nm was observed. The fluorescence intensity increased with the increase of R6G concentration. A strong blank fluorescence spectrum was also observed.
fluorescence spectrometer maker. An optical spectrum was also recorded after each sample was injected into the LCW.
RESULT AND DISCUSSION The excitation laser used in this work is a strong light source. The focus point of the lens in the optical fiber collimator–coupler used to inject laser beam into the optical fiber is larger than the core diameter (50 lm) of the excitation optical fiber. A small fraction of the excitation laser was guided through the cladding of the excitation optical fiber and leaked into the LCW, which was then guided through the LCW, coupled into the BHF48-600 optical fibers, and delivered to the fluorescence spectrometer. Although a mini monochromator was used to filter out the leaked excitation laser, it was observed that a very low intensity of excitation light was still detected by the optical fiber-compatible fluorescence spectrometer as shown in Fig. 3. It was believed that the 532 nm excitation light escaped the monochromator by diffuse reflection. This leaked excitation light intensity is very low and can be completely separated from the fluorescence light with the optical fibercompatible spectrometer as shown in recorded spectra in Fig. 3. It was observed that, without the use of the mini monochromator, the spectrometer detected a strong excitation light, which cannot have a baseline separation from the fluorescence spectrum and interferes with the detection of fluorescence photons. The recorded spectra and integrated light intensity from 554–558 nm (fluorescence wavelength) are showing in Figs. 3 and 4. The preliminary test results clearly demonstrate the feasibility of using LCW to collect optical fiber TIRF photons. A calibration curve indicating the relationship of fluorescence intensity with R6G concentration was established using the recorded integrated fluorescence intensity. The integrated fluorescence intensity has a linear relationship with R6G concentration as indicated by the following equation,
FIG. 4. Time response of the integrated fluorescence intensity (554 to 558 nm) during the test. The inserted graph is a calibration curve indicating the relationship of integrated fluorescence intensity with R6G concentration.
which is obtained by using integrated fluorescence intensity and R6G concentration data in Fig. 4. IFL ¼ 22:81CR6G ðnMÞ þ 1980:2
ð1Þ
The detection limit of the system built in this work for detecting R6G can be calculated from the slope of the linear calibration curve and the standard deviation of the blank sample signal, which is calculated to be 6.15 counts/s using data from Fig. 4. The calculated detection limit (3r) is 8.1*1010 mole/L. In considering the fact that the evanescent wave penetration depth is comparable to the wavelength of light guided within the waveguide, only the fluorescence molecules in the solution of a layer of 532 nm surrounding the 43 lm optical fiber core were excited to emit fluorescence photons. The volume of that very thin layer solution can be calculated using the measured excitation fiber core diameter (43 lm), evanescent wave penetration depth (532 nm), and the length of the excitation fiber core (4 cm). The calculated volume is 5.78*109 L. Therefore, the detection limit of the optical fiber TIRF system of this work expressed in number of moles is calculated to be 4.6*1018 mole (2.7*106 molecules). The detection limit of the optical fiber TIRF system of this work is not as good as those achieved with commercial TIRF microscopic equipment.7,8 Two factors limited the system of this work to achieve a lower detection limit. First, the laser power is still not high enough to reach a saturated excitation. The fluorescence intensity can be increased with the use of a stronger laser. Second, an intensive blank signal was observed at the R6G fluorescence wavelength in this work. The recorded spectra clearly indicates that an intensive fluorescence spectrum with an integrated intensity of 1980 counts/s was recorded when DI water was injected as a blank sample. The origin of the blank fluorescence signal is not experimentally examined. The source of this blank fluorescence signal could be from fluorescence compounds in the fiber jacket (acrylate polymer), the
plastic T-connector, or the epoxy glue used to glue the fibers.
CONCLUSION The optical fiber total internal reflection fluorescence technique developed from this work is more robust compared with the present single molecular detection fluorescence technologies such as TIRF microscopy and confocal microscopy. This makes the technology useful in field applications such as environmental investigation, food safety inspection, and point-of-care disease diagnosis. Our further work will be focused on (1) eliminating the blank signal in order to achieve a lower detection limit, and (2) integrating the optical fiber TIRF developed from this work with fluorescence sandwich immunoassay for detecting biospecies. ACKNOWLEDGMENT This work was supported by the U.S. Department of Agriculture (Award No. 2012-69003-19620). 1. F.G. Prendergast. ‘‘Time-Resolved Fluorescence Techniques: Methods and Applications in Biology’’. Curr. Opin. Struct. Biol. 1991. 1(6): 1054-1059. 2. W.E. Moerner, D.P. Fromm. ‘‘Methods of Single-Molecule Fluorescence Spectroscopy and Microscopy’’. Rev. Sci. Instrum. 2003. 74: 3597-3619. 3. S.M. Nie, R.N. Zare. ‘‘Optical Detection of Single Molecules’’. Ann. Rev. Biophys. Biomol. Struct. 1997. 26: 567-596. 4. X.H. Xu, E.S. Yeung. ‘‘Direct Measurement of Single-Molecule Diffusion and Photodecomposition in Free Solution’’. Science. 1997. 275(5303): 1106-1109. 5. T.H. Wang, Y.H. Peng, C.Y. Zhang, P.K. Wong, C.M. Ho. ‘‘SingleMolecule Tracing on a Fluidic Microchip for Quantitative Detection of Low-Abundance Nucleic Acids’’. J. Am. Chem. Soc. 2005. 127(15): 5354-5359. 6. J.M. Song, T. Inoue, H. Kawazumi, T. Ogawa. ‘‘Single Molecule Detection by Laser Two-Photon Excited Fluorescence in a Capillary Flow Cell’’. Anal. Sci. 1998. 14: 913-916. 7. W.P. Ambrose, P.M. Goodwin, J.P. Nolan. ‘‘Single-Molecule Detection with Total Internal Reflection Excitation: Comparing Signal-to-Background and Total Signals in Different Geometries’’. Cytometry. 1999. 36(3): 224-231.
APPLIED SPECTROSCOPY
611
8. M. Tokunaga, K. Kitamura, K. Saito, A. Hikikoshi, T. Yanagida. ‘‘Single Molecule Imaging of Fluorophores and Enzymatic Reactions Achieved by Objective-Type Total Internal Reflection Fluorescence Microscopy’’. Biochem. Biophys. Res. Commun. 1997. 235(1): 47-53. 9. X.H. Fang, W.H. Tan. ‘‘Imaging Single Fluorescent Molecules at the Interface of an Optical Fiber Probe by Evanescent Wave Excitation’’. Anal. Chem. 1999. 71(15): 3101-3105. 10. R. Altkorn, I. Koev, R.P. Van Duyne, M. Litorja. ‘‘Low-Loss LiquidCore Optical Fiber for Low-Refractive-Index Liquids: Fabrication, Characterization, and Application in Raman Spectroscopy’’. Appl. Opt. 1997. 36(34): 8992-8998. 11. R.H. Byrne, W.H. Yao, E. Kaltenbacher, R.D. Waterbury. ‘‘Construction of a Compact Spectrofluorometer/Spectrophotometer System Using a Flexible Liquid Core Waveguide’’. Talanta. 2000. 50(6): 1307-1312.
612
Volume 69, Number 5, 2015
12. T. Le, S. Tao. ‘‘Intrinsic UV Absorption Spectrometry Observed with a Liquid Core Waveguide as a Sensor Technique for Monitoring of Ozone in Water’’. Analyst. 2011. 136(16): 3335-3342. 13. S. Tao, S.F. Gong, L. Xu, J.C. Fanguy. ‘‘Mercury Atomic Absorption by Mercury Atoms in Water Observed with a Liquid Core Waveguide as a Long Path Absorption Cell’’. Analyst. 2004. 129(4): 342-346. 14. S. Tao, C.B. Winstead, H. Xian, S. Krunal. ‘‘A Highly Sensitive Hexachromium Monitor Using a Water Core Optical Fiber with UV LED’’. J. Environ. Monit. 2002. 4(5): 815-818. 15. Q. Li, K.J. Morris, P.K. Dasgupta, I.M. Raimundo, H. Temkin. ‘‘Portable Flow-Injection Analyzer with Liquid-Core Waveguide Based Fluorescence, Luminescence, and Long Path Length Absorbance Detector’’. Anal. Chim. Acta. 2003. 479(2): 151-165.
Total Internal Reflection Fluorescence Spectrometry Using a Dual Optical Fiber Sample Cell Cole A. Guyer, Shiquan Tao* Department of Mathematics, Chemistry and Physics, West Texas A&M University, WTAMU Box 60787, Canyon, TX 79016 USA
A novel total internal reflection fluorescence (TIRF) spectrometric technique using a dual optical fiber sample cell was developed. A conventional silica optical fiber was used for exciting fluorescence compounds in its evanescent wave field. A liquid core waveguide (LCW) was used to collect the fluorescence photons emitted from fluorescence compounds existing in an excitation fiber’s evanescent wave field. The collected fluorescence photons were guided through the LCW and sent to a fluorescence spectrometer for detection. The spatial separation of excitation light and fluorescence light reduces the excitation-light-related optical noise signal, which is the major factor limiting fluorescence techniques from achieving lower detection limit. Preliminary results obtained from this work indicate that the optical fiber TIRF system of this work can detect 4.6*1018 mole rhodamine 6G (2.7*106 molecules) existing in the evanescent wave field of the excitation optical fiber. Index Headings: Total internal reflection fluorescence; TRIF; Evanescent wave excited fluorescence; Liquid core waveguide; Dual optical fiber sample cell; Rhodamine 6G; R6G.
INTRODUCTION Fluorescence spectrometry is a critical chemical– biochemical detection method and has found a broad spectrum of applications in biology, medicine, environmental science, and agriculture. In fluorescence spectrometry, optical noise signal is a major factor prohibiting the technique to achieve a lower detection limit. The major optical noise sources include scattering-caused optical noise signals (Rayleigh scattering, Raman scattering), diffuse reflected excitation light, and fluorescence photons emitted from sample matrix compounds. Usually both the fluorescence intensity of a targeted fluorescence compound and noise signal caused by scattering–diffuse reflection increase with the increase of excitation light intensity. Therefore, in fluorescence spectrometry, increasing excitation light intensity is needed, but not enough to achieve lower detection limit. Approaches must also be developed to reduce scattering reflection-caused optical noise intensity. Received 22 September 2014; accepted 5 November 2014. * Author to whom correspondence should be sent. E-mail: stao@ wtamu.edu. DOI: 10.1366/14-07744
608
Volume 69, Number 5, 2015
Fluorescence technologies developed for reducing excitation light-caused optical noise intensity include time-resolved fluorescence spectroscopy,1 confocal fluorescence microscopy,2–4 laser induced fluorescence of micro sample channel,5,6 and total internal reflection fluorescence (TIRF) spectroscopy.7,8 In time-resolved fluorescence spectroscopy, a pulsed light source is used to excite the fluorescence molecules in a sample.1 The fluorescence photons are collected and detected after the excitation pulse. The timely separation of excitation and fluorescence photon collection makes it possible to eliminate excitation light interference. This technique has significantly improved the detection capability of fluorescence spectroscopy. However, expensive equipment, such as a nanosecond pulsed laser, fast response photon detector with nanosecond resolution, and data acquisition systems, are needed in order to separate fluorescence excitation from fluorescence photon detection. Both confocal microscopy and microsample channel fluorescence spectroscopy reduce the excitation light interference through reducing the volume from which fluorescence photons are collected. In a confocal microscope, the photons (fluorescence photons, scattering-generated interference photons, and diffuse reflection-generated interference photons) from a volume of 1018 L range are collected.2 The number of interference photons is limited from such a small volume and can be eliminated by using a filter before sending the collected photons to a detector. Confocal microscopy and microsample channel fluorescence techniques have been reported achieving the ultimate limit of chemical detections: detecting a single fluorescence molecule. However, the confocal microscopy and microsample channel fluorescence spectroscopy need precise optics systems and are very difficult to operate. This is especially true if such systems have to be used outside of scientific research labs. Total internal reflection fluorescence (TRIF) is a different technique in which fluorescence molecules on the surface of an optical waveguide are excited by the evanescent wave of excitation light guided inside the optical waveguide. The excitation light and the fluorescence photons are spatially separated. Therefore, a lens used to collect the fluorescence photons will not pick up
0003-7028/15/6905-0608/0 Q 2015 Society for Applied Spectroscopy
APPLIED SPECTROSCOPY
FIG. 1. A diagrammatic graph of the optical fiber TIRF system of this work. The green lines show the excitation light that was guided through the excitation optical fiber, emerging at the end of the fiber that is outside of the LCW. The red arrows show the fluorescence photons emitted from the excitation optical fiber surface and guided within the LCW through total internal reflections.
scattered excitation light, which is the major noise signal when an intensive light source, such as a laser, is used for excitation. In TIRF, the diffuse reflection-caused noise is negligible as long as a high-quality waveguide (a prism or a microscope slide) is used as an excitation element. Using an optical filter can eliminate the limited Raman scattering noise. Due to its extremely low noise level, TIRF has also achieved single molecular detection (SMD). Planar waveguides and prisms have been used in TIRF as optical waveguides.7,8 In addition to SMD, TIRF can also be used to image the two-dimensional distribution of a limited number of fluorescence molecules absorbed on the waveguide’s surface if a highresolution optics system and a sensitive array detector are used to detect the fluorescence photon distribution.8 In TIRF, a lens has to be used to collect fluorescence photons from a large waveguide element. The noise originating from an instrument’s inner surface diffuse reflection is the factor limiting its detection capability. An optical fiber is a cylindrical optical waveguide. Light is guided inside an optical fiber through total internal reflection. If a fluorescence molecule exists on the surface of an optical fiber core, the molecule can be excited by the evanescent wave of light guided within the fiber core to emit fluorescence photons.9 Theoretically, this optical fiber TIRF phenomenon can be used for detecting the fluorescence compound. However, with traditional fluorescence detection methods, such as using a large lens to collect the optical fiber TIRF fluorescence photons, the lens will collect fluorescence photons from only a small area where the fiber is located, but also pick up noise photons from a large area where there is no optical fiber. In this paper, we present a special method for efficiently collecting the optical fiber TIRF photons. A liquid core waveguide (LCW) is a special optical fiber which has a liquid (water or an organic solvent) filled inside a capillary made from a special material as a fiber core for guiding light. The Teflon AF (DuPont) polymer has a refractive index between 1.29 and 1.31, which is lower than the refractive index of water (1.33) and most organic solvents. Therefore, a water-filled Teflon AF capillary can function as an optical fiber with water as the fiber core. Presently, a Teflon AF capillary filled with water or an organic solvent is the most-reported LCW for applications in optical spectrometry.10–14 In the reported works, the LCWs were used as long path-length sample cells for the purpose of increasing the sensitivity of optical spectrometric methods. However, when fluorescence molecules
in an LCW are excited with a transversely illuminating excitation method, the fluorescence signal-to-noise ratio does not have much improvement15 because the optical noise signal also increases with the increase of sample cell path-length. In this work, the LCW is used to collect fluorescence photons, which is different from the reported LCW fluorescence techniques.
EXPERIMENTAL The sample cell of our optical fiber TIRF is shown in Fig. 1. The sample cell consists of two optical fibers: a conventional silica optical fiber for exciting the fluorescence molecules and an LCW for collecting the fluorescence photons emitted from the silica optical fiber surface. The LCW used in this work is a 7 cm Teflon AF 2400 capillary (Random Technologies LLC, San Francisco, CA), which has an inner diameter of 800 lm. The Teflon AF 2400 polymer has a refractive index of 1.29 and is optically transparent in broad wavelength range from deep ultraviolet (UV) to near infrared. Deionized (DI) water is filled into the Teflon AF 2400 capillary to form an LCW. The optical fiber used for exciting the fluorescence compound is a silica optical fiber (AFS50/125Y, Thorlabs Inc.). This fiber has a 50 lm core diameter and 125 lm cladding diameter and has an acrylate polymer jacket (250 lm diameter). A 5 cm acrylate jacket of the silica optical fiber from one end of the fiber was removed by using a flame. The jacket-removed fiber was then inserted in 50% hydrofluoric acid for 20 min to etch off the silica cladding and part of the fiber core. The etched fiber was rinsed with DI water and was then inserted into the LCW (Teflon AF 2400 capillary). The diameter of the etched optical fiber was measured with an optical microscope (Micromaster, Fisher Scientific) by comparing the diameter of the etched silica fiber with the originally cladded fiber, which has a claimed cladding diameter of 125 lm. An optical microscopic imaging of the etched fiber and original jacket-removed fiber is shown in Fig. 2. The diameter of the etched fiber was measured to be 43 lm. The two ends of the LCW were connected to two barbed T-connectors as shown in Fig. 1. The etched fiber core was pushed through the T-connectors, and about 1 cm of the fiber core is outside of the T-connector in the far end. The excitation light will only be guided through the excitation silica fiber (shown as a green line in Fig. 1) by pushing the excitation optical fiber outside of the T-connector, and the LCW will collect only the evanescent wave excited
APPLIED SPECTROSCOPY
609
FIG. 2. An optical microscopic image of an etched optical fiber (left) and the original AFS50/125Y optical fiber (jacket removed). The diameter of the etched fiber was measured by comparing with the original fiber that has a cladding diameter equal to 125 lm.
fluorescence photons (shown as red arrows in Fig. 1). A second silica optical fiber (BHF48-600, Thorlabs Inc.), which has a 600 lm core diameter and 630 lm cladding diameter, was inserted into the T-connector as shown in Fig. 1. This silica optical fiber was inserted into a black tube in order to avoid coupling the excitation light from the excitation optical fiber end into the optical fiber. The distal end of the BHF48-600 optical fiber was connected to an optical fiber-compatible mini monochromator (MC1-03, Optometrics Corp.). The exit wavelength of the small monochromator was set to the peak fluorescence emission wavelength (556 nm) of rhodamine 6G (R6G). This mini monochromator was used as a filter to exclude excitation light leaked to the LCW. The exit port of the monochromator was connected with an optical fiber-compatible fluorescence spectrometer (QE Pro, OceanOptics Inc.) by using another piece of BHF48-600 optical fiber, which was also inserted inside a black tube in order to reduce the chance of the optical fiber picking up diffuse reflected photons in the working environment. A continuous wave 532 nm laser (1500 mW, 532 nm DPSS laser system, Opto Engine LLC) was used as an excitation light source. The light beam from the laser was focused into the excitation silica optical fiber by using a fiber port collimator–coupler (PAF-SMA-5-B. Thorlabs Inc.). Rhodamine 6G solutions were used in this work as samples. The R6G aqueous solutions were prepared by dissolving the fluorescence compound (rhodamine 6G, C28H31N2O3Cl, 95%, Sigma-Aldrich Co.) in DI water and appropriate diluting prepared solutions with DI water. During a test, the laser was operated at its maximum power output (1500 mW). A sample solution was injected into the LCW via the vertical port of a T-connector by using a syringe. During the test, the integrated light intensity from 554 to 558 nm was continuously monitored by using a data acquisition program (OceanView, OceanOptics Inc.) provided by the optical fiber compatible
610
Volume 69, Number 5, 2015
FIG. 3. Recorded spectra after injecting R6G sample solutions into the CLW. A low laser intensity at 532 nm was observed. The fluorescence intensity increased with the increase of R6G concentration. A strong blank fluorescence spectrum was also observed.
fluorescence spectrometer maker. An optical spectrum was also recorded after each sample was injected into the LCW.
RESULT AND DISCUSSION The excitation laser used in this work is a strong light source. The focus point of the lens in the optical fiber collimator–coupler used to inject laser beam into the optical fiber is larger than the core diameter (50 lm) of the excitation optical fiber. A small fraction of the excitation laser was guided through the cladding of the excitation optical fiber and leaked into the LCW, which was then guided through the LCW, coupled into the BHF48-600 optical fibers, and delivered to the fluorescence spectrometer. Although a mini monochromator was used to filter out the leaked excitation laser, it was observed that a very low intensity of excitation light was still detected by the optical fiber-compatible fluorescence spectrometer as shown in Fig. 3. It was believed that the 532 nm excitation light escaped the monochromator by diffuse reflection. This leaked excitation light intensity is very low and can be completely separated from the fluorescence light with the optical fibercompatible spectrometer as shown in recorded spectra in Fig. 3. It was observed that, without the use of the mini monochromator, the spectrometer detected a strong excitation light, which cannot have a baseline separation from the fluorescence spectrum and interferes with the detection of fluorescence photons. The recorded spectra and integrated light intensity from 554–558 nm (fluorescence wavelength) are showing in Figs. 3 and 4. The preliminary test results clearly demonstrate the feasibility of using LCW to collect optical fiber TIRF photons. A calibration curve indicating the relationship of fluorescence intensity with R6G concentration was established using the recorded integrated fluorescence intensity. The integrated fluorescence intensity has a linear relationship with R6G concentration as indicated by the following equation,
FIG. 4. Time response of the integrated fluorescence intensity (554 to 558 nm) during the test. The inserted graph is a calibration curve indicating the relationship of integrated fluorescence intensity with R6G concentration.
which is obtained by using integrated fluorescence intensity and R6G concentration data in Fig. 4. IFL ¼ 22:81CR6G ðnMÞ þ 1980:2
ð1Þ
The detection limit of the system built in this work for detecting R6G can be calculated from the slope of the linear calibration curve and the standard deviation of the blank sample signal, which is calculated to be 6.15 counts/s using data from Fig. 4. The calculated detection limit (3r) is 8.1*1010 mole/L. In considering the fact that the evanescent wave penetration depth is comparable to the wavelength of light guided within the waveguide, only the fluorescence molecules in the solution of a layer of 532 nm surrounding the 43 lm optical fiber core were excited to emit fluorescence photons. The volume of that very thin layer solution can be calculated using the measured excitation fiber core diameter (43 lm), evanescent wave penetration depth (532 nm), and the length of the excitation fiber core (4 cm). The calculated volume is 5.78*109 L. Therefore, the detection limit of the optical fiber TIRF system of this work expressed in number of moles is calculated to be 4.6*1018 mole (2.7*106 molecules). The detection limit of the optical fiber TIRF system of this work is not as good as those achieved with commercial TIRF microscopic equipment.7,8 Two factors limited the system of this work to achieve a lower detection limit. First, the laser power is still not high enough to reach a saturated excitation. The fluorescence intensity can be increased with the use of a stronger laser. Second, an intensive blank signal was observed at the R6G fluorescence wavelength in this work. The recorded spectra clearly indicates that an intensive fluorescence spectrum with an integrated intensity of 1980 counts/s was recorded when DI water was injected as a blank sample. The origin of the blank fluorescence signal is not experimentally examined. The source of this blank fluorescence signal could be from fluorescence compounds in the fiber jacket (acrylate polymer), the
plastic T-connector, or the epoxy glue used to glue the fibers.
CONCLUSION The optical fiber total internal reflection fluorescence technique developed from this work is more robust compared with the present single molecular detection fluorescence technologies such as TIRF microscopy and confocal microscopy. This makes the technology useful in field applications such as environmental investigation, food safety inspection, and point-of-care disease diagnosis. Our further work will be focused on (1) eliminating the blank signal in order to achieve a lower detection limit, and (2) integrating the optical fiber TIRF developed from this work with fluorescence sandwich immunoassay for detecting biospecies. ACKNOWLEDGMENT This work was supported by the U.S. Department of Agriculture (Award No. 2012-69003-19620). 1. F.G. Prendergast. ‘‘Time-Resolved Fluorescence Techniques: Methods and Applications in Biology’’. Curr. Opin. Struct. Biol. 1991. 1(6): 1054-1059. 2. W.E. Moerner, D.P. Fromm. ‘‘Methods of Single-Molecule Fluorescence Spectroscopy and Microscopy’’. Rev. Sci. Instrum. 2003. 74: 3597-3619. 3. S.M. Nie, R.N. Zare. ‘‘Optical Detection of Single Molecules’’. Ann. Rev. Biophys. Biomol. Struct. 1997. 26: 567-596. 4. X.H. Xu, E.S. Yeung. ‘‘Direct Measurement of Single-Molecule Diffusion and Photodecomposition in Free Solution’’. Science. 1997. 275(5303): 1106-1109. 5. T.H. Wang, Y.H. Peng, C.Y. Zhang, P.K. Wong, C.M. Ho. ‘‘SingleMolecule Tracing on a Fluidic Microchip for Quantitative Detection of Low-Abundance Nucleic Acids’’. J. Am. Chem. Soc. 2005. 127(15): 5354-5359. 6. J.M. Song, T. Inoue, H. Kawazumi, T. Ogawa. ‘‘Single Molecule Detection by Laser Two-Photon Excited Fluorescence in a Capillary Flow Cell’’. Anal. Sci. 1998. 14: 913-916. 7. W.P. Ambrose, P.M. Goodwin, J.P. Nolan. ‘‘Single-Molecule Detection with Total Internal Reflection Excitation: Comparing Signal-to-Background and Total Signals in Different Geometries’’. Cytometry. 1999. 36(3): 224-231.
APPLIED SPECTROSCOPY
611
8. M. Tokunaga, K. Kitamura, K. Saito, A. Hikikoshi, T. Yanagida. ‘‘Single Molecule Imaging of Fluorophores and Enzymatic Reactions Achieved by Objective-Type Total Internal Reflection Fluorescence Microscopy’’. Biochem. Biophys. Res. Commun. 1997. 235(1): 47-53. 9. X.H. Fang, W.H. Tan. ‘‘Imaging Single Fluorescent Molecules at the Interface of an Optical Fiber Probe by Evanescent Wave Excitation’’. Anal. Chem. 1999. 71(15): 3101-3105. 10. R. Altkorn, I. Koev, R.P. Van Duyne, M. Litorja. ‘‘Low-Loss LiquidCore Optical Fiber for Low-Refractive-Index Liquids: Fabrication, Characterization, and Application in Raman Spectroscopy’’. Appl. Opt. 1997. 36(34): 8992-8998. 11. R.H. Byrne, W.H. Yao, E. Kaltenbacher, R.D. Waterbury. ‘‘Construction of a Compact Spectrofluorometer/Spectrophotometer System Using a Flexible Liquid Core Waveguide’’. Talanta. 2000. 50(6): 1307-1312.
612
Volume 69, Number 5, 2015
12. T. Le, S. Tao. ‘‘Intrinsic UV Absorption Spectrometry Observed with a Liquid Core Waveguide as a Sensor Technique for Monitoring of Ozone in Water’’. Analyst. 2011. 136(16): 3335-3342. 13. S. Tao, S.F. Gong, L. Xu, J.C. Fanguy. ‘‘Mercury Atomic Absorption by Mercury Atoms in Water Observed with a Liquid Core Waveguide as a Long Path Absorption Cell’’. Analyst. 2004. 129(4): 342-346. 14. S. Tao, C.B. Winstead, H. Xian, S. Krunal. ‘‘A Highly Sensitive Hexachromium Monitor Using a Water Core Optical Fiber with UV LED’’. J. Environ. Monit. 2002. 4(5): 815-818. 15. Q. Li, K.J. Morris, P.K. Dasgupta, I.M. Raimundo, H. Temkin. ‘‘Portable Flow-Injection Analyzer with Liquid-Core Waveguide Based Fluorescence, Luminescence, and Long Path Length Absorbance Detector’’. Anal. Chim. Acta. 2003. 479(2): 151-165.
