Open Access
Add to BibSonomyAbstract
We report on a mid-IR optical parametric oscillator (OPO)-based high resolution transient absorption spectrometer for state-resolved collisional energy transfer. Transient Doppler-broadened line profiles at λ = 3.3 μm are reported for HCl R7 transitions following gas-phase collisions with vibrationally excited pyrazine. The instrument noise, analyzed as a function of IR wavelength across the absorption line, is as much as 10 times smaller than in diode laser-based measurements. The reduced noise is attributed to larger intensity IR light that has greater intensity stability, which in turn leads to reduced detector noise and better frequency locking for the OPO.
© 2014 Optical Society of America
1. Introduction
Recent developments in IR light sources are providing new opportunities for precise and sensitive measurements of transient molecular processes. Continuous-wave (CW) mid-IR optical parametric oscillators (OPOs) based on quasi-phase-matching of nonlinear crystals are especially useful for molecular spectroscopy based on their high spectral resolution (<1 MHz) [1,2], broad tunability [3,4], and stable, high-power output (>1 W). The use of OPOs as a source of IR light has existed for some time, but recent advances in the development of quasi-phase-matched crystals such as periodically-poled lithium niobate (PPLN) crystals and high pump power sources such as fiber lasers [5–7] have led to commercially available OPO devices [8]. A number of groups have taken advantage of these features of mid-IR OPOs for studies in the areas of frequency combs [9,10], laser-cooled atoms [11] and gas-surface chemistry [12]. Here, we describe the use of a mid-IR OPO for high-resolution transient IR absorption spectroscopy to investigate energy transfer in molecular collisions.
High-resolution transient mid-IR absorption spectroscopy is a powerful technique for studying the dynamics of non-equilibrium chemical processes [13–20]. The ideal IR light for such measurements is stable, high-resolution, broadly tunable and sufficiently powerful to access molecular transitions that have a broad range of oscillator strengths. Transient IR absorption yields detailed quantum state-resolved information on nascent species, thereby revealing the microscopic mechanisms for a wide variety of chemical phenomena including chemical reactions, cluster formation and inelastic collisions.
In the recent past, lead-salt diode lasers and color-center lasers have been the IR sources of choice for high-resolution transient IR absorption spectroscopy because they meet many of the requirements for transient detection [13–22]. They provide high-resolution output (<10 MHz) that is often narrow enough for measuring Doppler-broadened line profiles of individual ro-vibrational transitions, their output wavelengths are tunable to individual molecular transitions, and they are amenable to active feedback stabilization schemes to control the output frequency. However, they have a number of disadvantages that impact their effectiveness. Lead-salt diode lasers have relatively low output powers (~200 µW or less), have reduced performance at wavelengths shorter than λ = 4 μm and are tunable over limited ranges (~200 cm−1 or less) with non-continuous tuning. Color-center lasers have larger output powers (~5 mW) and more continuous tuning capability, but require complex optical set-ups for single-mode tuning and it has become difficult to obtain functioning crystals. Using a mid-IR OPO alleviates problems associated with low power and limited tuning range, and offers enhanced spectral and intensity stability at many IR wavelengths.
In this paper, we describe a frequency-controlled mid-IR OPO-based transient absorption spectrometer designed to measure collisional energy transfer to HCl molecules. Feedback control enables controlled frequency tuning of the IR light for periods of one hour or more. Ro-vibrational transitions of HCl at λ = 3.3 μm are used to monitor individual states of HCl that scatter from highly vibrationally excited pyrazine (C4H4N2) molecules. The ability to measure molecular processes in the IR near λ = 3 μm opens an important spectral window for accessing the high frequency stretching modes of many environmentally-relevant species. Our initial attempts to measure these processes with lead salt diode lasers were plagued by low signal to noise ratios (SNRs) due to amplitude instability and low IR power. While lead salt diode lasers have been used successfully in many studies at wavelengths of λ = 4 μm and longer, the λ = 3.3 μm spectral region is outside their optimal operating range. Subsequent measurements with the OPO yielded transients with much improved SNRs. Here we compare transient signals, line profiles and noise characteristics for HCl R7 transitions using the two IR sources and illustrate how the improved SNRs enable detailed features of a transient line profile to be distinguished using the HCl R4 transition.
2. Transient absorption spectroscopy of non-equilibrium molecular systems
The scheme for measuring collisional energy transfer with high-resolution transient IR absorption spectroscopy in the gas phase involves three overall steps. In the first step, the energy-donors, in this case pyrazine molecules, are excited with pulsed λ = 266 nm light from a Nd:YAG laser. Radiationless decay converts the photon energy into vibrational energy of the donor molecules [23]. The second step involves excited donor molecules colliding under low pressure conditions with the energy-acceptors, in this case HCl molecules; the collisions induce population changes in individual rotational states of HCl identified by the quantum number J. The time between collisions is much longer than the instrument response time so that nascent populations are measured at early times following optical excitation of donors. In the third step, the population changes in individual J-states of scattered HCl molecules are measured as a function of IR wavelength and J-quantum number using transient IR absorption at λ = 3.3 μm for R-branch transitions (ΔJ = 1) of the vibrational fundamental (Δv = 1).
Two types of measurements are made with transient IR absorption. In the first, the nascent translational energy distributions are determined from the Doppler-broadened line profiles measured at short times relative to the time between collisions. The transient line profiles are measured by locking the IR light to a fringe of a scanning Fabry-Perot (FP) etalon and collecting transient signals stepwise at a series of IR wavelengths (30-50 steps) over the IR transition. Each transient signal is the average of 50-80 UV laser pulses corresponding to data collection times of 1-2 minutes. The data collection time for an entire line profile is an hour or more. As such, these measurements require continuous, precise control of the IR frequency over an extended time period. In the second type of measurement, population changes at line center ν0 of an HCl ro-vibrational transition are monitored by locking the IR light to the peak of the absorption line and collecting transient absorption at a single IR wavelength. Population distributions and rate constants are determined by combining line center and line width measurements for a number of rotational states.
3. An OPO-based transient IR absorption spectrometer
The OPO transient IR absorption spectrometer is described here. The OPO (Aculight Argos, module C) contains a MgO:PPLN crystal that is pumped by 1064 nm light from a Ytterbium-doped fiber laser (Koheras, 5 mW, linewidth < 100 kHz) and amplifier (IPG Photonics, 5W). The pump energy is split by the OPO to generate a tunable signal (λ = 1.46 to 1.60 µm) beam and tunable idler (λ = 3.2 to 3.9 µm) beam with linewidth < 1 MHz. The idler output is split into reference and sample beams; the reference beam is used for active feedback control of the IR wavelength and the sample beam is used for transient detection of HCl.
Using a CW light source for high-resolution transient IR absorption spectroscopy of collisional energy transfer entails setting up active feedback to control the output wavelength and synchronizing that control with the pulse that initiates the transient event, in our case the 266 nm laser pulse. The active feedback system for wavelength control was designed and constructed in our laboratory as part of the transient IR absorption spectrometer. Figure 1 shows an overview of the feedback schemes for the OPO spectrometer and the lead salt diodespectrometer. For precise wavelength control of the OPO output, the IR optical frequency is modulated at 100 Hz by dithering the 1064 nm pump wavelength by means of a piezo-mounted mirror (PZT) in the pump laser. For the diode laser (Laser Photonics) spectrometer, the IR wavelength is modulated at 1 kHz by dithering the diode current using the commercial current controller (Laser Photonics). For both IR sources, the modulated IR light passes through a tunable FP etalon (or reference gas cell) and is collected with phase-sensitive detection. An error signal from a lock-in amplifier is fed back into the pump laser (for the OPO) or the current controller (diode) to lock the IR frequency to an etalon fringe or the peak of a molecular transition. The frequency and phase of the IR modulation are synchronized to the firing of the pulsed laser so that transient absorption is measured at a well-defined IR frequency within the modulation cycle. Similar modulation amplitudes were used for locking both IR sources.
Fig. 1 Overview of schemes for modulation and phase-sensitive detection of (a) the mid-IR OPO and (b) the lead-salt diode laser as part of a mid-IR transient IR absorption spectrometer.
The frequency-controlled OPO output is used in a transient IR absorption spectrometer, as shown in Fig. 2. A master timing pulse coordinates the firing of the 10 Hz UV pulse (to excite pyrazine molecules) and the IR frequency control loop. The timing scheme in the transient spectrometer is inherently coupled with the energy transfer system under investigation and the use of a long path absorption cell. For the pyrazine(E)/HCl system, increasing the path length to a 3-m collision cell enhances the detection sensitivity for IR absorption of HCl, but reduces the gas-pumping capacity of the vacuum system. This is turn leads to a reduced repetition rate for optical excitation of pyrazine to 1 Hz. The 100 ms interval between 10 Hz Nd:YAG pulses is insufficient to completely relax high energy pyrazine molecules (E = 38000 cm−1) through collisions (at 15 mTorr) or to completely refresh the 3-m collision cell. Thus, to avoid multi-pulse excitation of pyrazine, a shutter reduces the repetition rate of the UV excitation pulses to 1 Hz. For shorter sample cells and/or studies at lower donor energies, higher repetition rates are likely to be acceptable.
Fig. 2 Schematic diagram of a transient IR absorption spectrometer that use a frequency-locked mid-IR OPO to investigate molecular energy transfer.
A digital pulse generator (Stanford Research Systems) serves as the master clock that coordinates the firing of the Nd:YAG laser with the IR modulation cycle. The master timer provides a 100-Hz pulse train for the IR frequency control loop and a synchronized 10-Hz pulse train for the Nd:YAG firing. The 100-Hz pulse triggers a voltage ramp from a digital function generator (Stanford Resarch Systems, with a 90:10 duty cycle) that provides the dither voltage to the PZT amplifier of the OPO pump laser. For transient line profile measurements, the IR light is locked to the top of a fringe of the scanning FP etalon (finesse = 8, FSR = 0.01 cm−1) so that the IR frequency overlaps an HCl transition. Typical modulation amplitudes are ± 0.002 cm−1 (or less) relative to the center frequency of the fringe. The modulated IR reference beam is collected with an InSb detector, the output of which is input to a lock-in amplifier (Stanford Research Systems, using a 1 ms time constant, or less) and proportional integrator (Laser Photonics). The error signal from the lock-in amplifier is added to the dither voltage for the pump laser PZT in a homebuilt summing junction. The error signal adjusts the idler-wavelength to maintain a lock on the etalon fringe, thus completing the servo loop and controlling the IR output frequency. A dc voltage (Thorlabs, 0-150 V) is added in the summing junction to provide a constant offset voltage that prevents the PZT amplifier from accidental input of negative voltage from the modulation ramp.
The dc voltage also allows for broad IR tuning (with the servo loop open) in order to locate transitions of interest. For coarse (open loop) tuning of the OPO idler-wavelength, the crystal position is manually translated and the angle of an intracavity etalon is adjusted to locate the spectral region of interest. The finest tuning of the OPO is achieved by adjusting the dc voltage to the fiber laser PZT amplifier with a tuning ratio of 95.2 V/cm−1.
Transient Doppler-broadened line profile measurements are acquired stepwise with the IR light locked to the scanning FP etalon. Each transient is the average of 50-80 UV pulses from the Nd:YAG laser. Once a transient signal has been acquired at one wavelength, the etalon cavity length is tuned by computer-controlled rotation of an internal CaF2 plate and the locked IR light tracks the etalon to a new IR frequency where the next transient signal is collected. This procedure is repeated until 30-50 transient signals have been collected across the HCl line profile. For transient measurements at line center of a transition, the FP etalon is replaced with a reference gas cell containing HCl and the IR transition is used for locking.
The transient signals reported here were collected by passing the sample IR beam through a 3-m cell containing a 15 mTorr, 1:1 mixture of pyrazine and HCl vapor. The UV and IR beams were overlapped collinearly over the length of the cell and then separated using dichroic beamsplitters. The delay between the IR modulation and the 1-Hz, 266 nm pulse was set so that the UV pulse arrived at the center of the IR modulation cycle. A monochromator grating (Laser Photonics) was used to remove any residual/stray UV light from the transmitted IR. An InSb detector (Judson) with a 300 ns risetime collected the transient signals and a digital oscilloscope (LeCroy 9304A) recorded and averaged transient signals (50-80 shots/average) from the sample detector following the UV pulses.
4. Characterizing signal-to-noise levels for transient IR absorption.
Transient measurements of pyrazine/HCl energy transfer were collected on the same spectrometer using two different IR sources: a mid-IR OPO and a lead salt diode laser. Here, we compare data from the two instruments to characterize the overall performance. Transient signals for HCl (v = 0, J = 7) are shown in Fig. 3(a) for the OPO and in Fig. 3(b) for the diodelaser. Both transient signals show similar rates for the appearance of HCl in J = 7 when normalized to UV power and sample pressure. The transients are fit to a flat pre-trigger and an exponential rise for t>0. The residuals are shown in the lower plots. The root-mean-square (rms) noise (based on the residuals) is a factor of 5 smaller for OPO-based data, corresponding to signal levels with SNRs that are 5 times larger for the OPO data. At 15 mTorr total pressure, the average time between collisions is 5 μs and data at t = 1 μs corresponds to the single-collision regime. The averaged signals at t = 1 µs have a SNR = 17.5 for the OPO compared to a SNR = 3.5 for the diode laser.
Fig. 3 Transient absorption and kinetic fitting for the HCl R7 transition following collisions with vibrationally hot pyrazine collected with (a) the OPO-based spectrometer and (b) the diode laser spectrometer. The rms noise from the residuals is shown in lower plots.
The improved SNR for the OPO-based data has direct impact on detection sensitivity for molecules involved in transient events. Based on a minimum SNR = 2 at t = 5 μs or earlier, the OPO-based spectrometer (with a 3-m path length) can detect as few as 2 × 109 HCl molecules cm−3 per quantum state, compared to 1 × 1010 HCl molecules cm−3 per quantum state for the diode-based spectrometer. The minimum number density that can be detected depends on the IR oscillator strength as well as the spectrometer characteristics. For comparison, the oscillator strength for the strong IR absorption for CO2 at 2349 cm−1 is a factor of 13 larger than that for HCl. Correspondingly, we estimate that the CO2 detection limit for this band is on the order of 1 × 108 molecules cm−3 per quantum state for a 3-m path length [24].
The enhanced detection sensitivity of the OPO-based instrument allows for higher quality measurements of transient line profiles since there is better SNR in the Doppler-broadened wings near baseline. Figure 4 shows two Doppler-broadened transient absorption line profiles for HCl (v = 0, J = 7) based on transient absorption at t = 1 µs, one using the OPO [Fig. 4(a)] and the other using the diode laser [Fig. 4(b)]. The fractional absorption intensity at each IR wavelength is determined from a fit to the transient signal (as in Fig. 3) and the wavelength-dependent intensity data are fit with a Gaussian function, with residuals shown in the lower plots. The full width at half maximum line width Δν of the Gaussian fit yields the nascent translational temperature for appearance Tapp of the scattered molecules. Both measurements yield similar translational temperatures, but the rms residuals are a factor of two smaller for the OPO data. Based on the Gaussian fits, the line width uncertainty is approximately 10 times less for the OPO data.
Fig. 4 Doppler-broadened transient absorption line profiles collected at t = 1 μs for the HCl R7 transition measured with the a) OPO and b) diode laser spectrometers. Each data set is fit with Gaussian function. The OPO data have smaller residuals.
5. Noise characteristics
The overall noise in an absorption spectrometer results from noise in the IR source, detectors and electronics [25]. The same liquid-nitrogen cooled detectors and electronics were used for both spectrometers but differences in the IR sources and in their operation affect the overall noise. Here we consider how the overall noise of both spectrometers is affected by IR intensity, IR wavelength and the quality of the frequency lock. For these measurements, the OPO intensity is a factor of 10 greater than for the diode, which reduces shot noise for the OPO. The diode laser is modulated at 1 kHz (compared to the 100 Hz OPO modulation) which reduces the 1/f noise for the diode laser [26]. The spectral resolution of the OPO (Δνopo <1 MHz or 3 × 10−5 cm−1) is a factor of ten higher than that of the diode laser (Δνdiode~10 MHz or 3 × 10−4 cm−1), but the resolution of both light sources is more than sufficient to measure HCl line profiles at 298 K (Δν298 = 6 × 10−3 cm−1).
The power-dependent rms noise for fractional absorption of R7 is shown in Fig. 5 for two IR wavelengths: at line center and at baseline in the wings of the absorption line profile. With the OPO wavelength tuned to baseline, the rms noise (blue circles) decreases linearly with incident IR intensity. A measurement with the OPO at line center (blue triangle) fits extremely well with the detuned data. The diode laser power was insufficient for an extensive power-dependent measurement but a pair of wavelength-dependent measurements shows several informative results. The rms noise with the diode laser detuned from line center (red circle) is much greater (by a factor of 2.8) than for the detuned OPO. For the diode laser tuned to line center (red triangle), the rms noise increases by an additional factor of 1.9. This result shows that the diode laser spectrometer data has additional noise beyond detector noise and that this additional noise source is frequency-dependent.
Fig. 5 IR power dependence of rms noise for transient absorption of HCl R7. The noise with the OPO detuned from transition center (blue circles) decreases linearly with increasing IR power. A measurement at line center (blue triangle) is consistent with the noise for the detuned light, showing a high degree of wavelength stability. In contrast, the noise for the diode laser spectrometer is a factor of 3 larger for detuned IR and a factor of 5 larger at line center than for comparable OPO power.
A more detailed comparison of the wavelength-dependent noise for the two spectrometers is reported here, based on transient absorption measurements across the R7 spectral line. The upper plots of Fig. 6 show the transient noise at two wavelengths for each IR source: Fig. 6(a) is at baseline and Fig. 6(b) is at line center. In both cases, the rms noise (in brackets) for the diode-based data is larger than for the OPO data. At baseline, the diode-based noise is greater by more than a factor of 5 and at line center, the diode:OPO noise ratio increases to almost 8. The wavelength dependence of the rms noise across the line shape is shown in Fig. 6(c). The Gaussian fit to the transient line shape (from Fig. 4) is included as a wavelength reference but its intensity is not to scale. The rms noise for the OPO data is essentially constant for the wavelength range shown, and has an average value of 3.5 × 10−4. The rms noise for the diode-based data is larger at all wavelengths: the average rms noise is1.5 × 10−3 at baseline and as large as of 3.4 × 10−3 at line center. The frequency-dependent transient absorption noise for the diode-based data is consistent with the larger intensity fluctuations of the diode laser, as shown in Fig. 6(b).
Fig. 6 Noise at (a) baseline and (b) line center of HCl R7 with the OPO and diode laser spectrometers. Rms noise values are listed in parentheses. (c) The wavelength dependence of rms noise for the R7 line profile measured with the OPO (red circles) and the diode laser (grey squares). The Gaussian profile from Fig. 4 is shown for reference (intensity is not to scale). The noise amplitude for the diode spectrometer is larger than that of the OPO spectrometer and shows much more scatter across the line profile.
The noise of both instruments was characterized using fast Fourier transform analysis of transient signals (from Fig. 3). The results are consistent with the data in Fig. 6. The transient noise power spectra for HCl R7 are shown in Fig. 7 for both IR spectrometers at(a) IR baseline and (b) line center. The plots show that 1) the noise amplitude for both spectrometers decreases with increasing frequency; 2) at base line and line center, the diode-based data have larger amplitude noise than the OPO-based data; 3) the noise amplitude for the OPO data is nearly the same at baseline and line center and 4) the noise for the diode data is larger at line center than at base line.
Fig. 7 The power spectrum of the transient noise at (a) baseline and (b) line center of the HCl R7 transition for the OPO and diode laser spectrometers. The noise amplitude decreases with increasing frequency for both IR sources. The diode-based noise has greater amplitude than the OPO system and increases at line center.
6. Measurements of double-Gaussian transient line profiles
The improved sensitivity and spectral characteristics of the OPO-based spectrometer allow detection of fine details of transient Doppler-broadened line profiles. The energy transfer kinetics of low-J HCl states that are thermally populated at 300 K involve two simultaneous processes: appearance of population into state J due to collisions and depletion of initial population in that state, also due to collisions [20]. Positive-going appearance signals are broadened by collisional energy transfer and combine with negative-going depletion signals of thermal HCl molecules (at 300 K). Each subpopulation has a unique Doppler-broadened line profile associated with its velocity distribution. The Gaussian distributions of the two subpopulations combine to yield overall transient line profiles that have double Gaussian structure, as shown in Fig. 8(a). Such profiles are seen for HCl in rotational states with J<6. Additional IR stability and resolution is required to distinguish these features.
Fig. 8 Doppler-broadened transient absorption line profiles for the HCl R4 transition collected with the (a) OPO and (b) diode laser spectrometers. The OPO-based instrument has sufficient stability and frequency resolution to reveal the double Gaussian nature of the line profile. These features are obscured by noise in the diode-based instrument.
Doppler-broadened transient line profiles for the HCl R4 transition (collected at t = 1 μs) are shown in Fig. 8. The double Gaussian line shape is clearly seen with the OPO spectrometer [Fig. 8(a)], whereas noise obscures the diode-based profile [Fig. 8(b)]. The OPO data are fit (unconstrained) by the sum of two Gaussian profiles, yielding nascent translational temperatures for appearance Tapp and depletion Tdep, along with time-dependent intensities for each Gaussian. Attempts to fit an unconstrained double Gaussian function to the diode-based data were unsuccessful. Instead, the line profile fitting results for the OPO-based data are used in a constrained fit for the diode-based data. The result is shown in Fig. 8(b), highlighting the capabilities of the OPO-based spectrometer in measuring details of molecular dynamics.
The enhanced SNR for the OPO-based instrument has important benefits for transient absorption measurements: detection of smaller population changes with higher SNRs, shorter averaging times and reduced pressure conditions that are beneficial in minimizing secondary collisions that interfere with nascent measurements. These features combine to enable more sensitive and detailed measurements of transient processes in molecules and chemical reactions.
7. Conclusion
We have incorporated a mid-IR OPO into a transient IR absorption spectrometer and demonstrated the advantages of this instrument for measuring state-resolved collisional energy transfer. Compared to diode-laser measurements at λ ≈3.3 µm, the OPO-spectrometer exhibits reduced noise and enhanced detection sensitivity. The comparison of the noise characteristics reveals that the reduced noise of the OPO-based spectrometer is attributed to smaller amplitude noise in IR intensity and greater IR intensity. The increased spectral resolution and sensitivity of the OPO spectrometer enable measurements of fine details of transient IR line profiles that provide chemical dynamics information.
Acknowledgments
The authors thank Profs. Thomas E. Murphy and John T. Fourkas for helpful discussions on feedback loops and noise analysis. We also gratefully acknowledge research support from DOE BES DE-FG02-06ER15761. M.D.S. was a Beckman Scholar and received research support from the Arnold and Mabel Beckman Foundation.
References and links
1. I. Ricciardi, E. De Tommasi, P. Maddaloni, S. Mosca, A. Rocco, J.-J. Zondy, M. De Rosa, and P. De Natale, “A narrow-linewidth optical parametric oscillator for mid-infrared high resolution spectroscopy,” Mol. Phys. 110(17), 2103–2109 (2012). [CrossRef]
2. I. Ricciardi, E. De Tommasi, P. Maddaloni, S. Mosca, A. Rocco, J.-J. Zondy, M. De Rosa, and P. De Natale, “Frequency-comb-referenced singly-resonant OPO for sub-Doppler spectroscopy,” Opt. Express 20(8), 9178–9186 (2012). [CrossRef] [PubMed]
3. E. Andrieux, T. Zanon, M. Cadoret, A. Rihan, and J. J. Zondy, “500 GHz mode-hop-free idler tuning range with a frequency-stabilized singly resonant optical parametric oscillator,” Opt. Lett. 36(7), 1212–1214 (2011). [CrossRef] [PubMed]
4. J. Courtois, R. Bouchendira, M. Cadoret, I. Ricciardi, S. Mosca, M. De Rosa, P. De Natale, and J.-J. Zondy, “High-speed multi-THz-range mode-hop-free tunable mid-IR laser spectrometer,” Opt. Lett. 38(11), 1972–1974 (2013). [CrossRef] [PubMed]
5. L. E. Myers and W. R. Bosenberg, “Periodically poled lithium niobate and quasi-phase-matched optical parametric oscillators,” IEEE J. Quantum Electron. 33(10), 1663–1672 (1997). [CrossRef]
6. P. Gross, M. E. Klein, T. Walde, K. J. Boller, M. Auerbach, P. Wessels, and C. Fallnich, “Fiber-laser-pumped continuous-wave singly resonant optical parametric oscillator,” Opt. Lett. 27(6), 418–420 (2002). [CrossRef] [PubMed]
7. A. Henderson, R. Stafford, and J. H. Miller, “Continuous wave optical parametric oscillators break new spectral ground,” Spectroscopy 20, 16–18 (2005).
8. A. Henderson, R. Stafford, and P. Hoffman, “High-power CW OPOs span the spectrum,” Laser Focus World 44, 65–69 (2008).
9. F. Adler, P. Masłowski, A. Foltynowicz, K. C. Cossel, T. C. Briles, I. Hartl, and J. Ye, “Mid-infrared Fourier transform spectroscopy with a broadband frequency comb,” Opt. Express 18(21), 21861–21872 (2010). [CrossRef] [PubMed]
10. M. Vainio, M. Merimaa, and L. Halonen, “Frequency-comb-referenced molecular spectroscopy in the mid-infrared region,” Opt. Lett. 36(21), 4122–4124 (2011). [CrossRef] [PubMed]
11. P. G. Mickelson, Y. N. M. de Escobar, P. Anzel, B. J. DeSalvo, S. B. Nagel, A. J. Traverso, M. Yan, and T. C. Killian, “Repumping and spectroscopy of laser-cooled Sr atoms using the (5s5p)3P2-(5s4d)3D2 transition,” J. Phys. At. Mol. Opt. Phys. 42(23), 235001 (2009). [CrossRef]
12. B. L. Yoder, R. Bisson, and R. D. Beck, “Steric effects in the chemisorption of vibrationally excited methane on Ni(100),” Science 329(5991), 553–556 (2010). [CrossRef] [PubMed]
13. G. W. Flynn and R. E. Weston Jr., “Hot atoms revisited: laser photolysis and product detection,” Annu. Rev. Phys. Chem. 37(1), 551–585 (1986). [CrossRef]
14. D. J. Nesbitt, “High-resolution infrared spectroscopy of weakly bound molecular complexes,” Chem. Rev. 88(6), 843–870 (1988). [CrossRef]
15. P. F. Bernath, “High resolution infrared spectroscopy of transient molecules,” Annu. Rev. Phys. Chem. 41(1), 91–122 (1990). [CrossRef]
16. R. E. Weston Jr and G. W. Flynn, “Relaxation of molecules with chemically significant amounts of vibrational energy: the dawn of the quantum state resolved era,” Annu. Rev. Phys. Chem. 43(1), 559–589 (1992). [CrossRef]
17. P. B. Davies, “High resolution tunable infrared laser spectroscopy of transient molecules,” Annu. Rep. Prog. Chem. Sect. C: Phys. Chem. 89, 89–110 (1992). [CrossRef]
18. M. S. Elioff, M. Fang, and A. S. Mullin, “Methylation effects in state resolved quenching of highly vibrationally excited azabenzenes (Evib~38500 cm−1). I. Collisions with water,” J. Chem. Phys. 115(15), 6990–7001 (2001). [CrossRef]
19. E. M. Miller, L. Murat, N. Bennette, M. Hayes, and A. S. Mullin, “Relaxation dynamics of highly vibrationally excited picoline isomers (E(vib) = 38 300 cm(-1)) with CO2: the role of state density in impulsive collisions,” J. Phys. Chem. A 110(9), 3266–3272 (2006). [CrossRef] [PubMed]
20. D. K. Havey, Q. Liu, Z. Li, M. Elioff, M. Fang, J. Neudel, and A. S. Mullin, “Direct determination of collision rates beyond the Lennard-Jones model through state-resolved measurements of strong and weak collisions,” J. Phys. Chem. A 111(13), 2458–2460 (2007). [CrossRef] [PubMed]
21. J. V. V. Kasper, C. R. Pollock, R. F. Curl Jr, and F. K. Tittel, “Computer control of broadly tunable lasers: conversion of a color center laser into a high resolution laser spectrometer,” Appl. Opt. 21(2), 236–247 (1982). [CrossRef] [PubMed]
22. H. K. Haugen, W. P. Hess, and S. R. Leone, “Application of semiconductor diode lasers to probe photodissociation dynamics,” AIP Conf. Proc. 160, 566–568 (1987). [CrossRef]
23. I. Yamazaki, T. Murao, T. Yamanaka, and K. Yoshihara, “Intramolecular electronic relaxation and photoisomerization processes in the isolated azabenzene molecules pyridine, pyrazine, and pyrimidine,” Faraday Discuss. Chem. Soc. 30, 395–405 (1983). [CrossRef]
24. L. A. Pugh and K. N. Rao, “Intensities from Infrared Spectra,” in Molecular Spectroscopy, Modern Research: Volume II, K. N. Rao, ed. (Academic Press, 1976), pp. 165–227.
25. P. Werle and F. Slemr, “Signal-to-noise ratio analysis in laser absorption spectrometers using optical multipass cells,” Appl. Opt. 30(4), 430–434 (1991). [CrossRef] [PubMed]
26. D. S. Bomse, A. C. Stanton, and J. A. Silver, “Frequency modulation and wavelength modulation spectroscopies: comparison of experimental methods using a lead-salt diode laser,” Appl. Opt. 31(6), 718–731 (1992). [CrossRef] [PubMed]
