- Email: [email protected]
Evaluation of instrument response functions for lifetime imaging detectors using quenched Rose Bengal solutions
Chemical Physics Letters 471 (2009) 153–159
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Evaluation of instrument response functions for lifetime imaging detectors using quenched Rose Bengal solutions Mariusz Szabelski a,b,*, Rafal Luchowski a, Zygmunt Gryczynski a,c, Peter Kapusta d, Uwe Ortmann d, Ignacy Gryczynski a,c a
Center for Commercialization of Fluorescence Technologies, Department of Molecular Biology and Immunology, UNT HSC, Fort Worth, TX, USA Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80-952 Gdansk, Poland c Department of Cell Biology and Genetics, UNT HSC, Fort Worth, TX, USA d PicoQuant GmbH, Berlin, Germany b
a r t i c l e
i n f o
Article history: Received 12 December 2008 In final form 2 February 2009 Available online 5 February 2009
a b s t r a c t Instrument response functions (IRF) in time-domain fluorescence are usually recorded as reflected or scattered excitation light, which is at shorter wavelengths than the observed fluorescence emission. However, its often more appropriate to measure the IRF in the emission spectral region. In this work we show that Rose Bengal water solutions quenched by potassium iodide can be used to measure instrument response functions of single photon detectors in the orange-red wavelength region. We used the quenched RB emission as a reference in time-domain measurements with common detectors and we got practically the same results as with scattering. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction Recording the correct instrument response function (IRF) in order to obtain correct fluorescence lifetimes is an important task in time-resolved fluorescence spectroscopy. IRF is usually recorded as reflected or scattered excitation light, which has a shorter wavelength than the observed fluorescence emission. Sometimes the excitation and emission wavelengths differ significantly and the timing characteristics of photon detectors (especially that of APDs) depend on the wavelength. This problem can be solved by using known standard dyes as a reference, preferably with very short lifetimes. Water solution of erythrosine B was used to obtain the IRF corresponding to the laser pulse convoluted with the detection response in the time-resolved flavin fluorescence experiments [1,2]. Hanley et al. described a multi-point method for calibrating a frequency domain FLIM system. For this purpose they used Rhodamine 6G solutions quenched with iodide, exhibiting single exponential decays [3]. In one of the recent work we showed that erythrosine B can be quenched by KI to about 25 ps and serves as a reference for a green region of spectrum (530–570 nm) [4]. It should be noted that the idea of using a reference dye to avoid col-
Abbreviations: QY, quantum yield; RB, Rose Bengal; KI, potassium iodide; IRF, instrument response function; RhB, Rhodamin B; Rh6G, Rhodamin 6G; APD, avalanche photo diode; MCP, microchannel plate; PMT, photomultiplier tube; FWHM, full-width at half of maximum. * Corresponding author. Address: Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80-952 Gdansk, Poland. Fax: +48 58 341 31 75. E-mail address: fi[email protected] (M. Szabelski). 0009-2614/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2009.02.001
or effects is older. Already Harris and Lytle [5], Zuker et al. [6] and Van Den Zegel [7] were well aware of this problem. The xanthene dyes are very well-known, and because of their photochemical properties often used fluorophores. Rose Bengal (RB) is a one of them. Its chemical structure is similar to that of fluorescein. Both dyes have the same carbon aromatic skeleton but differ in aromatic substitutions. Because of chlorines and iodines on the xanthene ring, RB features a heavy-atom effect which results in a high efficiency of intersystem crossing to the triplet state [8]. RB is a water-soluble photosensitizer with a high absorption coefficient in the red region of the spectrum and a affinity to transfer electrons from its excited triplet state, producing long-lived radicals [9,10]. As a photosensitizer, RB was used to inactivate microorganisms such as viruses [11,12], gram-positive bacterial species [13] and protozoa [14]. The liposomal encapsulation of RB was studied to improve the applicability of the drug in clinical application [15]. Rose Bengal stains damaged conjuctival and corneal cells. It is used as eye drops to identify damage to the surface of the eye [16]. In new technologies dye-sensitized solar cell is fabricated using Rose Bengal dye for sensitization of nanocrystalline TiO2 and that imparts extension in spectral response towards visible region by modifying the semiconductor surface [17]. In this work we present an ultra-short fluorescence standard in the orange-red region (540–610 nm) which has a lifetime of 16 ps. We demonstrate that such a quenched RB fluorescence can be used as an instrument response function in reconvolution fitting of fluorescence intensity decays. We also evaluated the color dependence of APD responses using a quenched RB reference.
154
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
2. Materials and methods
Intensity [a.u.]
All chemicals were analytical reagents or the best grade commercially available. All solutions of chemicals were prepared in deionized water purified by using a Milli-Q Synthesis A10 system produced by Millipore. Rose Bengal, KI, KOH, Rhodamine B and Rhodamine 6G, Ludox (30 wt.% suspension in H2O) were purchased from Sigma–Aldrich. Rose Bengal solutions for spectroscopic measurements were prepared by adding 100 ll stock solution of RB in water to 2000 ll KI in water. KI solutions were prepared over concentration range from 0 M to saturate (6.03 M). The pH of all solutions was 9.8 and it was achieved by addition 30 ll of stock solution of KOH (0.004 M). Final volume of each sample was 2130 ll. Emission spectra and steady-state fluorescence anisotropies were obtained
220 200 180 160 140 120 100 80 60 40 20 0
with 1 cm quartz cuvettes using a Varian Cary Eclipse spectrofluorimeter. For lifetime measurements samples were excited at 470 nm and emission collected at 572 nm. Intensity decays were collected by time-domain technique using a FluoTime 200 lifetime spectrometer (PicoQuant GmbH) equipped with R3809U-50 microchannel plate photomultiplier (MCP-PMT, Hamamatsu) and PicoHarp300 TCSPC module. The excitation source was an LDH470 pulsed laser diode (470 nm, optical pulse duration: 65 ps FWHM) driven by a PDL800-B driver. Polarizers were set to magic angle conditions and the fluorescence was observed through a 100 mm focal length single grating emission monochromator (ScienceTech). Decays were analyzed using the FLUOFIT software package (version 4.2.1). The analysis involved iterative reconvolution fitting of a sum of exponentials to the experimentally recorded decays:
0 M KI
5.66 M KI
550
600
650
700
750
Wavelength [nm]
b F0/F and τ0/τ
8
F0/F
-1
7
Ksv= 1.08 M
6
kq= 1.4 x 10 M s
10
-1 -1
2
R = 0.9987
5
τ0/τ
4 3
-1
Ksv= 0.67 M
2
9
-1 -1
kq= 8 x 10 M s
1
2
R = 0.9990
0 0
1
2
3
4
5
6
4
5
6
KI [M]
c
0.37 0.36 0.35
Anisotropy
0.34 0.33 0.32 0.31 0.30 0.29 0.28 0
1
2
3
KI [M] Fig. 1. Fluorescence steady-state measurements for Rose Bengal. (a) Quenching of RB in water by KI. (b) Stern–Volmer dependence in the entire range of quencher concentration. Fluorescence quenching results in seven-fold decrease in the fluorescence intensity in the presence of 5.66 M KI. (c) Fluorescence anisotropy of Rose Bengal in water in presence of KI. The increase in the steady-state anisotropy is due to shorter lifetimes.
155
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159 Table 1 Fluorescence lifetime, and quality of fit (v2) for Rose Bengal in water and different concentration of KI.
where ai and si are the pre-exponential factor and fluorescence lifetime, respectively.
Concentration of KI (mol/dm3)
Lifetime (ps)
v2
3. Results and discussion
0 1.13 2.26 5.66
77 ± 3 48 ± 2 31 ± 1 16 ± 1
1.024 1.147 1.221 1.125
3.1. Steady-state fluorescence of RB in absence and presence of KI
IðtÞ ¼
X
ai expðt=si Þ;
ð1Þ
i
We measured fluorescence spectra of RB in absence and presence of KI (Fig. 1a). The intensity of the RB progressively decreases but the shape of the spectrum does not change. Also, the absorption spectra of RB do not change in the presence of KI (not shown). We believe that KI acts as an efficient collisional quencher of RB and does not react with the RB chromophore. This conclusion is
15000
Scatterer λobs= 470 nm
Intensity
10000
FWHM = 111 ps
111 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
3.7
3.8
3.7
3.8
15000
Rose Bengal in water λobs= 572 nm
Intensity
10000
FWHM= 188 ps
188 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
15000
Rose Bengal in 5.66M KI λobs= 572 nm
Intensity
10000
FWHM= 115 ps
115 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
3.7
3.8
Fig. 2. Time-domain results in linear scale collected using 470 nm pulsed laser diode excitation and Hamamatsu R3809U-50 microchannel plate PMT as a detector. These TCSPC histograms were recorded for a scatterer and Rose Bengal in absence and presence of 5.66 M KI. The measured FWHMs are: (a) 111 ps for scatterer; (b) 188 ps for Rose Bengal in absence of quencher and (c) 115 ps Rose Bengal in presence of 5.66 M KI.
156
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
Fig. 3. Intensity decay of the Rhodamin B–Rhodamin 6G mixture in water. The IRF was measured with a scatterer (a). Fractional intensities of the positive decay components (a, top) are 40% for Rhodamin B and 60% for Rhodamin 6G. Graphs (b) and (c) are intensity decays of pure RhB and Rh6G solutions, respectively,
Fig. 4. Intensity decay of Rhodamin B–Rhodamin 6G mixture in water. Quenched RB fluorescence is used as an IRF (a). Fractional intensities of the positive decay components (a, top) are 40% for Rhodamin B and 60% for Rhodamin 6G. Graphs (b) and (c) are intensity decays for RhB and Rh6G solutions, respectively.
157
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
based on the following experimental observations: The Stern– Volmer plot (Fig. 1b) is clearly linear. The quantum yield of the quenched RB is relatively high 0.003 and the brightness of the sample is reasonable and still two orders of magnitude stronger than the Raman scattering. The fluorescence anisotropy (Fig. 1c) increases with KI concentration in accordance with the expected decrease of fluorescence lifetime. The shorter lifetime means less time available for effective fluorescence depolarization by molecular rotation. 3.2. Time-resolved fluorescence of RB in absence and presence of KI Next, we measured the fluorescence intensity decays of RB in absence and presence of KI. The results are presented in Table 1. The errors presented in Table 1 are standard deviation and corresponding to an assumption of 68% confidence [18]. In the presence of 5.66 M KI the lifetime is only 16 ps, which is close to the decay resolution limit of the FT200. The responses of RB fluorescence in absence and presence of KI, as well as the response from colloidal silica (Ludox) scatter are presented in Fig. 2. In order to visualize details, we presented these responses in linear scale. It is clear that
a
the response for the quenched RB (Fig. 2c) at 572 nm observation and the IRF at the 470 excitation (Fig. 2a) are almost identical. 3.3. Collisional quenching of RB by KI We used both area under emission spectra and the lifetimes of progressively quenched Rose Bengal (Table 1) to construct a dynamic version of Stern–Volmer dependence (Fig. 1b). The plot s0/ s versus KI concentration reveal the Stern–Volmer quenching constant of 0.67 M1 and bimolecular quenching constant of about 8 109 M1 s1, taking into account that unquenched RB has the lifetime s0 = 77 ps. However, when fluorescence intensities were used to construct Stern–Volmer dependence, higher values of quenching constants were obtained (Ksv = 1.08 M1, 1 1 kq = 1.4 M s ). We believe this because of a static-quenching at high KI concentrations. 3.4. Use of quenched RB fluorescence as IRF in lifetime measurements First we measured lifetimes of two fluorophores, RhB and Rh6G (about 1 lM each) using scattering as IRF (Fig. 3b and c). These
14000 12000
Scatterer λobs= 470 nm
10000
Intensity
FWHM = 579 ps 8000
579 ps
6000 4000 2000 0 1.0
1.5
2.0
2.5
3.0
3.5
Time [ns]
b
14000 12000
Rose Bengal in 5.66M KI λobs= 572 nm
10000
FWHM = 471 ps
Intensity
8000
471 ps
6000 4000 2000 0 1.0
1.5
2.0
2.5
3.0
3.5
Time [ns] Fig. 5. Impulse response function in linear scale obtained with a scatterer (a) and quenched Rose Bengal (b) for Perkin Elmer APD (SPCM-AQR-14). FWHMs are showed on the graphs.
158
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
compounds show single exponential decays with lifetimes of 1.57 ns (RhB) and 3.90 ns (Rh6G). Than we prepared a mixture of these fluorophores and measured the decay kinetic of it (Fig. 3a). This intensity decay is heterogeneous and can be described with a double exponential model. Recovered lifetimes correspond well to the individual fluorophores. We analyzed these decays using two IRFs. One is obtained by scattering (Fig. 3), the other one is a decay of quenched RB at 572 nm (Fig. 4). Practically identical parameters were recovered (compare Figs. 3 and 4).
gle photon sensitive APDs using the MicroTime200 microscope (PicoQuant GmbH). For both detectors, the IRF FWHM recorded using scattered 470 nm excitation is found to be broader (by 20%) than the decay of quenched RB at 572 nm. However, due to the fast response of MPD APD this broadening is less than 20 ps, whereas for Perkin–Elmer APD it amounts to more than 100 ps under the same conditions.
3.5. Evaluation of IRFs of APD detectors
We demonstrated that using KI quencher, the RB fluorescence emission can be shortened to 16 ps. The fluorescence signal from quenched RB solution is more than two orders of magnitude stronger than Raman scattering and can be easily measured with a commercial fluorometer. This signal can be effectively used as IRF at the emission wavelength in decay curve analysis including reconvolution fitting of more complex decay models. We also showed
During the experiments described above we used an MCP-PMT with negligible color dependence of the timing performance. The temporal response of APD detectors, commonly used today in time-resolved microscopy, are more sensitive to the observation wavelength. We tested Perkin–Elmer (Fig. 5) and MPD (Fig. 6) sin-
a
4. Conclusions
12000
10000
Scatterer λobs= 470 nm
Intensity
8000
FWHM = 120 ps 6000
120 ps
4000
2000
0 1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Time [ns]
b
12000
10000
Rose Bengal in 5.66M KI λobs= 572 nm
Intensity
8000
FWHM= 102 ps 6000
102 ps
4000
2000
0 1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Time [ns] Fig. 6. Impulse response function in linear scale obtained with a scatterer (a) and quenched Rose Bengal (b) for MPD APD (PDM 1CTC). FWHMs are showed on the graphs.
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
that ultra-short fluorescence pulses from quenched RB solutions can be used to evaluate responses from electronic detectors such as APDs.
[4] [5] [6] [7]
Acknowledgements
[8] [9]
This work was supported by Texas Emerging Technologies Fund Grant (CCFT). Mariusz Szabelski and Rafal Luchowski are the recipients of a researcher’s mobility program from the Polish Ministry of Science and Higher Education. References
[10] [11] [12] [13] [14] [15]
[1] P.I.H. Bastiaens, A. van Hoek, W.F. Wolkers, J.-C. Brochon, A.J.W.G. Visser, Biochemistry 31 (1992) 7050. [2] P.A.W. van Den Berg, S.B. Mulrooney, B. Gobets, I.H.M. van Stokkum, A. van Hoek, C.H. Williams Jr., A.J.W.G. Visser, Protein Sci. 10 (2001) 2037. [3] Q.S. Hanley, V. Subramaniam, D.J. Arndt-Jovin, T.M. Jovin, Cytometry 43 (2001) 248.
[16] [17] [18]
159
M. Szabelski et al., Appl. Spectrosc. 63 (2009) 363. J.M. Harris, F.E. Lytle, Rev. Sci. Instrum. 48 (1977) 1469. M. Zuker, A.G. Szabo, L. Bramall, D.T. Krajcarski, Rev. Sci. Instrum. 56 (1985) 14. M. Van Den Zegel, N. Boens, D. Daems, F.C. De Schryver, Chem. Phys. 101 (1986) 311. D.C. Neckers, O.M. Valdes-Aguilera, Adv. Photochem. 18 (1993) 315. C. Giulivi, M. Sarcansky, E. Rosenfeld, A. Boveris, Photochem. Photobiol. 52 (1990) 745. I.E. Kochevar, C.R. Lambert, M.C. Lynch, A.C. Tedesco, Biochim. Biophys. Acta 1280 (2) (1996) 223. J. Lenard, A. Rabson, R. Vanderoef, Proc. Natl. Acad. Sci. USA 90 (1993) 158. J. Chodosh, M.C. Banks, W.G. Stroop, Invest. Ophthalmol. Vis. Sci. 33 (1992) 2520. T.A. Dahl, W.R. Midden, D.C. Neckers, Photochem. Photobiol. 48 (1988) 607. F.S. Cruz, L.A.V. Lopes, W. De Souza, S.N.J. Moreno, R.0. Mason, R. De Campo, Acta Trop. 41 (1984) 99. C.-C. Chang, Y.-T. Yang, J.-C. Yang, H.-D. Wu, T. Tsai, Dyes Pigments 79 (2008) 170. A.J. Bron, Surv. Ophthalmol. 45 (2001) S221. M.S. Roy, P. Balraju, M. Kumar, G.D. Sharma, Sol. Energy Mater. Sol. Cells 92 (2008) 909. J.R. Lakowicz, Principles of Fluorescence Spectroscopy, third edn., Kluwer Academic/Plenum Publishers, 2006.
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Evaluation of instrument response functions for lifetime imaging detectors using quenched Rose Bengal solutions Mariusz Szabelski a,b,*, Rafal Luchowski a, Zygmunt Gryczynski a,c, Peter Kapusta d, Uwe Ortmann d, Ignacy Gryczynski a,c a
Center for Commercialization of Fluorescence Technologies, Department of Molecular Biology and Immunology, UNT HSC, Fort Worth, TX, USA Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80-952 Gdansk, Poland c Department of Cell Biology and Genetics, UNT HSC, Fort Worth, TX, USA d PicoQuant GmbH, Berlin, Germany b
a r t i c l e
i n f o
Article history: Received 12 December 2008 In final form 2 February 2009 Available online 5 February 2009
a b s t r a c t Instrument response functions (IRF) in time-domain fluorescence are usually recorded as reflected or scattered excitation light, which is at shorter wavelengths than the observed fluorescence emission. However, its often more appropriate to measure the IRF in the emission spectral region. In this work we show that Rose Bengal water solutions quenched by potassium iodide can be used to measure instrument response functions of single photon detectors in the orange-red wavelength region. We used the quenched RB emission as a reference in time-domain measurements with common detectors and we got practically the same results as with scattering. Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction Recording the correct instrument response function (IRF) in order to obtain correct fluorescence lifetimes is an important task in time-resolved fluorescence spectroscopy. IRF is usually recorded as reflected or scattered excitation light, which has a shorter wavelength than the observed fluorescence emission. Sometimes the excitation and emission wavelengths differ significantly and the timing characteristics of photon detectors (especially that of APDs) depend on the wavelength. This problem can be solved by using known standard dyes as a reference, preferably with very short lifetimes. Water solution of erythrosine B was used to obtain the IRF corresponding to the laser pulse convoluted with the detection response in the time-resolved flavin fluorescence experiments [1,2]. Hanley et al. described a multi-point method for calibrating a frequency domain FLIM system. For this purpose they used Rhodamine 6G solutions quenched with iodide, exhibiting single exponential decays [3]. In one of the recent work we showed that erythrosine B can be quenched by KI to about 25 ps and serves as a reference for a green region of spectrum (530–570 nm) [4]. It should be noted that the idea of using a reference dye to avoid col-
Abbreviations: QY, quantum yield; RB, Rose Bengal; KI, potassium iodide; IRF, instrument response function; RhB, Rhodamin B; Rh6G, Rhodamin 6G; APD, avalanche photo diode; MCP, microchannel plate; PMT, photomultiplier tube; FWHM, full-width at half of maximum. * Corresponding author. Address: Institute of Experimental Physics, University of Gdansk, Wita Stwosza 57, 80-952 Gdansk, Poland. Fax: +48 58 341 31 75. E-mail address: fi[email protected] (M. Szabelski). 0009-2614/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.cplett.2009.02.001
or effects is older. Already Harris and Lytle [5], Zuker et al. [6] and Van Den Zegel [7] were well aware of this problem. The xanthene dyes are very well-known, and because of their photochemical properties often used fluorophores. Rose Bengal (RB) is a one of them. Its chemical structure is similar to that of fluorescein. Both dyes have the same carbon aromatic skeleton but differ in aromatic substitutions. Because of chlorines and iodines on the xanthene ring, RB features a heavy-atom effect which results in a high efficiency of intersystem crossing to the triplet state [8]. RB is a water-soluble photosensitizer with a high absorption coefficient in the red region of the spectrum and a affinity to transfer electrons from its excited triplet state, producing long-lived radicals [9,10]. As a photosensitizer, RB was used to inactivate microorganisms such as viruses [11,12], gram-positive bacterial species [13] and protozoa [14]. The liposomal encapsulation of RB was studied to improve the applicability of the drug in clinical application [15]. Rose Bengal stains damaged conjuctival and corneal cells. It is used as eye drops to identify damage to the surface of the eye [16]. In new technologies dye-sensitized solar cell is fabricated using Rose Bengal dye for sensitization of nanocrystalline TiO2 and that imparts extension in spectral response towards visible region by modifying the semiconductor surface [17]. In this work we present an ultra-short fluorescence standard in the orange-red region (540–610 nm) which has a lifetime of 16 ps. We demonstrate that such a quenched RB fluorescence can be used as an instrument response function in reconvolution fitting of fluorescence intensity decays. We also evaluated the color dependence of APD responses using a quenched RB reference.
154
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
2. Materials and methods
Intensity [a.u.]
All chemicals were analytical reagents or the best grade commercially available. All solutions of chemicals were prepared in deionized water purified by using a Milli-Q Synthesis A10 system produced by Millipore. Rose Bengal, KI, KOH, Rhodamine B and Rhodamine 6G, Ludox (30 wt.% suspension in H2O) were purchased from Sigma–Aldrich. Rose Bengal solutions for spectroscopic measurements were prepared by adding 100 ll stock solution of RB in water to 2000 ll KI in water. KI solutions were prepared over concentration range from 0 M to saturate (6.03 M). The pH of all solutions was 9.8 and it was achieved by addition 30 ll of stock solution of KOH (0.004 M). Final volume of each sample was 2130 ll. Emission spectra and steady-state fluorescence anisotropies were obtained
220 200 180 160 140 120 100 80 60 40 20 0
with 1 cm quartz cuvettes using a Varian Cary Eclipse spectrofluorimeter. For lifetime measurements samples were excited at 470 nm and emission collected at 572 nm. Intensity decays were collected by time-domain technique using a FluoTime 200 lifetime spectrometer (PicoQuant GmbH) equipped with R3809U-50 microchannel plate photomultiplier (MCP-PMT, Hamamatsu) and PicoHarp300 TCSPC module. The excitation source was an LDH470 pulsed laser diode (470 nm, optical pulse duration: 65 ps FWHM) driven by a PDL800-B driver. Polarizers were set to magic angle conditions and the fluorescence was observed through a 100 mm focal length single grating emission monochromator (ScienceTech). Decays were analyzed using the FLUOFIT software package (version 4.2.1). The analysis involved iterative reconvolution fitting of a sum of exponentials to the experimentally recorded decays:
0 M KI
5.66 M KI
550
600
650
700
750
Wavelength [nm]
b F0/F and τ0/τ
8
F0/F
-1
7
Ksv= 1.08 M
6
kq= 1.4 x 10 M s
10
-1 -1
2
R = 0.9987
5
τ0/τ
4 3
-1
Ksv= 0.67 M
2
9
-1 -1
kq= 8 x 10 M s
1
2
R = 0.9990
0 0
1
2
3
4
5
6
4
5
6
KI [M]
c
0.37 0.36 0.35
Anisotropy
0.34 0.33 0.32 0.31 0.30 0.29 0.28 0
1
2
3
KI [M] Fig. 1. Fluorescence steady-state measurements for Rose Bengal. (a) Quenching of RB in water by KI. (b) Stern–Volmer dependence in the entire range of quencher concentration. Fluorescence quenching results in seven-fold decrease in the fluorescence intensity in the presence of 5.66 M KI. (c) Fluorescence anisotropy of Rose Bengal in water in presence of KI. The increase in the steady-state anisotropy is due to shorter lifetimes.
155
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159 Table 1 Fluorescence lifetime, and quality of fit (v2) for Rose Bengal in water and different concentration of KI.
where ai and si are the pre-exponential factor and fluorescence lifetime, respectively.
Concentration of KI (mol/dm3)
Lifetime (ps)
v2
3. Results and discussion
0 1.13 2.26 5.66
77 ± 3 48 ± 2 31 ± 1 16 ± 1
1.024 1.147 1.221 1.125
3.1. Steady-state fluorescence of RB in absence and presence of KI
IðtÞ ¼
X
ai expðt=si Þ;
ð1Þ
i
We measured fluorescence spectra of RB in absence and presence of KI (Fig. 1a). The intensity of the RB progressively decreases but the shape of the spectrum does not change. Also, the absorption spectra of RB do not change in the presence of KI (not shown). We believe that KI acts as an efficient collisional quencher of RB and does not react with the RB chromophore. This conclusion is
15000
Scatterer λobs= 470 nm
Intensity
10000
FWHM = 111 ps
111 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
3.7
3.8
3.7
3.8
15000
Rose Bengal in water λobs= 572 nm
Intensity
10000
FWHM= 188 ps
188 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
15000
Rose Bengal in 5.66M KI λobs= 572 nm
Intensity
10000
FWHM= 115 ps
115 ps 5000
0 2.9
3.0
3.1
3.2
3.3 3.4 Time [ns]
3.5
3.6
3.7
3.8
Fig. 2. Time-domain results in linear scale collected using 470 nm pulsed laser diode excitation and Hamamatsu R3809U-50 microchannel plate PMT as a detector. These TCSPC histograms were recorded for a scatterer and Rose Bengal in absence and presence of 5.66 M KI. The measured FWHMs are: (a) 111 ps for scatterer; (b) 188 ps for Rose Bengal in absence of quencher and (c) 115 ps Rose Bengal in presence of 5.66 M KI.
156
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
Fig. 3. Intensity decay of the Rhodamin B–Rhodamin 6G mixture in water. The IRF was measured with a scatterer (a). Fractional intensities of the positive decay components (a, top) are 40% for Rhodamin B and 60% for Rhodamin 6G. Graphs (b) and (c) are intensity decays of pure RhB and Rh6G solutions, respectively,
Fig. 4. Intensity decay of Rhodamin B–Rhodamin 6G mixture in water. Quenched RB fluorescence is used as an IRF (a). Fractional intensities of the positive decay components (a, top) are 40% for Rhodamin B and 60% for Rhodamin 6G. Graphs (b) and (c) are intensity decays for RhB and Rh6G solutions, respectively.
157
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
based on the following experimental observations: The Stern– Volmer plot (Fig. 1b) is clearly linear. The quantum yield of the quenched RB is relatively high 0.003 and the brightness of the sample is reasonable and still two orders of magnitude stronger than the Raman scattering. The fluorescence anisotropy (Fig. 1c) increases with KI concentration in accordance with the expected decrease of fluorescence lifetime. The shorter lifetime means less time available for effective fluorescence depolarization by molecular rotation. 3.2. Time-resolved fluorescence of RB in absence and presence of KI Next, we measured the fluorescence intensity decays of RB in absence and presence of KI. The results are presented in Table 1. The errors presented in Table 1 are standard deviation and corresponding to an assumption of 68% confidence [18]. In the presence of 5.66 M KI the lifetime is only 16 ps, which is close to the decay resolution limit of the FT200. The responses of RB fluorescence in absence and presence of KI, as well as the response from colloidal silica (Ludox) scatter are presented in Fig. 2. In order to visualize details, we presented these responses in linear scale. It is clear that
a
the response for the quenched RB (Fig. 2c) at 572 nm observation and the IRF at the 470 excitation (Fig. 2a) are almost identical. 3.3. Collisional quenching of RB by KI We used both area under emission spectra and the lifetimes of progressively quenched Rose Bengal (Table 1) to construct a dynamic version of Stern–Volmer dependence (Fig. 1b). The plot s0/ s versus KI concentration reveal the Stern–Volmer quenching constant of 0.67 M1 and bimolecular quenching constant of about 8 109 M1 s1, taking into account that unquenched RB has the lifetime s0 = 77 ps. However, when fluorescence intensities were used to construct Stern–Volmer dependence, higher values of quenching constants were obtained (Ksv = 1.08 M1, 1 1 kq = 1.4 M s ). We believe this because of a static-quenching at high KI concentrations. 3.4. Use of quenched RB fluorescence as IRF in lifetime measurements First we measured lifetimes of two fluorophores, RhB and Rh6G (about 1 lM each) using scattering as IRF (Fig. 3b and c). These
14000 12000
Scatterer λobs= 470 nm
10000
Intensity
FWHM = 579 ps 8000
579 ps
6000 4000 2000 0 1.0
1.5
2.0
2.5
3.0
3.5
Time [ns]
b
14000 12000
Rose Bengal in 5.66M KI λobs= 572 nm
10000
FWHM = 471 ps
Intensity
8000
471 ps
6000 4000 2000 0 1.0
1.5
2.0
2.5
3.0
3.5
Time [ns] Fig. 5. Impulse response function in linear scale obtained with a scatterer (a) and quenched Rose Bengal (b) for Perkin Elmer APD (SPCM-AQR-14). FWHMs are showed on the graphs.
158
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
compounds show single exponential decays with lifetimes of 1.57 ns (RhB) and 3.90 ns (Rh6G). Than we prepared a mixture of these fluorophores and measured the decay kinetic of it (Fig. 3a). This intensity decay is heterogeneous and can be described with a double exponential model. Recovered lifetimes correspond well to the individual fluorophores. We analyzed these decays using two IRFs. One is obtained by scattering (Fig. 3), the other one is a decay of quenched RB at 572 nm (Fig. 4). Practically identical parameters were recovered (compare Figs. 3 and 4).
gle photon sensitive APDs using the MicroTime200 microscope (PicoQuant GmbH). For both detectors, the IRF FWHM recorded using scattered 470 nm excitation is found to be broader (by 20%) than the decay of quenched RB at 572 nm. However, due to the fast response of MPD APD this broadening is less than 20 ps, whereas for Perkin–Elmer APD it amounts to more than 100 ps under the same conditions.
3.5. Evaluation of IRFs of APD detectors
We demonstrated that using KI quencher, the RB fluorescence emission can be shortened to 16 ps. The fluorescence signal from quenched RB solution is more than two orders of magnitude stronger than Raman scattering and can be easily measured with a commercial fluorometer. This signal can be effectively used as IRF at the emission wavelength in decay curve analysis including reconvolution fitting of more complex decay models. We also showed
During the experiments described above we used an MCP-PMT with negligible color dependence of the timing performance. The temporal response of APD detectors, commonly used today in time-resolved microscopy, are more sensitive to the observation wavelength. We tested Perkin–Elmer (Fig. 5) and MPD (Fig. 6) sin-
a
4. Conclusions
12000
10000
Scatterer λobs= 470 nm
Intensity
8000
FWHM = 120 ps 6000
120 ps
4000
2000
0 1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Time [ns]
b
12000
10000
Rose Bengal in 5.66M KI λobs= 572 nm
Intensity
8000
FWHM= 102 ps 6000
102 ps
4000
2000
0 1.9
2.0
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
Time [ns] Fig. 6. Impulse response function in linear scale obtained with a scatterer (a) and quenched Rose Bengal (b) for MPD APD (PDM 1CTC). FWHMs are showed on the graphs.
M. Szabelski et al. / Chemical Physics Letters 471 (2009) 153–159
that ultra-short fluorescence pulses from quenched RB solutions can be used to evaluate responses from electronic detectors such as APDs.
[4] [5] [6] [7]
Acknowledgements
[8] [9]
This work was supported by Texas Emerging Technologies Fund Grant (CCFT). Mariusz Szabelski and Rafal Luchowski are the recipients of a researcher’s mobility program from the Polish Ministry of Science and Higher Education. References
[10] [11] [12] [13] [14] [15]
[1] P.I.H. Bastiaens, A. van Hoek, W.F. Wolkers, J.-C. Brochon, A.J.W.G. Visser, Biochemistry 31 (1992) 7050. [2] P.A.W. van Den Berg, S.B. Mulrooney, B. Gobets, I.H.M. van Stokkum, A. van Hoek, C.H. Williams Jr., A.J.W.G. Visser, Protein Sci. 10 (2001) 2037. [3] Q.S. Hanley, V. Subramaniam, D.J. Arndt-Jovin, T.M. Jovin, Cytometry 43 (2001) 248.
[16] [17] [18]
159
M. Szabelski et al., Appl. Spectrosc. 63 (2009) 363. J.M. Harris, F.E. Lytle, Rev. Sci. Instrum. 48 (1977) 1469. M. Zuker, A.G. Szabo, L. Bramall, D.T. Krajcarski, Rev. Sci. Instrum. 56 (1985) 14. M. Van Den Zegel, N. Boens, D. Daems, F.C. De Schryver, Chem. Phys. 101 (1986) 311. D.C. Neckers, O.M. Valdes-Aguilera, Adv. Photochem. 18 (1993) 315. C. Giulivi, M. Sarcansky, E. Rosenfeld, A. Boveris, Photochem. Photobiol. 52 (1990) 745. I.E. Kochevar, C.R. Lambert, M.C. Lynch, A.C. Tedesco, Biochim. Biophys. Acta 1280 (2) (1996) 223. J. Lenard, A. Rabson, R. Vanderoef, Proc. Natl. Acad. Sci. USA 90 (1993) 158. J. Chodosh, M.C. Banks, W.G. Stroop, Invest. Ophthalmol. Vis. Sci. 33 (1992) 2520. T.A. Dahl, W.R. Midden, D.C. Neckers, Photochem. Photobiol. 48 (1988) 607. F.S. Cruz, L.A.V. Lopes, W. De Souza, S.N.J. Moreno, R.0. Mason, R. De Campo, Acta Trop. 41 (1984) 99. C.-C. Chang, Y.-T. Yang, J.-C. Yang, H.-D. Wu, T. Tsai, Dyes Pigments 79 (2008) 170. A.J. Bron, Surv. Ophthalmol. 45 (2001) S221. M.S. Roy, P. Balraju, M. Kumar, G.D. Sharma, Sol. Energy Mater. Sol. Cells 92 (2008) 909. J.R. Lakowicz, Principles of Fluorescence Spectroscopy, third edn., Kluwer Academic/Plenum Publishers, 2006.