Beruflich Dokumente
Kultur Dokumente
Department of Chemistry and Biochemistry, Uniersity of Colorado at Boulder, Boulder, CO 80309-0215, USA
Received 14 August 1998; in final form 31 August 1998
Abstract
Two-dimensional Fourier transform electronic spectroscopy is demonstrated. Non-collinear three-pulse scattering allows
selection of electronic coherence order by choice of phase-matching geometry, temporal pulse order, and Fourier transform
variables. Detection of the scattered field at the sample allows separation of the real absorptive and emissive. and imaginary
refractive. contributions to the 2D spectra. The experiments reported here are electronic analogs of the 2D magnetic
resonance techniques known as COSY correlated spectroscopy. and NOESY nuclear Overhauser effect spectroscopy. and
should allow detection of electronic multiple quantum coherence. q 1998 Elsevier Science B.V. All rights reserved.
1. Introduction
Despite the many differences between spin flips
and electronic excitations w1,2x, the extension of
magnetic resonance methods to optical spectroscopy
has been fruitful w36x. In magnetic resonance,
Fourier transform spectra with two frequency dimensions 2D spectra. w7,8x can resolve overlapping peaks
by spreading them out in a second dimension, correlate linked spectroscopic transitions using crosspeaks, map excitation transfer pathways and measure
rates by following the growth of new cross-peaks,
and have decisive advantages over double resonance
for the study of multiple quantum transitions. Ernst
and co-workers also suggested that 2D rotational and
vibrational spectroscopy might be useful w7x. In the
last five years, there has been some progress in
extending 2D spectroscopy beyond magnetic resonance. Tanimura and Mukamel examined 2D Raman
spectroscopy theoretically w9x. Tokmakoff et al. w10x
)
For the second-order nonlinearity of doubling crystals, measurement of the sum frequency and Fourier transform recovery of
the difference frequency allows calculation of either optical frequency, but difference frequencies are not sufficient to determine
the optical frequency for third-order nonlinearities.
0009-2614r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
PII: S 0 0 0 9 - 2 6 1 4 9 8 . 0 1 1 4 0 - 3
308
2. Theory
In order to understand 2D electronic spectroscopy, it is necessary to include the phase of the
signal field. Since signal phases had not been experimentally measured, previous theoretical work on
three-pulse scattering w4,6x did not calculate the
phase. For simplicity, three-pulse scattering signals
will be discussed using a Bloch model. This is
followed by examination of the spatial storage of the
phase information at long times when spectral diffusion w4,6x eliminates frequency gratings and only
population gratings remain.
A three-pulse scattering experiment w4,6x uses
pulses with wavevectors k a , k b , and k c to generate
a signal in the direction k c q k b y k a . For a two-level
system, physical pictures based on scattering of pulse
c off a grating formed by pulses a and b are helpful.
The origin of time is chosen so that pulse c is
centered at time t s 0, and the pulse delays are
t ' t b y t a and T ' min< t a <, < t b <.. When pulses a and
b overlap in the sample, interference creates a spatially periodic intensity pattern w4,6x. The resulting
spatially periodic excitation probability acts as both
an amplitude and phase diffraction grating w18x. For
inhomogeneously broadened transitions, the spatial
population grating is gradually replaced by a joint
spatial-frequency grating when pulses a and b no
longer overlap in time w6x. The spatial-frequency
grating produces delayed echo signals and decays by
spatial diffusion and spectral diffusion.
Examination of the signal field obtained for a
two-level system in the optical Bloch limit using d
function excitation pulses provides some insight into
2D spectra. The third-order nonlinear polarization
with wave-vector k c q k b y k a is readily calculated
from the response functions given by Mukamel w19x.
A frequency-independent pr2 phase shift between
the emitted field and the source polarization, predicted by Maxwells equations, has been included.
For t G 0, the signal field is given by
E t , t , T . A exp yG t q <t < . exp w yTrT1 x
=cos veg t y t . x t y t . ,
1.
G 2 q i G v t q veg .
2
2
v t q veg . q G 2 vt y veg . q G 2
G 2 q i G v t y veg .
2
2
v t y veg . q G 2 vt q veg . q G 2
2.
This complex 2D single quantum spectrum has real
2D absorption lineshapes peaked at v t s yvt s
"veg and imaginary dispersion lineshapes along
v t . Inhomogeneous broadening yields a convolution with the 2D inhomogeneous lineshape
x v t y vt . r'2 . d v t q vt . r'2 . , which broadens the signal along the diagonal line v t s yvt but
not perpendicular to the diagonal.
Storage of the electronic frequency dependent
phase in a population grating can be illuminated by
considering interference between monochromatic
plane waves. Crossing waves produce a spatial intensity grating
I r . s 2 I 1 q cos k a y k b . P r y vt
.,
3.
3
This definition of the Fourier transform matches Ref. w15x and
seems to be most common in optical spectroscopy e.g., Ref. w2x..
Unfortunately, the sign of i is reversed from Ref. w8x, requiring
complex conjugation before comparison with most work on 2D
NMR.
309
310
3. Experiment
Three-pulse scattering signals w4x were generated
by pulses from a 10 kHz cavity dumped mode-locked
Ti:sapphire laser. Prisms pre-compensated for dispersion in the five-beam interferometer w15x. Excitation
pulses had a 41 nm wide spectrum centered at 797
nm and a 29 fs pulse duration determined by
second-harmonic FROG w16x. Beams a, b, and c
; 270 pJrpulse. were aligned to three corners of a
square and focused to a 40 mm spot in the sample to
generate a four-wave mixing signal with wave-vector
k s s k c q k b y k a on the fourth corner w15x. After
collimation, the signal was spatially combined with a
temporally non-overlapping reference at a beamsplitter and imaged into a spectrograph with a CCD array
which measured the spectral interferogram w17,22,23x.
This spectral interferogram determined the signal
field amplitude e v . and the phase difference
D f v . between the signal and reference at the
beamsplitter. Characterization of a tracer beam focused into the sample along the path of signal emission k t s k s . and measurement of the tracer-reference phase difference by spectral interferometry determined the signal amplitude and phase at the sample exit w15x.
A 0.4 mM solution of IR144 in methanol w24x
flowed at 2 mrs through a 100 mm pathlength
sample cell nominal OD f 0.15.. The experimental
conditions reproduced previous three-pulse echo peak
shift results w25x, yielded signal fields linear in each
excitation field, and produced signal spectra and
phases which are independent of pulse energy. To
reference the phase to t s 0, the delay of pulse c and
the tracer were fixed at t s 0 while the delays for
pulses a and b were scanned in 0.66 fs steps so as to
vary only t or only T. At each delay, a spectral
interferogram was recorded using 0.11 s CCD exposures, depending on T. Recording 601 spectral
interferograms as t was scanned from y200 to
q200 fs for a 2D electronic spectrum took a total of
315 min. After recovery of the signal field in the
frequency domain at each delay w15x, the data were
Fig. 1. 2D power spectrum square of the inverse Fourier transform with respect to t and T . of a transient grating scan t s 0.
extending to the Nyquist frequency of the indirectly detected
dimension v T .. The directly detected dimension v t . extends
from 1.994 to 2.750 radrfs spectrograph bandwidth. and is
filtered by the excitation pulse spectra. This scan is sensitive to
population relaxation and vibrational coherence within each electronic state, but not sensitive to electronic coherence. This 2D
spectrum represents an electronic v t . Raman v T . correlation
spectrum. Contours are at 10%, 50%, and 90% of the maximum.
311
4. Results
Fig. 1 shows the 2D power spectrum of a transient grating experiment T scanned at fixed t s 0,
Fourier transformations with respect to t and T ..
During T, the experiment is sensitive to population
dynamics and vibrational wavepacket motion on each
electronic surface, but is blind to single quantum
electronic coherence. The resulting 2D spectrum
peaks around zero frequency in the indirectly detected dimension v T .. Fig. 2 shows the 2D power
spectrum of a three-pulse echo experiment t scanned
at T s 0, Fourier transformations with respect to t
and t .. During t , the experiment is sensitive to
single quantum electronic coherence, and the resulting 2D spectrum is peaked on or near the diagonal at
v t f yvt f 2.34 radrfs see Fig. 3..
Fig. 3 shows the real absorptiveremissive. and
imaginary dispersive. parts of the 2D spectrum obFig. 3. Real top. and imaginary bottom. parts of the 2D spectrum of a three-pulse echo scan with T s 0. The real part represents signal caused by changes in absorption and stimulated
emission. Contours are at 20%, 30%, 40%, 50% bold., 60%,
70%, 80%, and 90% of the maximum. The imaginary part represents signal caused by refractive index changes. Contours are
drawn at "30%, "40%, "50% bold., "60%, and "70% of
the absorptive real. maximum. Dotted contours indicate negative
values and solid contours indicate positive values. The half-maximum contour of the 2D absorption signal is overlaid as a dashed
curve. The interferometric stage calibration was not in register
with the scan so the indirectly detected frequencies vt . have
error bars of ; 0.04 radrfs. The 2D absorption peak is elongated
along the diagonal, suggesting the presence of inhomogeneous
broadening.
312
5. Discussion
Electric field detection should allow direct transposition of 2D NMR techniques to electronic and
infrared. spectroscopy. Transitions with a common
initial or intermediate level will produce cross-peaks
at T s 0, analogous to correlation spectroscopy
COSY 2D NMR. w8x. The diagonal 2D absorption at
6. Conclusions
Although the experiments reported here are in
some respects preliminary, there is sufficient agreement between the data and models of the 2D spectra
to confirm the basic theory. A non-collinear geometry is possible for 2D electronic spectroscopy if the
signal field is measured so that phase modulation at
the initial absorption frequency can be used to correlate excitation and scattering frequencies. This phase
modulation is stored in the robust position of population grating maxima. A non-collinear geometry can
probe a selected electronic coherence order w8x or
set of density matrix pathways. w2x, effectively substituting for the phase cycling procedures used in
NMR w8x. 2D studies of excitonic coupling and
energy transfer in aggregates are feasible with available femtosecond lasers. The prospect of following
electrons with extremely broadband 2D or 3D. 5
electronic spectroscopy is tantalizing.
Acknowledgements
We thank Arthur Pardi and Warren Warren for
helpful discussions of 2D NMR. This work was
supported by the Petroleum Research Fund administered by the American Chemical Society, the Camille
and Henry Dreyfus Foundation, the David and Lucile
Packard Foundation, and the National Science Foundation.
References
w1x W.S. Warren, A.H. Zewail, J. Phys. Chem. 85 1981. 2309.
w2x S. Mukamel, Principles of Nonlinear Optical Spectroscopy,
Oxford University Press, New York, 1995.
w3x I.D. Abella, N.A. Kurnit, S.R. Hartmann, Phys. Rev. 141
1966. 391.
w4x A.M. Weiner, S. De Silvestri, E.P. Ippen, J. Opt. Soc. Am. B
2 1985. 654.
w5x L. Allen, J.H. Eberly, Optical Resonance and Two-Level
Atoms, Dover, New York, 1987.
w6x D.A. Wiersma, K. Duppen, Science 237 1987. 1147.
w7x W.P. Aue, E. Bartholdi, R.R. Ernst, J. Chem. Phys. 64
1976. 2229.
5
Fourier transformation of the third-order polarization resulting
from a causal third-order nonlinear response with respect to the
two pulse delays and the time after the third pulse yields a 3D
frequency domain polarization which is a product of a triple
FourierLaplace transform of the third-order time-domain response function and the Fourier transforms of each of the three
time-domain electric fields the signs of all frequencies for pulses
which contribute a negative wavevector to the signal direction are
reversed.. The effect of the excitation pulse spectra is simply to
filter and possibly phase distort via chirp. the 3D frequency-domain response function. 3D spectroscopy using broadband, fully
characterized pulses should thus allow measurement of the complete third-order response function.
313
w8x R.R. Ernst, G. Bodenhausen, A. Wokaun, Principles of Nuclear Magnetic Resonance in One and Two Dimensions,
Oxford University Press, Oxford, 1987.
w9x Y. Tanimura, S. Mukamel, J. Chem. Phys. 99 1993. 9496.
w10x A. Tokmakoff, M.J. Lang, D.S. Larsen, G.R. Fleming, V.
Chernyak, S. Mukamel, Phys. Rev. Lett. 79 1997. 2702.
w11x L. Lepetit, M. Joffre, Opt. Lett. 21 1996. 564.
w12x J.D. Hybl, A.W. Albrecht, S.M. Gallagher Faeder, D.M.
Jonas, in: W. Zinth, J.G. Fujimoto, T. Elsaesser, D. Wiersma
Eds.., Ultrafast Phenomena XI, Springer, Berlin in press..
w13x N.F. Scherer, R.J. Carlson, A. Matro, M. Du, A.J. Ruggiero,
V. Romero-Rochin, J.A. Cina, G.R. Fleming, S.A. Rice, J.
Chem. Phys. 95 1991. 1487.
w14x L. Xu, C. Spielmann, A. Poppe, T. Brabec, F. Krausz, T.W.
Hansch,
Opt. Lett. 21 1996. 2008.
1995. 2467.
w18x K.A. Nelson, R. Casalegno, R.J. Dwayne Miller, M.D. Fayer,
J. Chem. Phys. 77 1982. 1144.
w19x S. Mukamel, Annu. Rev. Phys. Chem. 41 1990. 647.
w20x W.S. Warren, W. Richter, A.H. Andreotti, B.T. Farmer II,
Science 262 1993. 2005.
w21x D.P. Weitekamp, Adv. Magn. Reson. 11 1983. 111.
w22x D.N. Fittinghoff, J.L. Bowie, J.N. Sweetser, R.T. Jennings,
M.A. Krumbugel,
K.W. DeLong, R. Trebino, I.A. Walmsley,
w25x T. Joo, Y. Jia, J.-Y. Yu, M.J. Lang, G.R. Fleming, J. Chem.
Phys. 104 1996. 6089.
w26x F.A. Jenkins, H.E. White, Fundamentals of Optics, 4th ed.,
McGraw-Hill, New York, 1976.
w27x P.H. Vaccaro, in: R.W. Field, E. Hirota, J.P. Maier, S.
Tsuchiya Eds.., Nonlinear Spectroscopy for Molecular
Structure Determination, Blackwell, Oxford, 1997, p.75.
w28x R. Trebino, K.W. DeLong, D.N. Fittinghoff, J.N. Sweetser,
M.A. Krumbugel,
B.A. Richman, Rev. Sci. Instrum. 68
1997. 3277.
w29x T.W. Mossberg, R. Kachru, S.R. Hartmann, A.M. Flusberg,
Phys. Rev. A 20 1979. 1976.
w30x M. Bellini, A. Bartoli, T.W. Hansch,
Opt. Lett. 22 1997.
540.
w31x W.M. Zhang, V. Chernyak, S. Mukamel, in: W. Zinth, J.G.
Fuijimoto, T. Elsaesser, D. Wiersma Eds.., Ultrafast Phenomena XI, Springer, Berlin in press..