Anal. Chem. 2004, 76, 3794-3799
Development of a Total Internal Reflection Ultrafast Transient Lens Method for Studying Molecular Dynamics on an Interface Tsuyoshi Sugimoto,† Yasushi Hirose,† Hiroharu Yui,*,†,‡ and Tsuguo Sawada†,‡
Department of Advanced Materials Science, Graduate School of Frontier Sciences, The University of Tokyo, 5-1-5-401 Kashiwanoha, Kashiwa, Chiba 277-8561, Japan, and Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST), 5-1-5-401 Kashiwanoha, Kashiwa, Chiba 277-8561, Japan
We have developed the total internal reflection ultrafast transient lens (TIR-UTL) method to detect nonradiative chemical processes at interfaces and surfaces with subpicosecond time resolution. In the TIR-UTL measurements, the evanescent field of a pump beam irradiated under the TIR condition generates a refractive index change. The refractive index change is attributed to changes of the molecular electronic state, of density by molecular orientation/structure change, and of temperature by vibrational relaxation processes. The refractive index change is detected as a change of the power intensity of the probe beam adjusted coaxially with the pump beam. At first, we disscuss a theoretical principle of a coaxial configuration in the TIR-UTL measurement. This configuration has an advantage of versatility over the established TIR configuration. Then, we evaluate time resolution of TIR-UTL and obtain a value of less than 400 fs. We measure the ultrafast molecular dynamics of the cationic chromophore Auramine O (AuO) at a silica/water interface. Two slow time constants originating from AuO adsorbed on the silica surface are detected by TIR-UTL. These are attributed to AuO, whose twisting motion is strongly hindered by adsorption on a silica surface. It is important both industrially and scientifically to observe chemical phenomena at interfaces and surfaces.1 These areas play critical roles in electrochemical, catalytic, and biological reactions. Various kinds of ultrathin devices utilizing interfaces and surfaces have been developed from advances made in nanotechnology in the past decade.2-7 For example, applications of photofunctional * To whom correspondence should be addressed. E-mail: yui@ laser.t.u-tokyo.ac.jp. † The University of Tokyo. ‡ CREST/JST. (1) Adair, J. H., Casey, J. A., Venigalla, S., Eds. Handbook on Characterization Techniques for the Solid-Solution Interface; American Ceramic Society: Westerville, OH, 1993; Chapter 3, p 151. (2) Maenosono, S.; Dushkin, C. D.; Yamaguchi, Y. Jpn. J. Appl. Phys. 2000, 39, 4006. (3) Lewis, L. N. Chem. Rev. 1993, 93, 2693. (4) Huynh, W. U.; Peng, X.; Alivisatos, A. P. Adv. Mater. 1999, 11, 923. (5) Kitano, H.; Taira, Y.; Yamamoto, H. Anal. Chem. 2000, 72, 2979. (6) Seki, T.; Sakuragi, M.; Kawanishi, Y.; Suzuki, Y.; Tamaki, T.; Langmuir 1993, 9, 218.
3794 Analytical Chemistry, Vol. 76, No. 13, July 1, 2004
membranes and nanoparticles coated on substrates have been put forth for such industrial products as recording media,2 catalysts,3 solar cells,4 molecular sensors,5 and display devices.6 Although the materials at interfacial areas are of importance and interest, the amounts there are generally very minute compared with those in the bulk. It is crucial to reduce the bulk contribution to selectively obtain a signal from samples at interfacial areas. Total internal reflection (TIR) spectroscopy has been developed for this purpose.11-19 Since the TIR configuration allows penetration of the evanescent field of an incident light into the sample only adjacent to the interface, it lowers the bulk contribution. TIR spectroscopy allows measurement of chemical phenomena at the interface and surface with high selectivity. For example, attenuated total internal infrared spectroscopy (ATRIR), slab optical waveguide spectroscopy, and total internal reflection fluorescence spectroscopy have been developed as infrared, UV-visible absorption, and luminescence spectroscopy, respectively.8-15 Interfacial molecular dynamics,12-15 structure of polymer films,8 and bio-protein materials9-11 have been analyzed by these spectroscopic technique. Recently, much attention has been paid to nonradiative processes of the nanosecond to picosecond time scale at interfaces and surfaces. This is because nonradiative processes of the ultrafast time scale, such as the molecular orientation change of (7) Baba, T.; Minakawa, H.; Hato, M.; Motoki, A.; Hirano, M.; Zhou, D.; Kawasaki, K. Biochem. Biophys. Res. Commun. 1999, 265, 734. (8) Sammon, C.; Mura, C.; Yarwood, J.; Everall, N.; Swart, R.; Hodge, D. J. Phys. Chem. B 1998, 102, 3402. (9) Kato, K.; Takatsu, A.; Matsuda, N.; Azumi, R.; Matsumoto, M. Chem. Lett. 1995, 437. (10) Matsuda, N.; Takatsu, A.; Kato, K. Chem. Lett. 1996, 105. (11) Santos, J. H.; Matsuda, N.; Qui, Z. M.; Yoshida, T.; Takatsu, A.; Kato, K. Surf. Interface Anal. 2003, 35, 1. (12) Bell, M. A.; Crystall, B.; Rumbles, G.; Porter, G.; Klug, D. R. Chem. Phys. Lett. 1994, 221, 15. (13) De Mello, A. J.; Crystall, B.; Rumbles, G. J. Collid Interface Sci. 1995, 69, 169. (14) Ishizaka, S.; Nakatani, K.; Habuchi, S.; Kitamura, N. Anal. Chem. 1999, 71, 419. (15) Wirth, M. J.; Burbage, J. D. Anal. Chem. 1991, 63, 1311. (16) Shimosaka, T.; Sugii, T.; Uchiyama, K.; Hobo, T. Bunseki Kagaku 1998, 47, 1099. (17) Shimosaka, T.; Sugii, T.; Hobo, T.; Ross, J. B. A.; Uchiyama, K. Anal. Chem. 2000, 72, 3532. (18) Shimosaka, T.; Izako, M.; Uchiyama, K.; Hobo, T. Analyst 2003, 128, 562. (19) Fujinami, M.; Murakawa, H.; Sawada, T. Rev. Sci. Instrum. 2003, 74, 352. 10.1021/ac035526r CCC: $27.50
© 2004 American Chemical Society Published on Web 05/22/2004
photoswitching, local charge-transfer reactions in a microchannel, and molecular structure changes of bio-proteins by adsorption on a surface, play crucial roles in elemental chemical processes at interfaces and surfaces. The TIR-thermal lens (TIR-TL) method has also been developed over the past decade to measure nonradiative processes by refractive index change.14-19 Refractive index change indicates not only changes of the molecular electronic state but also changes of density with molecular orientation/structure changes and of temperature with vibrational relaxation processes.20-33 It is expected that combining the TIRTL method with absorption and luminescence spectroscopy will provide more complete information on chemical phenomena. Although the TIR-TL is a useful technique to measure nonradiative processes, its time resolution is currently limited to the millisecond to microsecond order. A way to measure ultrafast nonradiative processes at interfaces and surfaces is highly desirable. Here, we develop a total internal reflection ultrafast transient lens (TIR-UTL) method. The TIR-UTL principle is based on the UTL method,20-26 which we developed to measure ultrafast molecular dynamics in bulk solution. Time resolution of TIR-UTL is subpicosecond order, so ultrafast molecular nonradiative dynamics at the interface and surface can be measurable. In this study, we propose a coaxial configuration in the TIR-UTL measurement. Then, we construct the TIR-UTL apparatus and evaluate the time resolution of the apparatus. We measure an auramine-O (AuO) aqueous solution by TIR-UTL and compare the result with that of a bulk measurement. We are able to detect the signal from AuO adsorbed on a silica surface where its twisting motion is strongly hindered. EXPERIMENTAL SECTION TIR-UTL Principle. TIR-UTL utilizes a pump-probe technique to realize high time resolution. Figure 1 shows the principle of TIR-UTL. When a pump beam is incident to the hemispherical prism (fused silica φ ) 15 mm), an evanescent field is generated near the interface. The penetration depth of the evanescent field to medium on the sample side (dp) is defined by eq 1.
dp )
λ 2n1πxsin 2 θ - n122
(1)
Here n1 is refractive index of the medium on the prism side and n2 is refractive index of the medium on the sample side. n12 is (20) Ito, K.; Tsuyumoto, I.; Harata, A.; Sawada, T. Chem. Phys. Lett. 1997, 275, 349. (21) Furui, G.; Ito, K.; Tsuyumoto, I.; Harata, A.; Sawada, T. J. Phys. Chem. A 1999, 103, 7575. (22) Ito, K.; Muto, M.; Harata, A.; Sawada, T. Chem. Phys. Lett. 2000, 318, 1. (23) Hirose, Y.; Yui, H.; Fujinami, M.; Sawada, T. Chem. Phys. Lett. 2001, 341, 29. (24) Takei, M.; Hirose, Y.; Yui, H.; Sawada, T. J. Phys. Chem. A 2002, 105, 11395. (25) Yui, H.; Takei, M.; Hirose, Y.; Sawada, T. Rev. Sci. Instrum. 2003, 74, 907. (26) Hirose, Y.; Yui, H.; Fujinami, M.; Sawada, T. Rev. Sci. Instrum. 2003, 74, 898. (27) Terazima, M. Opt. Lett. 1995, 20, 25. (28) Chang, Y. J.; Simon, J. D. J. Phys. Chem. 1996, 100, 8613. (29) Terazima, M.; Hirota, N. J. Phys. Chem. 1993, 97, 10554. (30) Terazima, M.; Hara, T.; Hirota, N. J. Chem. Phys. 1994, 100, 2481. (31) Terazima, M. Chem. Phys. 1994, 189, 793. (32) Terazima, M. Chem. Phys. Lett. 1994, 230, 87. (33) Terazima, M. J. Chem. Phys. 1996, 105, 6587.
Figure 1. Principle of TIR-UTL measurements. (a) A pump beam is incident under the TIR condition. (b) The evanescent field of the pump beam generates the UTL effect (refractive index change) at the silica/water interface. (c) The UTL effect is detected by the change of the power density of the probe beam.
n2/n1, the relative refractive index. θ is incident angle. λ is wavelength of incident light. The evanescent field of the pump beam generates a spatial distribution of the refractive index change near the interface. The ultrafast refractive index change, which originates from the photochemical phenomena, is called the UTL effect. The origins of the UTL effect have been reported elsewhere.27-33 The UTL effect changes the power intensity of the probe beam adjusted coaxially with the pump beam through a spatial filter. The change of the power intensity is detected by an avalanche photodiode. The theoretical principle of TIR-UTL is given in Theoretical Principle of the TIR-UTL Configuration. TIR-UTL Apparatus. A Ti:sapphire laser (Coherent, MIRA 900 F; wavelength centered at 800 nm; repetition 76 MHz) was used as a light source for TIR-UTL and UTL. The laser beam was separated into two beams by a beam splitter. One beam, which went through the optical delay line, was the probe beam; the other was used as the pump beam. The latter was intensity-modulated (1.1 MHz) with AOM and frequency doubled by a BBO crystal. The pump beam and the probe beam were adjusted to be collinear. We fabricated a sample prism unit (Figure 2) and adjusted the incident angle and focal point of the pump beam and the probe beam on the interface by an xyzθ stage. The incident angle of the beams was set at θ ) 86° and dp was calculated as 100 nm from eq 1. The polarizations of the beams were set parallel. The UTL effect was detected by an avalanche photodiode (detection area, 0.1 mm) as a change of a probe beam intensity. The output of the Analytical Chemistry, Vol. 76, No. 13, July 1, 2004
3795
Figure 4. Relationship between reflection angle and phase of irradiated light to the interface. Figure 2. Sample-prism unit for the TIR-UTL measurement. The unit was set on an xyz θ stage for adjustment of the focal point and incident angle.
deformation of the reflection plane (Figure 4). The relation is described as follows.
dθ dh 1 ) ∆φ ) 2φ ∼ 2 k cos φ dx dx Figure 3. Molecular structure of AuO.
photodiode was enhanced by a preamplifier and cutoff highfrequency component by a passive band-pass filter, and then the signal was sent to a lock-in-amplifier. Sample Systems. Figure 3 shows the molecular structure of AuO. AuO has an absorbance at pump beam wavelength. AuO is a cationic chromophore; thus, AuO is adsorbed on a silica surface, which is negatively charged by ionization of silanol groups in water.10 This interaction increases the population of AuO at the interface. The increased population of AuO increases the signal originating from the interfacial molecular dynamics. Additionally, it is expected that the lifetime of AuO increases by adsorbing on the silica surface because the twisting motion of AuO that promotes its relaxation process is strongly hindered. For these reasons, AuO is suitable for a test sample of TIR-UTL. AuO (Wako Chemicals) was used without further purification. It was dissolved in purified water. The material of the prism is fused silica. The concentrations of AuO were 0.40 mM for UTL and 1.0 mM for TIR-UTL. RESULT AND DISCUSSION Theoretical Principle of the TIR-UTL Configuration. We adopted a coaxial configuration of a pump beam and a probe beam in the TIR-UTL measurement. Although a vertical configuration is generally adopted in the TIR-TL measurement,14-18 it is difficult to adjust the focal points of pump and probe beams temporally and spatially at the interface. On the other hand, the coaxial configuration is relatively simple and offers advantages of versatility and instrumentation over the vertical configuration. In the present discussion of the theoretical principle, we use a onedimensional theoretical model to simplify calculation. A detailed treatment has been reported elsewhere.34 Here, we treated the polarizations of a probe beam and a pump beam as parallel. When the phase of a light changes in spatial nonuniformity by reflection, the phase front of the light changes. The phase change can be equated with deflection of the angle.34,35 Therefore, the spatial distribution of the phase change is treated as a (34) Hirose, Y. Ultrafast dynamics of solution in restricted environments. Doctoral thesis, University of Tokyo, 2003. (35) Rothenberg, J. E. Opt. Lett. 1988, 13, 713.
3796
Analytical Chemistry, Vol. 76, No. 13, July 1, 2004
(2)
Here, k is a wavenumber, θ is a change of phase, ∆φ is deflection angle, and h is change of reflection plane. The phase change of the P polarized light under the TIR condition is defined by Fresnel’s formula. It is expressed as follows.
tan
θp xsin2 φ1 - n122 )2 n 2 cos φ 12
(3)
1
Here φ1 is reflection angle (sin φ1 > n12 ) and θp is deflection angle of P polarized light. Equation 4 is the differentiation of eq 3 with respect to the n12 direction.
dθp ) dn12
2n12 cos φ1(2 sin2φ1 - n122) (n124 cos2 φ1 + sin2 φ1 - n122)x sin2 φ1 - n122
(> 0) (4)
Equation 4 means the phase change depends on the relative refractive index, n12. For this reason, the spatial distribution of the refractive index change generates the phase change of reflection light. The intensity of the pump beam, I(x) is expressed as eq 5. We hypothesize that the refractive index on the sample side is described as eq 6.
I(x) ) I0 exp(- (x2/ω02))
(5)
n2 ) n20 + n2′I(x)
(6)
Here n20 is the refractive index of the sample side before light irradiation and n2′ is the refractive index change by irradiation. In this case, the spatial distribution of n2 follows a Gaussian distribution (eq 7). When the interface is irradiated by the pump
( )
dn2 2n2′ x2 ) - 2 x exp - 2 dx ω0 ω0
(7)
beam under the TIR condition, a phase change distribution is generated in the probe beam by reflection. The phase change can
Figure 5. Principle of the TIR-UTL with the coaxial configuration. (a) The increase of the refractive index (n2′ > 0) behaves like a convex mirror. (b) The decrease of the refractive index (n2′ < 0) behaves like a concave mirror.
be converted to the deformation of the reflection plane as shown in eq 2. The change of reflection plane is described as follows.
dhp 1 ∼ dx 2k cos φ1
{
-
2n2′ n1ω0
2n0 cos φ1(2 sin φ1 - n0 ) (n04 cos2 φ1 + sin2 φ1 - n02)xsin2 φ1 - n02
( )}
x exp -
2
2
2
x2 ω02
( )
) - n2βpx exp -
x2 (βp > 0) ω02 (8)
Here, n0 is n20/n1. When the change of the reflection plane is approximated by a spherical mirror (x , ω0), the curvature radius, R is described as follows.
R ) 1/-n2′β
(9)
The increase of the refractive index (n2′ > 0) behaves like a convex mirror (R < 0). On the other hand, the decrease of the refractive index (n2′ < 0) behaves like a concave mirror (R > 0) (Figure 5). We calculate the diameter change of the probe beam by the ABCD matrix method. The result is described as follows.
(1 + 2n2′βl)r0 + lr ′0 r ) rn2′)0 r0 + lr ′0
(10)
Figure 6. (a) TIR-UTL signals of AuO aqueous solution and pure water. (b) TIR-UTL signal of AuO aqueous solution (long time scale). The concentrations of AuO solution was 1.0 mM for TIR-UTL. The vertical axis of the signals, “normalized signal intensity” is the refractive index change of the samples. The horizontal axis is delay time of the probe beam. We standardized the TIR-UTL signal as the OKE is positive because it is well known that the refractive index change with OKE of water and fused silica is positive. Then we normalized the TIR-UTL signal of water at the maximum of the signal. The signal of AuO solution is normalized at the minimum of the signal. The black line is the TIR-UTL signal of AuO in water. The gray line is the TIR-UTL signal of pure water. The OKE was observed in both measurements. The decrease of the TIR-UTL signal originated from solute molecules.
r0 and r ′0 are the diameter and the divergence of the probe beam at the interface, respectively, and l is the distance from the reflection plane. 2n2′ βl is far smaller than 1 under the TIR-UTL experimental condition; therefore, the signal intensity of TIR-UTL, STIR-UTL is expressed as follows.
STIR-UTL )
I - I0 -2n2′βlr0 ∼ I0 r0 + lr ′0
(11)
Here, I and I0 are the power density of the probe beam with and without the pump beam, respectively. From eqs 10 and the 11, we conclude that the TIR-UTL signal reflects the change of refractive index at the interface. It should be noted that the sign of the TIR-UTL signal is opposite that of the UTL signal. When the refractive index changes, n2′, in the vertical configuration is approximated by a lens (wavelength thickness), the signal intensity of the vertical configuration can be calculated.34 The sensitivity of signal intensity in the coaxial optical configuration is at the same level as that of the vertical configuration; thus, we adopted the coaxial configuration for the reasons of its simplicity of alignment and versatility. Evaluation of Time Resolution of TIR-UTL. We measured the AuO aqueous solution and pure water to evaluate the time resolution of TIR-UTL. Figure 6 shows these TIR-UTL signals. Analytical Chemistry, Vol. 76, No. 13, July 1, 2004
3797
Figure 7. TIR-UTL signal and UTL signal of AuO aqueous solution. The concentrations of AuO solution were 0.4 mM for UTL and 1.0 mM for TIR-UTL. We normalized the signals in the same way of Figure 6. The black line is the TIR-UTL signal of AuO solution. The gray line is the UTL signal of AuO solution. Table 1. Time Constants of TIR-UTL and UTL Signalsa
τ1 τ2 τ3
UTL (ps)
TIR-UTL (ps)
