Fourier fluorescence spectrometer for excitation ... - OSA Publishing

Fourier fluorescence spectrometer for excitation emission matrix measurement Leilei Peng *, Joseph A. Gardecki, Brett E. Bouma and Guillermo J. Tearney Harvard Medical School and Wellman Center for Photomedicine, Massachusetts General Hospital, 55 Fruit Street, BAR 708, Massachusetts 02114 * [email protected]

Abstract: We demonstrate a fluorescence spectrometer that utilizes principles of Fourier transform spectroscopy to measure excitation emission matrices (EEM) rapidly and with high spectral resolution. For this EEM fluorometer, incoherent excitation light is first input into a differential-delay scanning Michelson interferometer. Light from the output port excites sample fluorescence. The fluorescence remitted from the sample is directed to a second Michelson interferometer, whose differential-delay scanning is synchronized with the first interferometer. The EEM is obtained by twodimensional Fourier analysis of the detected signal from the output port of the second interferometer. EEM results from the system are verified by comparing with results from a standard spectrometer. The system provides a wide spectral range, adjustable spectral resolution, and fast EEM acquisition speed, which allows EEM’s to be acquired in 40 seconds at a spectral resolution of 81-cm-1. ©2008 Optical Society of America OCIS codes: (300.6280) Spectroscopy, fluorescence and luminescence; (300.6300) Spectroscopy, Fourier transforms.

References and links 1. 2.

3. 4. 5. 6. 7.

8. 9.

10.

11.

M. B. Sinclair, D. M. Haaland, J. A. Timlin, and H. D. T. Jones, "Hyperspectral confocal microscope," Appl. Opt. 45, 6283-6291 (2006). V. L. Sutherland, J. A. Timlin, L. T. Nieman, J. F. Guzowski, M. K. Chawla, P. F. Worley, B. Roysame, B. L. McNaughtona, M. B. Sinclair, and C. A. Barnes, "Advanced imaging of multiple mRNAs in brain tissue using a custom hyperspectral imager and multivariate curve resolution," J. Neurosci. Methods 160, 144-148 (2007). T. Zimmermann, J. Rietdorf, and R. Pepperkok, "Spectral imaging and its applications in live cell microscopy," FEBS Letters 546, 87 (2003). M. Gouzman, N. Lifshitz, S. Luryi, O. Semyonov, D. Gavrilov, and V. Kuzminskiy, "Excitation–emission fluorimeter based on linear interference filters," Appl. Opt. 43, 3066-3072 (2004). S. J. Hart and R. D. JiJi, "Light emitting diode excitation emission matrix fluorescence spectroscopy," Analyst 127, 1693-1699 (2002). M. G. Müller, A. Wax, I. Georgakoudi, R. R. Dasari, and M. S. Feld, "A reflectance spectrofluorimeter for real-time spectral diagnosis of disease," Rev. Sci. Inst. 73, 3933-3937 (2002). A. R. Muroski, K. S. Booksh, and M. L. Myrick, "Single-Measurement Excitation/Emission Matrix Spectrofluorometer for Determination of Hydrocarbons in Ocean Water. 1. Instrumentation and Background Correction," Anal. Chem. 68, 3534-3538 (1996). C. D. Tran and R. J. Furlan, "Spectrofluorometer Based on Acousto-Optic Tunable Filters for Rapid Scanning and Multicomponent Sample Analyses," Anal. Chem. 65, 1675-1681 (1993). R. A. Zângaro, J. Landulfo Silveira, R. Manoharan, G. Zonios, I. Itzkan, R. R. Dasari, J. V. Dam, and M. S. Feld, "Rapid multiexcitation fluorescence spectroscopy system for in vivo tissue diagnosis," Appl. Opt. 35, 5211-5219 (1996). A. F. Zuluaga, U. Utzinger, A. Durkin, H. Fuchs, A. Gillenwater, R. Jacob, B. Kemp, J. Fan, and R. Richards-kortum, "Fluorescence Excitation Emission Matrices of Human Tissue: A System for in Vivo Measurement and Method of Data Analysis," Appll. Spectrosc. 53, 302-311 (1999). R. Heintzmann, K. A. Lidke, and T. M. Jovin, "Double-pass Fourier transform imaging spectroscopy," Opt. Exp. 12, 753-763 (2004).

#89327 - $15.00 USD

(C) 2008 OSA

Received 31 Mar 2008; revised 31 Mar 2008; accepted 9 Jun 2008; published 30 Jun 2008

7 July 2008 / Vol. 16, No. 14 / OPTICS EXPRESS 10493

12. 13. 14. 15. 16. 17. 18. 19.

J. T. Motz, D. Yelin, B. J. Vakoc, B. E. Bouma, and G. J. Tearney, "Spectral- and frequency-encoded fluorescence imaging," Opt. Lett. 30, 2760-2762 (2005). L. Peng, J. T. Motz, R. W. Redmond, B. E. Bouma, and G. J. Tearney, "Fourier transform emission lifetime spectrometer," Opt. Lett. 32, 421-423 (2007). L. Greengard and J.-Y. Lee, "Accelerating the Nonuniform Fast Fourier Transform," Siam Rev. 46, 443-454 (2004). J. A. Gardecki and M. Maroncelli, "Set of Secondary Emission Standards for Calibration of the Spectral Responsivity in Emission Spectroscopy," Appl. Spec. 52, 1179-1189 (1998). J. R. Lakowicz, Principles of fluorescence spectroscopy, 2nd ed. (Kluwer Academic/Plenum Publishers, New York, 1999). G. Genty, S. Coen, and J. M. Dudley, "Fiber supercontinuum sources (Invited)," J. Opt. Soc. Am. B 24, 1771-1785 (2007). G. J. Tearney, B. E. Bouma, and J. G. Fujimoto, "High-speed phase- and group-delay scanning with a grating-based phase control delay line," Opt. Lett. 22, 1811-1813 (1997). W. Y. Oh, S. H. Yun, G. J. Tearney, and B. E. Bouma, "115 kHz tuning repetition rate ultrahigh-speed wavelength-swept semiconductor laser," Opt. Lett. 30, 3159-3161 (2005).

Simultaneous detection of many markers in situ is a “holy grail” of molecular imaging/microscopy that would enable gene expression profiling and the study of pathways on a system level in living cells and organisms. The most commonly investigated method for multiplex microscopy is multispectral emission imaging, followed by spectral unmixing. Using this technique, researchers have demonstrated the capability to distinguish up to 5 distinct fluorophores with overlapping spectra in a single sample[1-3]. In theory, however, if both the excitation and the emission spectral properties, i.e. Excitation-Emission Matrices (EEM), were simultaneously measured, unmixing could occur in a two-dimensional space, and as a result, many more fluorophores could be distinguished. The most straightforward way of measuring an EEM is to use a white light source and two dispersive spectrometers, one for filtering the excitation and the other for filtering the emission. This method provides a broad spectral range and adjustable spectral resolution. However, this approach is time consuming, frequently taking minutes to acquire one EEM*. Other methods have been reported for faster EEM measurements[4-10]. Most of these EEM spectrometers replaced the emission spectrometer with an emission spectrogram imaging system[5-7, 9, 10], resulting in EEM acquisition speeds ranging from tens of seconds[7] to less than 200 ms[6] seconds. Increasing the speed of the excitation wavelength scan is the primary engineering task for fast EEM measurements. Approaches include fast variable excitation filters[4, 8] and tunable excitation sources, or dye lasers with fast-switching dye cells[6, 9]. While EEM’s have been rapidly obtained by these methods at speeds up to 5/s[6], in general these techniques have fixed and limited tuning ranges that depend on the tunable filter or light source. Additionally, all of these methods have relatively low excitation spectral resolutions, which range approximately from 5 to 10 nm, and are limited in the number of excitation wavelength samples. Another approach is to spatially distribute different excitation wavelengths across the sample and measure the EEM directly with a 2-D camera[5, 7]. This technique requires a homogeneous sample over the spatial extent of the dispersed excitation source and is therefore not well suited for imaging. In this paper, we report a versatile double Fourier transform spectrometer that is capable of rapidly obtaining EEM’s. Fourier transform EEM spectroscopy has several advantages including: 1) the spectral resolution is adjustable by changing interferometer scan lengths and there no fundamental limitation on the spectral resolution other than the signal to noise ratio; higher spectral resolutions will provide more data points and may provide higher accuracies when spectrally unmixing signals from multi-labeled samples; 2) all excitation wavelengths *

A state of the art commercial fluorescence spectrometer (FluoroMax 4, Jobin Yvon) can scan wavelengths at a maximal speed of 80 nm/sec, and, as a result, it would require 2 minutes to obtain a single 100×100 nm2 EEM. #89327 - $15.00 USD

(C) 2008 OSA



simultaneously illuminate the sample while all emission wavelengths are simultaneously detected, improving the signal to noise ratio and permitting more rapid acquisition of EEM spectra; 3) the system uses low-cost photomultipliers instead of high-end imaging systems. A formidable challenge in Fourier EEM measurements is implementing two high-speed synchronized and yet independently adjustable delay scans for excitation and emission Fourier transforms. One previously reported Fourier transform method used a single, double-passed interferometer to obtain both the excitation and emission spectra [11]. However, with this method only three angle-projections of the EEM, the emission projection, excitation projection and diagonal projection are measured. Information from these three angleprojections is not sufficient for full EEM reconstruction.

Fig. 1. Fourier fluorescence excitation emission spectrometer. The system consisted of a broadband light source (Xe lamp) and two Michelson interferometers: the excitation interferometer (M1, M2 and BS1), and the emission interferometer (M2, M3 and BS2). The HeNe laser and the photodetector DET1 monitored the motion of the excitation interferometer’s scanning mirror. The green laser and the photodetector DET2 monitored the position of the emission interferometer. The Photomultiplier PMT2 detected the modulated emission power. PMT1 and PMT3 detected the zero differential delay for the excitation and emission interferometers respectively. SP: short pass filter; LP: long pass filter; BS: beam splitter; BP: beam pickup; DM: dichroic mirror.

In the Fourier EEM spectrometer, one interferometer modulates the broadband illumination source with a series of wavelength-dependent frequencies[12, 13]. The other interferometer measures the autocorrelation of the fluorescence emission spectra. Figure 1 depicts a schematic of the instrument. The spectrometer’s light source was a xenon arc lamp (Oriel 75W). Collimated light from the lamp was delivered to the first Michelson interferometer (excitation interferometer) by a 1-mm diameter, 0.22 numerical aperture (NA), multimode fiber. The interferometer was built with a cube beamsplitter (BS1), a fixed mirror on one arm (M1), and a scanning mirror (M2) driven by a galvanometer (GSI Lumonics, HPLK Z focus module) over a maximum range of 1 mm and at a scan rate of 50 Hz. Illumination light from the output port of the first interferometer was directed towards the sample cuvette by a dichroic mirror DM1 (Semrock 555 nm). An objective lens (Nikon 10X, 0.45 NA) was used to focus the illumination light into the sample under investigation. Fluorescent emission was collected through the same objective. Emitted light was transmitted through DM1 and sent to the second interferometer (emission interferometer), which consisted of a cube beamsplitter (BS2), the fast scanning mirror (M2) and a slow scanning mirror M3, driven by a piezoelectric translation stage (Thorlabs P7TFD001/T). Light from the output port of the emission interferometer was detected by a photomultiplier PMT2 (Hamamtsu HC125-01). The emission interferometer was synchronized with the excitation interferometer through the shared fast scan mirror M2, which was double-sided. The slow

#89327 - $15.00 USD

(C) 2008 OSA



scanning mirror M3 controlled the optical path offset (-50 to +50 μm) between the first and the second interferometer. To monitor the motion of the scan mirror in the excitation interferometer, light from a single frequency HeNe laser (633 nm) was coupled into the interferometer through DM1, and co-propagated with the illumination light. HeNe interference signals were detected by a photodetector DET1 (Thorlabs PDA55). Similarly, a green diode laser (532 nm) was coupled into the emission interferometer through a dichroic mirror DM2 (Semrock 555 nm) to monitor the movement of the emission interferometer. The green laser interference signal was detected by the second photodetector DET2. In addition to monitoring the motion of the two scanning mirrors, the exact differential delay of each interferometer needed to be measured in order to accurately reconstruct EEM’s. Thus, two additional monitoring signals were detected by photomultipliers, PMT1 and PMT3. PMT1 detected light that was reflected by the right side of a beam-pickup (BP1, Thorlabs, 4% reflectivity). PMT1 signals were the interference fringes from the broadband illumination light passing through the excitation interferometer. These interference fringes were used to determine the time at which the differential delay of the excitation interferometer was zero. PMT3 detected a portion of the original illumination light, which was reflected by the left side of BP1, combined with green laser light by the dichroic mirror DM3 (Thorlabs magenta dichroic). Interference fringes detected by PMT3 were used to determine the time at which the differential delay of the emission interferometer was zero. The system therefore had one signal channel (PMT2) and a total of 4 monitoring channels (DET1, DET2, PMT1 and PMT3). All channels were digitized simultaneously at a constant acquisition rate with 12-bit dynamic range using two multi-channel DAQ cards (National Instruments 6115). In an EEM, the fluorescent signal is typically located in a region where the emission wavelength is larger than the excitation wavelength (λ2 >λ1), where as stray lights and scatter from the samples locate at λ2 =λ1. In theory if the system could operate without filters, its spectral range would only be limited by the light source and the PMT sensitivity, and there would be no restrictions on the minimal stokes shift. However, in reality, stay light and scatter can saturate the PMT and affect the signal. Thus, a series of filters (LP: long-pass filter, SP: short-pass filter, DM: dichroic mirror) were used in the system to ensure that the spectra of all five channels did not overlap. The spectrometer’s spectral range was set by these filters, the spectral range of the light source, and the sensitivity of PMT’s. The use of a dichroic mirror limits the minimum emission wavelength and the maximum excitation wavelength allowed by the instrument. As configured for these experiments, the device can simultaneously measure excitation wavelengths ranging from 425 to 550 nm and emission wavelengths ranging from 580 to 750 nm. The spectral range of the current system accommodates most popular fluorophores, especially GFP and its variants. The system was able to measure entire EEM’s in 40 seconds. We have previously demonstrated a Fourier transform excitation fluorometer that used a single interferometer to measure lifetimes and excitation spectra simultaneously[13]. The excitation interferometer of the EEM spectrometer functions the same way as the previous fluorometer. When the differential delay of the excitation interferometer (DD1) scans, the intensity of the illumination light after the excitation interferometer is modulated by interference. If the illumination is monochromatic with a total power P0 at a wavenumber of σ1, the illumination power on the sample is a function of DD1 which is given by PIllum ( DD1 ) ∝ P0 ⎡⎣1 + cos ( 2πσ 1 DD1 ) ⎤⎦ .

(1)

Radiative emission excited by the modulated illumination will carry the same modulation frequency. The emission spectrum as a function of DD1 is given by I Em ( DD1 , σ 2 ) ∝ EEM (σ 1 , σ 2 ) P0 ⎣⎡1 + cos ( 2πσ 1 DD1 ) ⎦⎤ , #89327 - $15.00 USD

(C) 2008 OSA

(2)



where σ2 is the wavenumber of the emission light, and EEM(σ ,σ2) is the steady state excitation emission matrix of the sample, which is a two-dimensional spectral density matrix of emission power. The emitted fluorescence light is sent into the emission interferometer. The total emission power after the two interferometers is 

PEm ( DD1 , DD2 ) ∝ ∫ EEM

(σ 1 ,σ 2 ) P0 ⎡⎣1 + cos ( 2πσ 1 DD1 ) ⎤⎦ {1 + cos ( 2πσ 2 DD2 )}dσ 2

,

(3)

where DD2 is the differential delay of the emission encoder. If the illumination light has a broadband spectrum S0(σ1), Equation (3) becomes PEm ( DD1 , DD2 ) ∝ ∫∫ EEM

(σ 1 , σ 2 ) S0 (σ 1 ) ⎡⎣1 + cos ( 2πσ 1 DD1 ) ⎤⎦ {1 + cos ( 2πσ 2 DD2 )}dσ 1dσ 2

.

(4)

Two-dimensional Fourier transform of Equation (4) yields I Em (σ 1 , σ 2 ) = δ (σ 1 ) δ (σ 2 ) ∫∫ EEM (σ 1 , σ 2 )d σ 1dσ 2 + δ (σ 2 ).∫ EEM (σ 1 , σ 2 ) S0 (σ 1 )dσ 2 +δ (σ 1 ) ∫ EEM (σ 1 , σ 2 ) S0 (σ 1 )dσ 1 + EEM (σ 1 , σ 2 ) S0 (σ 1 ) + c.c.

,

(5)

where c.c. refers to complex conjugates of all terms in Equation (5), and δ(σ) is defined by ⎧1, σ = 0 . ⎩ 0, σ ≠ 0

δ (σ ) = ⎨

(6)

Equation (5) has four terms: the first term is the total steady state emission power; the second term is the one-dimensional excitation spectrum SEx(σ1); the third term is the onedimensional emission spectrum SEm(σ2), and the last term is the excitation emission matrix. In the Fourier EEM spectrometer depicted in Figure 1, modulated emission power, described by Equation (5), is detected by PMT2. Two interferometers share the moving mirror, M2. The emission encoder has an additional moving mirror M3. If position offsets of M2 and M3 are x1 and x2 respectively, DD1 and DD2 are DD1 = 2 x1 , DD2 = 2 x1 + 2 x2 .

(7)

Thus, the Fourier transform of the PMT2 signal as a function of x1 and x2 is : F ⎣⎡ PEm ( x1 , x2 ) ⎦⎤ ∝ I Em (σ 1 , σ 1 + σ 2 ) .

(8)

As a result, the two-dimensional Fourier coefficients for the EEM can be obtained by Fourier transforming the interference signal detected at PMT2 at different mirror offsets. The EEM is then reconstructed by a two-dimensional Fourier transform. Traditional Fourier reconstruction requires data that has equal point spacing or at least precisely known spatial frequencies, which is difficult to achieve when using mechanical components at high scan speed. Fourier transform infrared spectroscopy often uses a single frequency source, usually a HeNe laser, to monitor the movement of scanning mirror, and samples data at zero crossings of HeNe fringes. This method cannot be applied to the Fourier EEM spectrometer for four reasons: 1) visible spectra require higher sampling rate and higher precision; 2) hardware triggering will filter out useful data points; 3) two-dimensional interpolation is very time consuming and unreliable when noise is high; 4) single wavelength light cannot measure the differential delay of the interferometer.

#89327 - $15.00 USD

(C) 2008 OSA



We addressed two-dimensional spatial frequency calibration by implementing both hardware and software solutions. In addition to single-wavelength laser interference signals (DET1 and DET2), broadband interference signals (PMT1 and PMT2) were also monitored in order to pinpoint the time at which the differential delays for each of two interferometers was zero. As a result, DD1 and DD2 were precisely determined throughout the measurements. Furthermore, the two interferometers shared the same fast moving mirror, M2. With an additional slow scanning mirror M3 in the emission interferometer, the spectrometer performed two-dimensional optical path length scans in a diagonal saw tooth pattern. Compared with using two independent scanners, this approach generated semi-synchronized two-dimensional scans with roughly straight and evenly-spaced scan lines. In software, single wavelength interference signals were used to calculate the relative position of the scanning mirror via Hilbert transform and phase unwrapping [13]. Interference fringes from PMT1 and PMT3 were compared between each scan, so that exact differential delays could be determined for each scan. Once DD1 and DD2 were calculated, raw data from the PMT2 channel were organized as PM(x1,x2), which then went through 2D non-uniform fast Fourier transform (NUFFT) [14] to reconstruct I(σ1,σ2), followed by an inverse coordinate transform to obtain EEM’s. We chose the fast 2D NUFFT algorithm [14] instead of 2D interpolation because, in our experience, the 2D NUFFT consumed much less computing time and was much less sensitive to noise.

(a)

(b)

(c)

Fig. 2. (a) Corrected emission and excitation spectra measured by the Fourier spectrometer (dots) and a conventional spectrometer (solid lines) . (b) Corrected EEM of 4dicyanomethylene-2-methyl-6-(p-dimethyl-aminostyryl)-4H-pyran (DCM), 2.5 μM in dimethyl sulfoxide (DMSO), measured by the Fourier EEM spectrometer The x-axis of the grayscale map represents excitation wavelengths. The y-axis represents emission wavelengths. Excitation light was cut-off by a 555 nm dichroic mirror. Emission light was filtered by a 568 long pass filter. The spectral resolution for this dataset was 81 cm-1. (c) Corrected EEM measured by a conventional fluorescent spectrometer (Jobin Yvon Fluoromax 3). The spectral resolution for this dataset was 5 nm. EEM intensities are displayed by a contour plot with a gray scale lookup table representing arbitrary units.

The performance of the Fourier EEM spectrometer was checked with standard fluorophore solutions. Raw data sets were taken at an M2 scan rate of 7.5 mm/s (150 µm at 50 Hz) and an M3 scan rate of 1.5 μm/s (60 µm at 0.025 Hz). A 150×60 μm2 (x1 × x2) twodimensional space was acquired. All signals were acquired at 1Ms/s acquisition rate. The data acquisition took 40 seconds. Raw data points that fell within a 60×60 μm2 (x1 × x2) box, centered at the differential delay zero for each interferometer, were processed using the NUFFT algorithm. After the coordinate transform, EEMs were reconstructed with an 81-cm-1 spectral resolution (equivalent to 2 nm at 500 nm, or 3.4 nm at 650 nm). Emission and excitation spectra were reconstructed simultaneously with EEM’s. The data set was corrected with excitation-emission spectral responses of the system, which is calculated based on a previously measured illumination spectrum S0(σ1), calibrated PMT gains and measured transmission curves of all optical parts. Finally, the data set was converted from wavenumber units to wavelength units by multiplying with a factor of 1/lamda2. Spectral resolutions of the

#89327 - $15.00 USD

(C) 2008 OSA



excitation and emission interferometers were verified by measuring line widths of the HeNe and green laser. Figure 2 shows excitation and emission spectra and EEM results from 2.5 µM 4dicyanomethylene-2-methyl-6-(p-dimethyl-aminostyryl)-4H-pyran (DCM) in dimethyl sulfoxide (DMSO), obtained with both the Fourier spectrometer and a conventional fluorescence spectrometer (Jobin Yvon FluoroMax 3). The Fourier EEM was obtained in 40 seconds and the Fluoromax in approximately 5 minutes. Excitation and emission resolutions for the Fourier and standard spectrometers were 81-cm-1 and 5 nm, respectively. The excitation and emission spectral response of the Fluoromax 3 were obtained with the use of secondary standards[15, 16]. Excitation and emission spectra measured by the Fourier EEM spectrometer agree with those measured with the conventional spectrometer (Fig. 2a). The EEM measured with the Fourier spectrometer (Fig. 2b) shows a good correspondence to that obtained with the conventional spectrometer (Fig. 2c) Figure 3(a) shows the result from 1.0 μM Rhodamine 6G in DMSO. In the gray scale contour map, Rhodamine 6G EEM’s peak intensity was located at an excitation wavelength of 535 nm. The EEM shows a steep emission cut-off which is a result of the long pass filter LP1 (568 nm).

(a)

(b)

Fig. 3. (a) EEM of Rhodamine 6G, 1.0 μM in DMSO; (b) EEM of a mixture of DCM (2.5 μM) and Rhodamine 6G (1.0 μM) in DMSO. The x-axis of the grayscale map represents excitation wavelengths. The y-axis represents emission wavelengths. Excitation light was cut-off by a 555 nm dichroic mirror. Emission light was filtered by a 568 long pass filter. EEM intensities are displayed by a contour plot with a gray scale lookup table representing arbitrary units.

Emission from pure fluorophore solutions, shown in Figure 2(a) and Figure 3(a), generally has the same spectral shape regardless of the excitation wavelength. EEM’s of such samples can be expressed as the product of the excitation and emission spectra EEM (σ 1 , σ 2 ) = S Ex (σ 1 ) • S Em (σ 2 ) ,

(9)

where the excitation (or emission) spectrum is measured with any suitable emission (or excitation) wavelength. In such cases, EEM’s do not provide additional information above and beyond the excitation and emission spectra. However, when the sample is a mixture of different fluorophores, Equation (9) does not hold. We measured the EEM of a mixture of Rhodamine 6G and DCM (1.0 μM and 2.5 μM respectively, in DMSO). Figure 3(b) shows the gray scale map of EEM obtained from the mixture. The EEM of the mixture has a strong peak at an excitation wavelength of 535 nm and an emission wavelength of 580 nm, which is the contribution from Rhodamine 6G. A second EEM peak around an excitation wavelength of 480 nm and an emission wavelength of 640 nm are mostly from DCM fluorescence emission. Emission intensities beyond 610 nm are seen in the shorter excitation range. The EEM clearly shows that the emission peak of the mixture shifts as the excitation wavelength changes. The EEM of the mixture demonstrates that the Fourier EEM spectrometer measures true EEMs, not just the product of excitation and emission spectra. In conclusion, we have demonstrated a Fourier spectrometer for fast fluorescence excitation emission matrix measurement, which has a broad spectral range, variable spectral resolution and relatively low equipment cost. The Fourier EEM spectrometer acquired EEMs #89327 - $15.00 USD

(C) 2008 OSA



in ~ 40 seconds with 82-cm-1 spectral resolutions in the 425~550 nm excitation range and the 580~750 emission range. The speed of current configuration is limited by the low power and brightness of the spatially incoherent illumination of the arc lamp light source, which requires a large aperture delay scanner and makes it difficult to obtain maximal fringe contrast. These limitations of our current light source require averaging of multiple spectra (~50 averages) in order to attain an acceptable signal to noise ratio. Spatially coherent, visible supercontinuum lasers[17], which are now commercially available, could overcome these limitations. With a broadband laser source and a rapidly scanning delay line[18, 19], the Fourier EEM spectrometer could be capable of measuring high resolution EEM’s in milliseconds. Acknowledgments This research was supported in part by the Department of Defense Medical Free Electron Laser Program FA9550-04-1-0079.

#89327 - $15.00 USD

(C) 2008 OSA

