Nonresonant Raman Effects on Femtosecond PumpâProbe with

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Letter

Nonresonant Raman Effects on Femtosecond Pump – Probe With Chirped White Light: Challenges and Opportunities. Itay Gdor, Tufan Ghosh, Oleg Lioubashevski, and Sanford Ruhman J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.7b00559 • Publication Date (Web): 07 Apr 2017 Downloaded from http://pubs.acs.org on April 9, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 14

The Journal of Physical Chemistry Letters

1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nonresonant Raman Effects on Femtosecond Pump – Probe with Chirped White-Light: Challenges and Opportunities. Itay Gdor,1 Tufan Ghosh,1 Oleg Liubashevski, 1 Sanford Ruhman1,* 1

Institute of Chemistry, the Hebrew University, Jerusalem 9190401, Israel.

Abstract Impulsive Raman excitation in neat organic liquids far from resonance is followed using chirped broadband supercontinuum probe pulses. Spectral modulations due to impulsively induced coherent vibrations vary in intensity tenfold as a function of the probe’s linear chirp. Simulations clarify why the vibrational signature is maximized for a Group Delay Dispersion (GDD) in reduced units of 0.5 while a probe GDD of twice that quenches the same

spectral modulations. Accordingly recent claims that chirped white-light probe pulses provide equivalent information on material response as their compressed analogues must be taken with caution. In particular interactions which induce spectral shifts in the probe depend crucially on the arrival chronology of the continuum colors. On one hand this presents limitations to application of chirped continuum radiation as-is in pump-probe experiments. It also presents the opportunity for using this dependence to control the relative amplitude of non-resonant interactions in to pump-probe signals such as that of solvent vibrations.

TOC

Keywords: Molecular vibrations, Impulsive Raman, Chirped pulses. *Corresponding author: [email protected]

ACS Paragon Plus Environment


2 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The effects of GDD on ultrafast pump-probe experiments has been extensively investigated, prompted both by the simplicity of envisioning linear chirp effects on the pump or probe process, as well as the ease with which quadratic spectral phase can be imparted on femtosecond pulses.1,2 For extremely short pulses the main challenge is to perfectly eliminate chirp in order to realize the transform limited duration at a desired point in the experiment.3,4,5 Under resonant conditions, a controlled amount of negative linear chirp was shown to enhance the excitation of ground state vibrational coherence or to minimize population transfer to the excited state.6,7,8,9,10,11 Probe chirp has also been shown under resonant conditions to be sensitive to momentum and not only to location in coordinate space of coherent vibrations.12 Another major incentive for studying probe chirp effects on pump-probe spectroscopy came with the advent of multi-channel transient absorption measurements with dispersed super-continuum pulses. Such pulses are routinely generated by self-phase modulation (SPM) of ultrafast pulses in transparent media. Starting with the work of Mataga and coworkers,13 this approach provides broad band transient spectra of the sample following photoexcitation. It is essential for separating spectral shifts from dipole strength variations in the spectral bands of nascent excited states under study. Since SPM naturally produces predominantly linearly chirped continuum pulses spanning thousands of wavenumbers, it is often impractical to compress these pulses before their application as probes. White light continua so generated can span the visible and extend into the near IR, and when generated with pulses a few tens of femtoseconds in duration, are produced with a GDD of at least 0.5 ps over this range.14 In most ultrafast experiments the target time resolution, defined by the cross correlation of pump and probe is a small fraction of a picosecond. Accordingly a dense coverage of pump-probe delays followed by spectral


Page 2 of 14

Page 3 of 14


3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

reconstruction have been the key to obtaining temporally uniform transient spectra. Transmission changes for different wavelengths are obtained from different temporally chirped spectra. Theoretical justification for this manipulation along with an assessment of the effective time resolution of the resulting spectra was recently put forward by Polli et.al.15,16 A comparison with theory shows that, limited only by the resolution of the recording spectrograph, data collected with and without probe GDD is nearly identical, and that time resolution is not forfeited by working with chirped white light probes without compression. This result is puzzling. The interaction of probe with an evolving sample must involve both dispersive and absorptive interactions, characterized by modulation of both real and imaginary parts of the refractive index. Modulation of the former will induce spectral shifts, which should effect the probe field very differently when the arrival times of its Fourier components is not uniform. Scattering of a compressed probe off a pump-induced coherent Raman active mode will, for instance, be Stokes or anti-Stokes shifted in its entirety depending on pump-probe delay.17,18 In contrast the different colors of a chirped pulse at a single nominal delay will alternate between these extremes depending upon what phase of the vibration they encounter etc. To test this we have chosen the extreme case of nonresonant impulsive Raman in transparent organic liquids where the material response essentially modulates only dispersion. As anticipated, unlike the results obtained on resonance, the observed effect of coherent vibrations is strongly affected by the chirp, altering the observed spectral modulations at a specific probe wavelength by more than an order of magnitude. Furthermore the maximal and minimal modulation depth obtained by this method lie above and below that observed for a transform limited probe. Finally, adaptation of a classical



4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

model for spectral effects of impulsive Raman to chirped probe pulses matches the experimental observations and justifies the qualitative picture suggested above. experimental setup - 30 fs pulses at 790 nm with 0.8 mJ of energy were derived from a homemade multi-pass amplified Ti:Sapph laser system at a 1 KHz repetition rate. One μJ was used to generate a single-filament white light continuum probe in 2 mm of sapphire (typical spectrum shown in Fig. 1). The continuum pulses were collimated and re-focused into the sample with reflective optics. The remaining fundamental was used to seed a one-stage noncollinear optical parametric amplifier (NOPA). The latter produced 2 µJ ultra-broadband pulses (~540-700 nm), the typical spectrum is shown in Fig. 1. These pulses were compressed to within 10% of their transform limit by a BK7 prism pair, followed by a 19channel aluminum coated deformable mirror (OKO Flexible Optical) positioned in the Fourier plane of a 4-f shaper. The resulting 7 fs pulse was focused in a 0.2 mm thick optical cuvette with 0.15 mm quartz windows (Starna Scientific Ltd.) to a spot of 0.1 mm, and overlapped with the probe whose diameters was two times smaller. Transmitted probe pulses were dispersed in an imaging CCD spectrograph and read shot-by-shot. Every other pump pulse was chopped using an optical chopper (Thorlabs Ltd.) and two consecutive readouts (pumpon/pump-off) were used to calculate the optical density difference (∆OD). Chirp control over the probe beam was achieved by adding different width glass windows in the beam path (after the white light generation), the GDD for each scan was characterized by fitting the nonlinear Kerr response across the probing range. All beams were polarized in parallel.


Page 4 of 14

Page 5 of 14


5

Pump Probe

Intensity

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

480

520

560

600

640

λ (nm)

Fig 1. Intensity spectra of the pulses used in the experiments. NOPA pump pulses are depicted in black and supercontinuum probe in red.

Fig. 2 depicts the initial 0.5 ps of a transient absorption spectrum from 450 to 700 nm following excitation of a benzene sample, using the previously described NOPA pulse. As seen in the figure, a short and intense scattering peak is apparent during pump-probe temporal overlap (this signal appears saturated in the color scheme chosen to emphasize the later vibrational coherence). Since we are dealing with a chirped probe, this intense peak appears at different delay times for each wavelength, providing a measurement of probe GDD which was quantified by fitting this peak position to a linear function against frequency. Since the time axis reflects an increased relative path-length added to the probe arm, the early appearance of overlap in the blue end of the probe spectrum reflects its relative delay in arrival.



Page 6 of 14

6

∆mOD

500

6.000

400 3.000

Time (fs)

300 0.000

200 100

-3.000

0 -6.000

500

550

600

650

λ (nm) Fig. 2: ΔOD map following excitation of the benzene sample measured with a chirped probe pulse.

Following the intense electronic response the ΔOD map does not decay to zero. Instead much fainter periodic modulations appear as ripples in the color map. These are assigned to impulsively induced Raman motions in the liquid. Not surprisingly the phase-front of the vibrational ripples follows the shape of the probe chirp. Taking a temporal cut at a particular delay TD after the excitation results in a sinusoidal signal which shifts in phase as TD is varied. Changing the chirp produces more interesting effects, as can be seen in Fig. 3, where the frequencies of the sinusoidal and, more importantly, the amplitude of this signal change. GDD;

0.75

0.9

1

1.3

12

∆mOD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

8

4

0

500

550

600

650

λ (nm)


Page 7 of 14


7 Fig. 3: temporal cuts measured 0.5 ps after excitation of a benzene sample taken with different probe chirp. The plots are offset in the Y axis to facilitate their comparison. A cut of the ΔOD map in Fig. 2 at a particular probe wavelength provides a record of the free induction decay of the induced vibrations as seen in panel A of Fig. 4. A Fourier transform of this signal will result in the Raman spectrum of benzene, with clear, strong vibrations at ca. 990 cm-1 associated with ring breathing. Since the dephasing time of these modulations is very long this process is not covered in our experiments, producing an artificially broadened Raman line-shape which reflects the range of pump-probe delay recorded.

550 nm

50

∆OD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

FFT Power

0 3

-50

0 -3

-100

800

0

400

900

800

Time (fs)

1200 800

900

1000

1100

1200

-1

Frequency (cm )

Fig. 4: (A) spectral cut at 550 nm in the map presented in Fig. 1 showing vibrational induced intensity modulation of the probe. (B) Power of FFT derived from modulations in transient transmission depicted in (A). In order to extend our investigation of chirp effects on the intensity of the impulsive Raman signals, similar experiments were conducted on dichloromethane (DCM). Unlike benzene, the impulsive Raman spectrum in DCM contains two strong vibronic bands; CCl2 scissoring at 280 cm-1 and CCl2 symmetric stretch at 700 cm-1. Accordingly time cuts such as those shown in figure 3 for DCM do not show a perfect sinusoidal modulation, more than one vibration contributes to the resulted spectrum. Fig. 5 presents spectral cuts at 550 nm for different chirps, in which, as for benzene, chirp modifies the intensity of the vibration. As two vibrations are active with a large frequency gap between them, one can clearly see how the



Page 8 of 14

8 ratio between their intensities changes as the chirp increases. This is obvious from the Fourier analysis of these modulations presented on the right side of figure 5. As expected, the amplitude of modulations either in time or in frequency were found to depend linearly on pump fluence over one order of magnitude, changing from 30-300 µJ/cm2.

-1

GDD = 0.5 (700 cm )

CH2Cl2

0 3

-40

0 -3

-80

800

1000 -1

GDD = 0.8 (700 cm )

FFT Power

0

∆mOD

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3

-40

0

-80

-3 800

1000 A

-1

GDD = 1 (700 cm )

0 2

-40

0 -2

-80

800

0

500

1000 A

1000

Time (fs)

1500

200

400

600

800 -1

Frequency (cm )

Fig. 5: Spectral cuts at 550 nm of DCM data at 3 different chirp values along with Fourier analysis obtained from the observed modulations. In order to follow the effect of probe chirp on the depth of spectral modulation, the FFT amplitude is plotted in figure 6 for both liquids as a function of probe GDD in reduced units of where phase, time and frequency are based on the vibrational period. As can be seen

the modulation amplitude peaks when the reduced GDD is 0.5±N and is minimized when the chirp equals 0±N where N is an integer. For reasons which remain to be determined the first maximum at GDD=0.5 as more pronounced than those which follow. None the less this trend of maxima and minima is conserved over the tested range.


Page 9 of 14


9 1.50

-1

-1

C6H6(1000 cm )

Normalized amplitude

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

CH2Cl2(285 cm )

-1

CH2Cl2(695 cm )

1.25

Eq 4

1.00 0.75 0.50 0.25 0.00 0.0

0.5

1.0

1.5

2.0

2.5

Reduced GDD

Fig. 6: FFT amplitude as a function of reduced GDD for the 1000 nm vibration of benzene (black), 695 nm vibration of DCM (red), and 285 band of DCM (blue). Prediction of the model (Eq 4) plotted in grey. To account for these results a simple classical theory of impulsive Raman spectroscopy due to Yan et.al.17 is extended to include probe pulse chirp. Equation 17 in reference 17 presents the delay dependent spectrum of a transform limited probe field in the presence of a coherent Raman active vibration in the sample which was initiated by a co-propagating ultrafast pump. After notational simplification the resulting probe field can be written: , + exp + − exp − − 1

is the frequency of the Raman active vibration, and TD the probe pulse delay with respect to the pump. The Gaussian probe field !"

#$ %$

, B is &⁄'( *+′-.′/ where Q/

is the vibrational mode amplitude, N is the number density of oscillators, α the differential polarizability tensor, ∆=(ω−ωL) where ωL is the central probe frequency, l is the sample thickness and τ the FWHM of the probe pulse. Ignoring homodyned scattering, the intensity spectrum of the probe is proportional to the absolute value squared of the field, giving:



Page 10 of 14

10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 , ∝ 1 + 2 !" 4

− 5 > 7 cos ;exp

Nonresonant Raman Effects on Femtosecond PumpâProbe with

Recommend Documents

Nonresonant Raman Effects on Femtosecond PumpâProbe with