Coherent anti-Stokes Raman spectroscopy (CARS) has been widely used as a powerful tool for chemical sensing, molecular dynamics measurements, and rovibrational spectroscopy since its development over 30 years ago, finding use in fields of study as diverse as combustion diagnostics, cell biology, plasma physics, and the standoff detection of explosives. The capability for acquiring resolved CARS spectra in multiple spatial dimensions within a single laser shot has been a long-standing goal for the study of dynamical processes, but has proven elusive because of both phase-matching and detection considerations. Here, by combining new phase matching and detection schemes with the high efficiency of femtosecond excitation of Raman coherences, we introduce a technique for single-shot two-dimensional (2D) spatial measurements of gas phase CARS spectra. We demonstrate a spectrometer enabling both 2D plane imaging and spectroscopy simultaneously, and present the instantaneous measurement of 15 000 spatially correlated rotational CARS spectra in N2 and air over a 2D field of 40 mm2.

Coherent anti-Stokes Raman spectroscopy (CARS) yields laser-like signals that are chemically selective, temperature sensitive, and spatially resolved. Due to the uniqueness of molecular rotational and vibrational Raman spectra, CARS has often been used as a chemical sensing tool. In this context, CARS techniques have been developed for species identification in the microscopic imaging of biological tissues,1,2 the standoff detection of explosives,3 and species concentration measurements in combustion.4,5 While some of the early work in gas-phase CARS required tuning the frequency of one of the beams to generate a CARS spectrum,6,7 multiplex CARS utilizing broadband preparation pulses has been implemented7–9 to measure the population distribution within the rotational and vibrational energy levels of the probed molecules within a single laser shot, making CARS an exquisite probe of the local instantaneous temperature,10 a scalar of vital importance to the probing of reacting flows.11 Chemical effects, such as reaction rate constants, and physical effects, such as gas expansion and heat transfer, are directly linked to the instantaneous thermal field,12 and CARS is often held as the gold standard for nonintrusive gas-phase measurements of thermometry. CARS signals are emitted as a coherent laser-like beam, yielding high collection efficiencies even at large distances away from the probed volume, making the technique ideally suited for probing in highly luminous or scattering environments, such as those often encountered in the study of combustion, plasmas, and particle-laden flows. Ultrafast, time-resolved, spectroscopy techniques have greatly extended the utility of CARS measurements recently. The use of transform limited femtosecond (fs) pulses in the preparation of Raman coherences has been shown to be extremely efficient13 as many pairs of photons within the bandwidth of the pump/Stokes pulse envelopes combine in phase to amplify the CARS signal. Hybrid CARS utilizes transform limited fs pulses to efficiently drive the Raman coherences, and a tailored probe pulse, often in the picosecond (ps) regime, to yield spectrally resolved signals in a single-laser shot.14,15 The applicability of the CARS technique has been extended using time-resolved probing to map the dephasing mechanisms for a number of molecules following coherent excitation.16–21 In rotational CARS, mixtures of multiple species yielding a complex spectrum can be simplified as the coherence from the small molecules is significantly longer-lived; thus, an appropriate probe delay can yield simplified spectra.22,23

One of the long-standing goals in CARS measurements has been the capability to obtain CARS spectra in multiple spatial dimensions simultaneously. Such single shot two-dimensional (2D) measurements would significantly add to the information available for the rigorous comparison between numerical simulations of complex dynamical systems and experiments for model validation and development. However, up to now, if a 2D image of CARS spectra was desired, sample- or laser-scanning techniques have been used to collect 2D data across many laser shots, thus forfeiting time-resolution altogether in favor of more complete spatial information. But in the probing of rapidly fluctuating systems, which are often encountered in turbulent combustion and fluid dynamics measurements, for instance, single-laser-shot data are needed to probe the system instantaneously. Single shot 2D measurements can potentially overcome many of the particular difficulties in comparing measurement results with the results a model provides, such as in the modeling of combusting flows,24 and fundamentally increase the ongoing progress of comparisons between numerical simulations of multidimensional phenomena and experiments.

In the pursuit of a multiplex 2D-CARS signal in a single laser shot, two significant hurdles are encountered. The first challenge is the generation of a spatially resolved 2D signal. In the conventional crossed-beam CARS phase-matching scheme for gas phase CARS measurements,25 there is no arrangement of pump, Stokes, and probe which follows the phase-matching condition that allows for the generation and probing of a 2D-CARS signal in a well spatially resolved manner. In this work, we first address the generation of a 2D-CARS signal by implementing a newly developed phase-matching scheme in which only two beams are required to generate a CARS signal as opposed to the three beams (pump/Stokes/probe) in the conventional gas-phase CARS arrangement. Crossed-beam CARS using only a single pump/Stokes coherence preparation beam and a probe beam has been recently discovered.26,27 In this scheme, both pump and Stokes photons used to drive the Raman coherence are obtained from the same broadband laser pulse. At low probe crossing angle, this two-beam nonlinear interaction is phase-matched over a large range of transition frequencies, enabling two-beam CARS measurements over a transition frequency range limited primarily only by the bandwidth of the excitation laser. For instance, at a two-beam crossing angle of 4°, the phase-mismatch induced CARS signal intensity reduction is less than 50% for Raman transitions up to 3000 cm−1 covering the fundamental vibrational and rotational transitions of most all molecules.28 

The second challenge associated with 2D-CARS is the spectrally resolved detection of the 2D image within a single laser shot. Here, we describe a spectrometer which allows for spectrally resolved 2D imaging of the isolated Raman transitions in a single laser shot. The spectrometer is conceptually similar to the tomographic hyperspectral imaging spectrometer,29 with the exception that we probe narrow, well-isolated spectral lines, significantly simplifying the data analysis. For instance, for the S-branch rotational transitions of N2 molecules, the Raman linewidth is <0.1 cm−1 at 1 bar, and the spacing of the rotational transitions is 8 cm−1.

Figure 1 displays the experimental setup for the 2D-CARS measurements reported here and the 2D imaging spectrometer developed. We implemented the hybrid fs-pump/ps-probe CARS scheme15 to probe the rotational manifold population distribution of the ground vibrational state of gas-phase N2 and O2. A nearly transform-limited 45 fs pump/Stokes beam at 800 nm is formed into a sheet by focusing it to the probe volume with a f = 1000 mm cylindrical lens. The 532 nm probe beam remains collimated, and intersects the pump/Stokes sheet at an angle of 6°. The probe beam was 90-ps in duration, providing high spectral resolution (∼0.25 cm−1) in the rotational CARS spectrum. The generated coherent 2D signal was relay imaged through a diffraction grating (3600 lines/mm), to the CCD.28 The diffraction grating was placed after the image collimating lens, and the diffracted light was sent back at Littrow's angle to the lens to suppress image distortions as seen in Fig. 1. The signal was then reflected to the CCD. The signal generation plane is thus relay imaged to the face of the CCD. Each of the rotational transitions probed get mapped to a different, isolated, position on the CCD. To assess the effect of the grating on the imaging quality a mask was placed in the probe beam to provide structured illumination. As seen in Fig. 1, the probe light passing through this mask was imaged to the beam crossing, and the signal from an individual rotational transition of N2, J = 8, is shown in blue. Qualitatively, it is clear from looking at the image scattered from a single rotational coherence that the image is well-maintained through the detection system. The structured illumination serves as a resolution target for evaluating the imaging quality and any distortions to the J-specific images. The 4-mm mask is down-collimated in the imaging optics 2:1 to the beam crossing. Given the crossing angle of 6°, this arrangement probes a 2D spatial field of ∼2 mm × 20 mm.

FIG. 1.

The experimental setup for the 2D-CARS measurements displaying the 2D imaging spectrometer system and the alignment of the laser beams. The details of the time-synchronized fs and ps laser systems employed are described in the supplementary material.28 The structured probe beam is formed by passing the probe beam through a mask (4 mm × 4 mm) which is relay imaged to the crossing of the pump/Stokes beam with a 2:1 magnification to increase the probe pulse irradiance. In light blue is the 2D-CARS image from the resolved J = 8 rotational transition. The optical detection system consists of a single “effective” lens used in combination with a diffraction grating to disperse the 2D-CARS light. L1, L2, and L3 – spherical lenses with f1 = 1 m, f2 = 0.5 m, f3 = 0.75 m, respectively. M – mirror, HWP – half wave plate, CL – cylindrical lens (f = 1 m / 0.3 m), BD – beam dump, PBS – polarizing beam splitter cube, SP – short wave pass filter, G – grating (3600 l/mm), CCD – charge coupled device camera, RF – 100 MHz radio frequency source to which both the fs and ps seed lasers are phase-locked.

FIG. 1.

The experimental setup for the 2D-CARS measurements displaying the 2D imaging spectrometer system and the alignment of the laser beams. The details of the time-synchronized fs and ps laser systems employed are described in the supplementary material.28 The structured probe beam is formed by passing the probe beam through a mask (4 mm × 4 mm) which is relay imaged to the crossing of the pump/Stokes beam with a 2:1 magnification to increase the probe pulse irradiance. In light blue is the 2D-CARS image from the resolved J = 8 rotational transition. The optical detection system consists of a single “effective” lens used in combination with a diffraction grating to disperse the 2D-CARS light. L1, L2, and L3 – spherical lenses with f1 = 1 m, f2 = 0.5 m, f3 = 0.75 m, respectively. M – mirror, HWP – half wave plate, CL – cylindrical lens (f = 1 m / 0.3 m), BD – beam dump, PBS – polarizing beam splitter cube, SP – short wave pass filter, G – grating (3600 l/mm), CCD – charge coupled device camera, RF – 100 MHz radio frequency source to which both the fs and ps seed lasers are phase-locked.

Close modal

Single-shot 2D-CARS signals from pure N2 at 295 K in the 40-mm2 probed field, recorded with and without a mask inserted in the probe beam path, are displayed in Fig. 2(a). Note that a frequency scale is not suitable in the employed representation, instead the individual rotational Raman S-branch transitions (JJ + 2) are being indicated with rotational quantum number J. The characteristic features of the rotational spectrum of N2 are evident: equidistant lines separated by ∼4B (=8 cm−1) and the characteristic 4:1 intensity alternation between the odd and even transitions, which is related to the nuclear spin degeneracy for N2. The image quality, quantified in terms of a horizontal and vertical line-spread function, was evaluated to be ∼3.5 pixels full-width at half maximum (fwhm) in the horizontal dimension (∼56 μm) and ∼2 pixels fwhm in the vertical dimension (∼32 μm), respectively, and the total image consists of ∼125 × 120 pixels which provides about 15 000 spectra generated in a single-laser-shot. The vertical extent of the imaged field here is limited only by the choice of mask, as most of the probe beam was reflected away by the mask, and a vertical size 3 to 4 times larger is achievable with the current laser setup. The extra pixel in the width of the horizontal line-spread function is because of the convolution of the image in this dimension with the rotational spectrum. In Fig. 2(b), a mixture of 21% O2 and 79% N2 (air) at 295 K is probed, and a vector diagram is illustrated to demonstrate the principle for extracting spectral information corresponding to a spatial location in the two-dimensional field. The presented spectra have been detected with larger dispersion, achieved by increasing the focal length of the “effective” lens in the detection system, from 0.75 m to 2 m. This is an important tuning ability of the technique, as the experimenter can tune the dispersion higher if needed to separate lines from multiple species in a mixture. Further, a larger dispersion may be used to allow for a larger 2D field to be imaged to the CCD if desired. For example, O2 spectral contributions are being indicated from a 2D-CARS spectrum recorded in air. With the presented dispersion, many of the O2 and N2 lines do not overlap, providing for simple data extraction.

FIG. 2.

(a) 2D-CARS single-shot spectra recorded in pure N2 at room temperature. In the second row of 2D spectra, the center structure of the mask is removed to provide a defined 2D imaging field. (b) Recorded spectra in pure N2 and N2 in a mixture with O2 (Air), detected with a larger dispersion, are shown to indicate the potential for separately analyzing the 2D spatial field in mixtures of species. A vector diagram is used to orientate each spatial location of the measured 2D field. The image consists of ∼125 × 120 pixels which give about 15 000 spectra provided by a single-laser-shot.

FIG. 2.

(a) 2D-CARS single-shot spectra recorded in pure N2 at room temperature. In the second row of 2D spectra, the center structure of the mask is removed to provide a defined 2D imaging field. (b) Recorded spectra in pure N2 and N2 in a mixture with O2 (Air), detected with a larger dispersion, are shown to indicate the potential for separately analyzing the 2D spatial field in mixtures of species. A vector diagram is used to orientate each spatial location of the measured 2D field. The image consists of ∼125 × 120 pixels which give about 15 000 spectra provided by a single-laser-shot.

Close modal

Each of the rotational-level-specific images in Fig. 2(a) was vectorized, and the data from a single height in the image (a single pixel row slice) are presented in Fig. 3. The rotational CARS intensity spectra30 were simulated using a time-domain CARS code similar to that used previously31,32 and the theoretical spectrum for the average evaluated temperature is shown in red. A library of theoretical rotational CARS spectra, generated over a range of temperatures, was fitted to the experimentally extracted signal intensities using a quick-fitting nonlinear interpolating procedure.33 The resulting average evaluated temperature was 299 K, an error in accuracy of 1.5%, and the standard deviation in temperature evaluation across the spectra was 1.4%.

FIG. 3.

Extracted rotational CARS spectra taken from a single height (Y) in the image from Fig. 2(a). The spectra originate from a specific height in the probed 2D plane and cover 20 mm (X) of space at this height. From the intensity spectra, the average temperature was evaluated to be 299 K with a standard deviation of 1.4%, this yields an accuracy error of 1.5%. There are 125 such pixel rows in the single laser shot 2D-CARS image, which each provide another set of data like that presented in this figure.

FIG. 3.

Extracted rotational CARS spectra taken from a single height (Y) in the image from Fig. 2(a). The spectra originate from a specific height in the probed 2D plane and cover 20 mm (X) of space at this height. From the intensity spectra, the average temperature was evaluated to be 299 K with a standard deviation of 1.4%, this yields an accuracy error of 1.5%. There are 125 such pixel rows in the single laser shot 2D-CARS image, which each provide another set of data like that presented in this figure.

Close modal

The development of 2D-CARS will open many new directions for research in the field of combustion diagnostics and pollutant formation. Thermal boundary layers near burner or bluff body surfaces are critical to the understanding of flame stabilization mechanisms, and also provide critical input parameters for the boundary conditions for the numerical simulation of combustors. Current methods for single-shot thermal field measurement, such as 2D Rayleigh scattering measurements, are very sensitive to light scattering and, in general, cannot be employed near surfaces. Detailed comparisons of turbulent heated flows can be compared to computational fluid dynamics (CFD) models for model validation and refinement with the high-precision and accuracy of single-shot CARS thermometry. Combined particle-imaging-velocimetry (PIV) measurements34 and 2D-CARS thermometry may provide simultaneous mapping of the flow- and temperature field. 2D thermal field measurements will be vitally important in the development and refinement of dynamical soot formation models,35,36 and to the understanding of energy dissipation following plasma formation.

Because the presented 2D-CARS technique is a coherent wave-mixing technique, allowing for the selection of wavelengths far from absorption, the beams are not significantly attenuated, and can be reflected back to the probe volume along a slightly displaced 2D imaging plane, and the 2D signal sent to a second CCD. By repeating this process a few 2D planes could be measured in a single laser shot, effectively creating a 3D-CARS measurement over a large field, and the instantaneous thermal field and relative species concentrations could be evaluated in 3D.

Funding provided by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences. Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.

1.
C. L.
Evans
,
X. Y.
Xu
,
S.
Kesari
,
X. S.
Xie
,
S. T. C.
Wong
, and
G. S.
Young
,
Opt. Express
15
,
12076
(
2007
).
2.
J. X.
Cheng
,
Y. K.
Jia
,
G. F.
Zheng
, and
X. S.
Xie
,
Biophys. J.
83
,
502
(
2002
).
3.
M. T.
Bremer
,
P. J.
Wrzesinski
,
N.
Butcher
,
V. V.
Lozovoy
, and
M.
Dantus
,
Appl. Phys. Lett.
99
,
101109
(
2011
).
4.
F.
Vestin
and
P. E.
Bengtsson
,
Proc. Combust. Inst.
32
,
847
(
2009
).
5.
P. R.
Regnier
and
J. P. E.
Taran
,
Appl. Phys. Lett.
23
,
240
(
1973
).
6.
I. R.
Beattie
,
T. R.
Gilson
, and
D. A.
Greenhalgh
,
Nature (London)
276
,
378
(
1978
).
7.
W. B.
Roh
,
P. W.
Schreiber
, and
J. P. E.
Taran
,
Appl. Phys. Lett.
29
,
174
(
1976
).
8.
D. V.
Murphy
and
R. K.
Chang
,
Opt. Lett.
6
,
233
(
1981
).
9.
M.
Alden
,
P. E.
Bengtsson
, and
H.
Edner
,
Appl. Opt.
25
,
4493
(
1986
).
10.
A. C.
Eckbreth
,
Laser Diagnostics for Combustion Temperature and Species
(
Gordon and Breach Publishers
,
1996
).
11.
A.
Bohlin
,
E.
Nordström
,
H.
Carlsson
,
X.-S.
Bai
, and
P.-E.
Bengtsson
,
Proc. Combust. Inst.
34
,
3629
(
2013
).
12.
R. A.
Patton
,
K. N.
Gabet
,
N.
Jiang
,
W. R.
Lempert
, and
J. A.
Sutton
,
Appl. Phys. B: Lasers Opt.
108
,
377
(
2012
).
13.
R. P.
Lucht
,
P. J.
Kinnius
,
S.
Roy
, and
J. R.
Gord
,
J. Chem. Phys.
127
,
044316
(
2007
).
14.
D.
Pestov
,
R. K.
Murawski
,
G. O.
Ariunbold
,
X.
Wang
,
M. C.
Zhi
,
A. V.
Sokolov
,
V. A.
Sautenkov
,
Y. V.
Rostovtsev
,
A.
Dogariu
,
Y.
Huang
, and
M. O.
Scully
,
Science
316
,
265
(
2007
).
15.
J. D.
Miller
,
M. N.
Slipchenko
,
T. R.
Meyer
,
H. U.
Stauffer
, and
J. R.
Gord
,
Opt. Lett.
35
,
2430
(
2010
).
16.
C. C.
Hayden
and
D. W.
Chandler
,
J. Chem. Phys.
103
,
10465
(
1995
).
17.
W. D.
Kulatilaka
,
P. S.
Hsu
,
H. U.
Stauffer
,
J. R.
Gord
, and
S.
Roy
,
Appl. Phys. Lett.
97
,
081112
(
2010
).
18.
J. D.
Miller
,
S.
Roy
,
J. R.
Gord
, and
T. R.
Meyer
,
J. Chem. Phys.
135
,
201104
(
2011
).
19.
A.
Bohlin
,
E.
Nordstrom
,
B. D.
Patterson
,
P.-E.
Bengtsson
, and
C. J.
Kliewer
,
J. Chem. Phys.
137
,
074302
(
2012
).
20.
C. J.
Kliewer
,
A.
Bohlin
,
E.
Nordstrom
,
B. D.
Patterson
,
P. E.
Bengtsson
, and
T. B.
Settersten
,
Appl. Phys. B: Lasers Opt.
108
,
419
(
2012
).
21.
Y.
Gao
,
A.
Bohlin
,
T.
Seeger
,
P.-E.
Bengtsson
, and
C. J.
Kliewer
,
Proc. Combust. Inst.
34
,
3637
(
2013
).
22.
T.
Seeger
,
J.
Kiefer
,
Y.
Gao
,
B.
Patterson
,
C.
Kliewer
, and
T.
Settersten
,
Opt. Lett.
35
,
2040
(
2010
).
23.
C. J.
Kliewer
,
Y.
Gao
,
T.
Seeger
,
J.
Kiefer
,
B. D.
Patterson
, and
T. B.
Settersten
,
Proc. Combust. Inst.
33
,
831
(
2011
).
24.
R. S.
Barlow
,
Proc. Combust. Inst.
31
,
49
(
2007
).
25.
A. C.
Eckbreth
,
Appl. Phys. Lett.
32
,
421
(
1978
).
26.
A.
Bohlin
,
B. D.
Patterson
, and
C. J.
Kliewer
,
J. Chem. Phys.
138
,
081102
(
2013
).
27.
A.
Satija
and
R. P.
Lucht
,
Opt. Lett.
38
,
1340
(
2013
).
28.
See supplementary material at http://dx.doi.org/10.1063/1.4810876 for both experimental details and calculations of the phase-matching limited excitation bandwidth.
29.
B. K.
Ford
,
M. R.
Descour
, and
R. M.
Lynch
,
Opt. Express
9
,
444
(
2001
).
30.
A.
Bohlin
,
P. E.
Bengtsson
, and
M.
Marrocco
,
J. Raman Spectrosc.
42
,
1843
(
2011
).
31.
S. P.
Kearney
and
D. J.
Scoglietti
,
Opt. Lett.
38
,
833
(
2013
).
32.
J. D.
Miller
,
S.
Roy
,
M. N.
Slipchenko
,
J. R.
Gord
, and
T. R.
Meyer
,
Opt. Express
19
,
15627
(
2011
).
33.
A. D.
Cutler
and
G.
Magnotti
,
J. Raman Spectrosc.
42
,
1949
(
2011
).
34.
A.
Braeuer
,
S. R.
Engel
,
R. F.
Hankel
, and
A.
Leipertz
,
Opt. Lett.
34
,
3122
(
2009
).
35.
A.
Krishnan
,
E.
Dujardin
,
M. M. J.
Treacy
,
J.
Hugdahl
,
S.
Lynum
, and
T. W.
Ebbesen
,
Nature (London)
388
,
451
(
1997
).
36.
S. L.
Fiedler
,
S.
Izvekov
, and
A.
Violi
,
Carbon
45
,
1786
(
2007
).

Supplementary Material