Realistic limits of subfemtosecond pulse generation by molecular modulation: dephasing, finite driving pulses, spatial effects, high-intensity nonlinearities, and energy conversion

In 1998, Harris and Sokolov proposed generating subfemtosecond optical pulses by molecular modulation: two phase-locked laser fields, detuned from an electronic resonance but near-resonant with a Raman transition, adiabatically prepare a near-maximal molecular coherence that phase-modulates a propagating field, producing a broad collinear Raman sideband comb with Bessel-function amplitudes and frequency-modulation phases that the same gas's group-velocity dispersion can compress into a train of subfemtosecond pulses (0.21 fs pulses spaced 11.1 fs predicted for deuterium). The quantitative predictions, however, rested on idealizations: no Raman dephasing, monochromatic driving fields of infinite transverse extent, no thermal distribution over rotational/vibrational levels, and no nonlinear processes (multiphoton ionization, nonlinear phase shifts, Raman saturation) at the enormous peak intensities reached after compression, while near-complete conversion of the incident energy into the phased comb was expected but not quantified. The open problem is to determine, theoretically and experimentally, how these realistic effects limit the achievable coherence, sideband comb, compressed pulse duration and contrast, peak intensity, peak power, energy, and conversion efficiency, and to validate whether high-energy subfemtosecond pulse trains are actually produced. Later work resolved substantial parts of the program (near-unity sideband conversion in the rotational variant, self-focusing control, transient and hollow-fiber generation, and single-cycle/sub-single-cycle pulse-train synthesis), but no complete quantitative account or experimental realization of the original self-compressed, high-energy subfemtosecond regime exists.

Background

Attosecond and subfemtosecond optical science needs pulses that are not only short but energetic. High-order harmonic generation, the dominant route to isolated attosecond pulses, converts only a very small fraction (of order $10^{-4}$) of the incident laser energy into the short-wavelength pulse, so attosecond pulse energies remain tiny. A complementary proposal by Harris and Sokolov in 1998 (Physical Review Letters 81, 2894) works in the near infrared and visible: 'molecular modulation'. Two continuous-wave-like laser fields at frequencies $\omega_0$ and $\omega_{-1}$, both far below electronic resonance but with a difference frequency $\omega_m=\omega_0-\omega_{-1}$ close to a Raman transition (a transition driven by the difference of two photon energies rather than by a single photon) of a molecular gas, adiabatically drive the transition. When the driving is adiabatic — the field envelopes vary slowly compared with the separation of the dressed-state eigenvalues — the molecules follow a single eigenstate in which the Raman coherence (the off-diagonal density-matrix element $\rho_{ab}$ between the two vibrational or rotational levels) approaches its maximum $|\rho_{ab}|\approx 0.5$, the regime of 'maximum coherence' related to electromagnetically induced transparency, in which the medium is transparent for the Raman process and all molecules scatter in phase. A propagating field then experiences a sinusoidally varying susceptibility at frequency $\omega_m$, which is exactly the condition for phase modulation: the sidebands generated at $\omega_0+q\omega_m$ acquire Bessel-function amplitudes $J_q(\gamma z)$ and the phases of a frequency-modulated (FM) carrier, with a peak phase deviation $\gamma z$ set by the coherence magnitude, gas density, and interaction length. Because FM spectra of this form have group delays ordered by frequency (blues leading or lagging reds depending on which dressed eigenstate, 'phased' or 'antiphased', is prepared), ordinary group-velocity dispersion (GVD — the frequency dependence of the group velocity, characterized by $\partial^2 k/\partial\omega^2$) of the same gas can rephase the comb into a pulse train, with the compression length fixed by $(\partial^2 k/\partial\omega^2)\,l=\pi/(2\omega_m^2\gamma L)$ for a generation length $L$. A full numerical propagation for the fundamental vibrational transition of deuterium ($\omega_m\approx 2994\ \mathrm{cm^{-1}}$, two-photon detuning $-1$ GHz, $10^{11}\ \mathrm{W/cm^2}$ in each drive, 3.6 cm cell at 1 atm), using published oscillator strengths of the Lyman and Werner bands (the electronic absorption bands of the hydrogen isotopologues), predicted 28 sidebands spanning 150 nm to 1.4 µm compressing to 0.21 fs pulses separated by 11.1 fs, and the authors proposed two diagnostics: sideband-intensity asymmetries that differ between the phased and antiphased prepared states, and nonlinear multiphoton responses (e.g., four-wave mixing — a third-order nonlinear process in which input fields generate a new field at a combination frequency — or ionization) as time-integrated signatures of the peak intensity. The predictions, however, were computed under idealized assumptions that the authors themselves flagged for further study: the model contains no dephasing (decay of $\rho_{ab}$ at the Raman linewidth, of order 0.3 GHz FWHM in D$_2$ at STP, i.e. a coherence time of order 1 ns, which also bounds the time for which nanosecond driving pulses sustain the coherence); the driving fields were treated as monochromatic plane waves, whereas the proposed experiment uses focused ~5-ns, ~50-mJ pulses whose peak power exceeds the self-focusing critical power of the gas (the power above which the intensity-dependent refractive index makes the beam collapse into a filament rather than propagate) and whose finite envelopes, bandwidth, and possible chirp were not modeled; no thermal population over rotational $J$ levels (a room-temperature gas populates many rotational states, so only a fraction of molecules sit on any single Raman line) and no saturation were included; and no nonlinear process at the post-compression intensities (a 0.21 fs pulse carrying the drive energy would exceed $10^{15}\ \mathrm{W/cm^2}$, well above the multiphoton-ionization threshold of the medium) was modeled. The paper also expected, but did not quantify, nearly complete conversion of the incident energy into the phased comb for driving times up to the Raman dephasing time. Which of these idealizations is decisive — and what pulse duration, contrast, peak power, and conversion efficiency actually survive when all of them are lifted — is the question posed here.

Problem Statement

Determine, by quantitative theory and experiment, whether subfemtosecond pulse generation by molecular modulation — adiabatic preparation of near-maximal Raman coherence by two detuned driving fields, generation of a collinear phased sideband comb with Bessel/FM amplitude and phase structure, and rephasing by group-velocity dispersion (the frequency dependence of the group velocity) — survives the realistic conditions excluded from the original 1998 treatment, and with what performance. Specifically: (1) how finite Raman coherence lifetimes (dephasing at the collisional Raman linewidth) and saturation of the Raman transition limit the achievable coherence $|\rho_{ab}|$ and modify sideband formation, for the quasi-continuous nanosecond-pulsed excitation actually available; (2) how finite-duration, finite-bandwidth, and possibly chirped driving pulses, and realistic transverse beam profiles with their self-focusing, filamentation, and transverse mode coupling, change the generation efficiency, the comb phase structure, and the compressed pulse quality; (3) what nonlinear processes (multiphoton ionization of the modulating gas, plasma formation and defocusing, additional nonlinear phase shifts, Raman saturation) are triggered by the extreme instantaneous peak intensities of the compressed pulses, and how they back-act on the spectrum and temporal waveform; (4) how the thermal distribution over rotational/vibrational levels of the modulating gas (a nonzero initial distribution over $J$ states) and depletion of the ground-state population affect the preparation of coherence and the required input powers and detunings; and (5) what fraction of the incident energy is actually converted into the phased comb and ultimately into the compressed pulses under pulsed excitation for times up to the Raman dephasing time, testing the expectation of nearly complete conversion. The question includes the experimental side: generation and characterization of the pulses, establishing the achievable coherence, sideband comb, compressed pulse duration and contrast, peak intensity, peak power, energy, and conversion efficiency, and the parameter regime (molecule, transition, pressure, detuning, drive intensity and duration, geometry) in which the technique does or does not work. The five effects (1)–(5) are mutually coupled under pulsed excitation, so the question is the combined limit together with the experimental characterization.

References

1. S. E. Harris and A. V. Sokolov, 'Subfemtosecond Pulse Generation by Molecular Modulation', Physical Review Letters 81, 2894 (1998), DOI 10.1103/PhysRevLett.81.2894, https://doi.org/10.1103/PhysRevLett.81.2894
2. A. V. Sokolov, D. D. Yavuz, and S. E. Harris, 'Subfemtosecond pulse generation by rotational molecular modulation', Optics Letters 24, 557 (1999), DOI 10.1364/OL.24.000557, https://doi.org/10.1364/OL.24.000557
3. Fam Le Kien, K. Hakuta, and A. V. Sokolov, 'Pulse compression by parametric beating with a prepared Raman coherence', Physical Review A 66, 023813 (2002), DOI 10.1103/PhysRevA.66.023813, https://doi.org/10.1103/PhysRevA.66.023813
4. D. D. Yavuz, D. R. Walker, G. Y. Yin, and S. E. Harris, 'Rotational Raman generation with near-unity conversion efficiency', Optics Letters 27, 769 (2002), DOI 10.1364/OL.27.000769, https://doi.org/10.1364/OL.27.000769
5. D. R. Walker, D. D. Yavuz, M. Y. Shverdin, G. Y. Yin, A. V. Sokolov, and S. E. Harris, 'Raman self-focusing at maximum coherence', Optics Letters 27, 2094 (2002), DOI 10.1364/OL.27.002094, https://doi.org/10.1364/OL.27.002094
6. A. V. Sokolov and S. E. Harris, 'Ultrashort pulse generation by molecular modulation', Journal of Optics B: Quantum and Semiclassical Optics 5, R1 (2003), DOI 10.1088/1464-4266/5/1/201, https://doi.org/10.1088/1464-4266/5/1/201
7. V. P. Kalosha and J. Herrmann, 'Ultrawide spectral broadening and compression of single extremely short pulses in the visible, uv-vuv, and middle infrared by high-order stimulated Raman scattering', Physical Review A 68, 023812 (2003), DOI 10.1103/PhysRevA.68.023812, https://doi.org/10.1103/PhysRevA.68.023812
8. E. Sali, K. J. Mendham, J. W. G. Tisch, T. Halfmann, and J. P. Marangos, 'High-order stimulated Raman scattering in a highly transient regime driven by a pair of ultrashort pulses', Optics Letters 29, 495 (2004), DOI 10.1364/OL.29.000495, https://doi.org/10.1364/OL.29.000495
9. M. Y. Shverdin, D. R. Walker, D. D. Yavuz, G. Y. Yin, and S. E. Harris, 'Generation of a Single-Cycle Optical Pulse', Physical Review Letters 94, 033904 (2005), DOI 10.1103/PhysRevLett.94.033904, https://doi.org/10.1103/PhysRevLett.94.033904
10. S. Sensarn, S. N. Goda, G. Y. Yin, and S. E. Harris, 'Molecular modulation in a hollow fiber', Optics Letters 31, 2836 (2006), DOI 10.1364/OL.31.002836, https://doi.org/10.1364/OL.31.002836
11. W.-J. Chen, Z.-M. Hsieh, S.-W. Huang, H.-Y. Su, C.-J. Lai, T.-T. Tang, C.-H. Lin, C.-K. Lee, R.-P. Pan, C.-L. Pan, and A. H. Kung, 'Sub-Single-Cycle Optical Pulse Train with Constant Carrier Envelope Phase', Physical Review Letters 100, 163906 (2008), DOI 10.1103/PhysRevLett.100.163906, https://doi.org/10.1103/PhysRevLett.100.163906
12. S. Baker, I. A. Walmsley, J. W. G. Tisch, and J. P. Marangos, 'Femtosecond to attosecond light pulses from a molecular modulator', Nature Photonics 5, 664 (2011), DOI 10.1038/nphoton.2011.256, https://doi.org/10.1038/nphoton.2011.256
13. H.-S. Chan, Z.-M. Hsieh, W.-H. Liang, A. H. Kung, C.-K. Lee, C.-J. Lai, R.-P. Pan, and L.-H. Peng, 'Synthesis and Measurement of Ultrafast Waveforms from Five Discrete Optical Harmonics', Science 331, 1165 (2011), DOI 10.1126/science.1198397, https://doi.org/10.1126/science.1198397
14. J. T. Karpel, D. Gold, and D. D. Yavuz, 'Prospects for continuous-wave molecular modulation in Raman-active microresonators', Optics Express 27, 34154 (2019), DOI 10.1364/OE.27.034154, https://doi.org/10.1364/OE.27.034154
15. A. G. Löhr, M. Yu. Ivanov, and M. A. Khokhlova, 'Controlled compression, amplification and frequency up-conversion of optical pulses by media with time-dependent refractive index', Nanophotonics 12, 2921 (2023), DOI 10.1515/nanoph-2022-0818, https://doi.org/10.1515/nanoph-2022-0818