Professional Documents
Culture Documents
Received 24 January 2018; revised 9 April 2018; accepted 9 April 2018 (Doc. ID 320533); published 8 May 2018
Backward stimulated Raman scattering in gases provides a promising route to the compression and amplification of a
Stokes seed pulse by counter-propagating against a pump pulse, as has been demonstrated already in various plat-
forms, mainly in free space. However, the dynamics governing this process when seeded by noise has not yet been
investigated in a fully controllable collinear environment. Here we report, to the best of our knowledge, the first
unambiguous observation of efficient noise-seeded backward stimulated Raman scattering in a hydrogen-filled hol-
low-core photonic crystal fiber. At high gas pressures, when the backward Raman gain is comparable to, but lower
than, the forward gain, we report quantum conversion efficiencies exceeding 40% to the backward Stokes at 683 nm
from a narrowband 532 nm pump. Efficiency increases to 65% when the backward process is seeded by a small amount
of back-reflected forward-generated Stokes light. At high pump powers, the backward Stokes signal, emitted in a clean
fundamental mode and spectrally pure, is unexpectedly always stronger than its forward-propagating counterpart. We
attribute this striking observation to the unique temporal dynamics of the interacting fields, which cause the Raman
coherence (which takes the form of a moving fine-period Bragg grating) to grow in strength toward the input end of
the fiber. A good understanding of this process, together with the rapid development of novel anti-resonant-guiding
hollow-core fibers, may lead to improved designs of efficient gas-based Raman lasers and amplifiers operating at
wavelengths from the ultraviolet to the mid-infrared. © 2018 Optical Society of America under the terms of the OSA
Open Access Publishing Agreement
OCIS codes: (060.5295) Photonic crystal fibers; (140.3550) Lasers, Raman; (290.5910) Scattering, stimulated Raman.
https://doi.org/10.1364/OPTICA.5.000570
material Raman gain and frequency shift of any gas, and is trans- (it diverges for very low pressures as it does not include the
parent down into the ultraviolet. Using narrowband pulses of Doppler-broadened linewidth as a limit). The first term on the
3.2 ns duration (physical length 96 cm) and a 114 cm length right-hand side in Eq. (1) describes the contribution from
of fiber with a core diameter of 22 μm filled with 38 bar of hydro- Dicke narrowing and the second term describes the contribution
gen, the noise-seeded backward Stokes signal (at 683 nm) was from collisional broadening. While Eq. (1) is not directly appli-
always stronger than the forward, reaching efficiencies exceeding cable to the backward case, it has been experimentally observed
40%. Even higher backward conversion efficiency (65%) was ob- that the backward Raman linewidth, Δνbr , also follows a similar
tained when a 14 cm length of fiber was used—in this case, the pressure dependence [18], although no fitting formula has been
pump pulse is substantially longer than the fiber, so that the back- reported. Collisional broadening in H2 starts to dominate over
ward Stokes process could be seeded by a small fraction of the Dicke narrowing at pressures above 3 bar for Δνfr and 31 bar
forward Stokes signal, back-reflected at the window of the gas cell. for Δνbr [18], the difference being attributed to the strong influ-
To understand the in-fiber temporal dynamics of backward ence of Doppler broadening in BSRS. However, beyond 31 bar,
Raman scattering, we recorded the temporal profiles of all inter- Δνbr increases linearly with pressure, whereas Δνfr –Δνbr decreases
acting sidebands along with their energy and compare the results and R → 1.
with numerical simulations based on a bi-directional set of The steady-state FSRS gain [cm/W] of the fundamental
Maxwell–Bloch equations [14,15,16]. This allowed us to relate vibrational transition of H2 , ΩR ∼125 THz (4155 cm−1 ), is given
the dynamics of the BSRS process to the evolution of forward by [20]
and backward Raman coherence waves (i.e., synchronous molecu- 9.37 × 106 52ρ∕Δνfr K B ∕0.658 × νP − 4155
lar oscillations driven by the beat-note created by the pump and γf , (2)
Stokes signals), which turns out unexpectedly to favor the onset 7.19 × 109 − ν2P 2
and amplification of the backward signal, even though the actual where K B 0.658 is the Boltzmann population factor at 298 K
gain coefficient for FSRS is higher than that for BSRS. for the J 1 rotational level in H2 [20]. Equation (2) is valid
The paper is organized as follows: in Section 2 we will briefly over the range of 1–100 amagats, depending as it does on Eq. (1).
revise general aspects of the Raman gain in hydrogen gas. In In the simulations, we used FSRS gain values from Eqs. (1)
Section 3, we present the experimental setup employed to dem- and (2). The BSRS gain was obtained using γ b Rγ f . At our
onstrate self- and noise-seeded BSRS in a HC-PCF, followed by a working pressure (38 bar), we used R 0.7 based upon exper-
discussion of the experimental and numerical results (Section 4). imental measurements of the forward and backward Raman
In Section 5, we discuss the key role of the interplay between linewidths [18]. This is justified since the 1064 nm parent laser
forward and backward coherence waves in the dominance of emits pulses with FWHM duration of ∼3.8 ns with a linewidth
BSRS over FSRS at high gas pressures. Finally, in Section 6 we of ∼130 MHz—parameters similar to those used in [18].
summarize the conclusions of the present study and provide an Moreover, at 38 bar, Δνfr ∼1.8 GHz, which is an order of mag-
outlook of potential implications of the results across different nitude larger than the laser linewidth.
fields.
3. EXPERIMENTAL SETUP
2. RAMAN GAIN IN MOLECULAR HYDROGEN The experimental setup is sketched in Fig. 1(a) and 1(b). A
The steady-state material Raman gain for the fundamental vibra- frequency-doubled Nd:YAG laser operating at a repetition rate
tional frequency shift of H2 is lower for BSRS than for FSRS. This of 3 kHz delivered linearly polarized 532 nm pump pulses with
asymmetry originates mainly from a difference in Raman line- FWHM duration of ∼3.2 ns (corresponding a pulse length of
width [2]. Moreover, it has also been predicted [2] and observed ∼96 cm). Two kagomé-style HC-PCFs with different core diam-
experimentally [6,17] that the efficiency of the BSRS process eters and fiber lengths were employed [see Fig. 1(c) for scanning
drops with increasing pump linewidth. In particular, when the electron micrographs (SEMs) of the fiber structures]. In the first
Raman linewidth of the FSRS process exceeds the pump line- set of experiments, we used a 14-cm-long fiber with a core
width (a condition fulfilled in our experiments), the BSRS/ diameter of ∼47 μm (Fiber 1). In the second set, we used a
FSRS gain ratio can be approximated as R ∼ Δνfr ∕Δνbr , where 114-cm-long fiber with a core diameter of ∼22 μm (Fiber 2).
Δνfr and Δνbr are, respectively, the forward and backward The first fiber was placed inside a 16-cm-long monolithic gas cell,
Raman scattering linewidths [2]. At room temperature, for pres- whereas the second fiber was placed in an arrangement of two
sures far below 1 bar, Δνfr and Δνbr are strongly influenced 8-cm-long gas cells connected by a ∼1-m-long tube. Both fibers
by Doppler broadening, which is more pronounced for BSRS, were filled with 38 bar of hydrogen—the highest pressure attain-
and R < 0.13 [18]. As the pressure increases, the Doppler- able with the gas system.
broadened Raman linewidths for both processes shrink under In all the experiments, the launch efficiency of the pump
the influence of Dicke narrowing [18]. With further increase pulses was ∼70%, limited by a mismatch between the spatial pro-
in pressure, collisional broadening starts to influence both Δνfr files of the focused pump light and the fundamental core mode. A
and Δνbr , eventually dominating at high enough pressure. A sim- dichroic mirror (DM1) transmitting the first vibrational Stokes
ple formula, based on fits to experimental data at 298 K, captures line at 683 nm and fully reflecting the 532 nm pump line was
the pressure dependence of the forward Raman linewidth [19]: placed in front of the in-coupling lens. Beams emerging at both
ends of the fiber were collimated and diverted to different detec-
309
Δνfr 51.8ρ, (1) tors. The energies of the pump and the generated forward and
ρ backward Raman pulses were measured with a calibrated power
where ρ is the density in amagats (1 bar corresponds to 0.9 ama- meter using customized band-pass filters. An optical spectrum
gats at 298 K) and the equation is valid from 1 to 100 amagats analyzer was used to measure the spectral content of the output
Research Article Vol. 5, No. 5 / May 2018 / Optica 572
FIAO Experiment
for normal incidence and an angled holder for 40° incidence. On BS 4.7µJ
4
these holders, 3-mm-thick MgF2 windows were mounted in
FS 7.4 µJ
such a way that the distance between the fiber output end and 2
RP 15.6µJ RP 17.4µJ FS 8.4 µJ
Simulation
6 FIFO FIAO
To demonstrate and unambiguously distinguish self- and noise- BS 4.0 µJ
seeded generation of the backward Stokes signal, two main gas- 3 RP 11.4µJ FS 8.2 µJ RP 12.7µJ FS 11.8 µJ
cell configurations were employed for both fibers [see Fig. 1(a)]:
BS 3 nJ
(a) a flat input window and a flat output window (FIFO), and 0
tf -6 tf tf +6 tf -6 tf tf +6
(b) a flat input window and an angled output window t (ns) t (ns)
(FIAO). Furthermore, for the experiments with the short fiber, Fig. 2. [(a), (b)] Experimental and [(c), (d)] simulated temporal pro-
an additional configuration with an angled input window and files of the generated sidebands for the FIFO (left) and FIAO (right) con-
a flat output window (AIFO) was employed. figurations in Fiber 1 filled with 38 bar of H2 . The measured energies for
When the uncoated flat output window is in place, a fraction the initial pump (IP), residual pump (RP), first forward Stokes (FS), and
of the noise-seeded forward Stokes pulse diverging from the backward Stokes (BS) are indicated in the panels.
Research Article Vol. 5, No. 5 / May 2018 / Optica 573
Power (kW) 12 4 BS
BS 29.3 µJ BS
0.4
2
FS RP
RP 22.2 µJ
6 0 0
0 1 2 3 4 5 6 7
FS 2.5 µJ Pump Energy (µJ)
(b) 8 (d) 1.6
Simulation Experiment IP
Power (kW)
tf -6 tf tf +6 BS
t (ns) 4 0.8 BS
The depletion of all the lines shown in Fig. 4(d) is due to the 800
appearance of a second forward Stokes signal at 953.6 nm
(not shown), where fiber loss is high (see Supplement 1 for a loss 700
measurement). We believe the measurements represent the
Frequency (THz)
ΩR
first unambiguous observation of efficient noise-seeded BSRS 600
in a hydrogen-filled HC-PCF.
ΩR
The spectra of the forward- and backward-propagating lines 500
for ∼24 μJ of pump energy in Fiber 2 are shown in Fig. 5.
The forward-propagating signal shows several Raman lines [see 400
Fig. 5(a)], generated from both vibrational and rotational SRS.
These may be explained by the long interaction length for 300
-20 -10 0 10 20
co-propagating pump and Stokes/anti-Stokes lines, along with (1/µ m)
a shallow dispersion landscape that permits interactions with
different forward coherence waves (which, among other effects, Fig. 6. Dispersion diagram for the forward and backward LP01 -like
modes of the HC-PCF. The solid blue arrows represent the coherence
allows generation of up-shifted anti-Stokes signals via molecular
waves involved in the various different SRS transitions. Backward
modulation [21–24]). In contrast, the counter-propagating anti-Stokes generation is very strongly dephased, as expected. FSRS is
Stokes light is concentrated solely at 683 nm, corresponding also dephased, but by a much smaller degree (too small to be seen on
to the first vibrational Stokes line [see Fig. 5(b)]. the plot). See text for more details.
As already mentioned, this behavior may be understood by
reference to the dispersion diagram in Fig. 6, which plots the
frequency-dependent propagation constants, βq , of the forward-
(q f ) and backward- (q b) propagating fundamental modes For efficient anti-Stokes generation, the momentum compo-
of Fiber 2 for 38 bar of H2 , based on the modified Marcatili– nent of a coherence wave must be closely similar for Stokes and
Schmelzer model (see Supplement 1). The solid arrows represent anti-Stokes processes. The dephasing rate can be written as
q q
the four vectors of the coherence waves (C w s) generated by ϑq j2β − βAS − βS j, which works out at 3.1 rad/cm for FSRS
5
and 4.8 × 10 rad∕cm for BSRS, corresponding, respectively,
interference between the pump and both the forward and
backward Stokes and anti-Stokes signals: to dephasing lengths π∕ϑq of 1 cm and 65 nm. This strongly
q q q q
favors the generation of a forward frequency comb, while very
C wS β − βS , ΩR , C wAS βAS − β, ΩR : (3) strongly suppressing the generation of a backward comb. This
explains the simplicity of the observed backward spectra and
the strength of the backward Stokes signal.
Furthermore, the backward Stokes signal is always emitted
0
(a) Vib. AS1 Pump Vib. FS1 in a clean fundamental core mode [see insets in Fig. 5(b) for a
Vib. FS2
-10 (435.7 nm) (532 nm) (683 nm)
(953.6 nm) far-field profile]. Very strong dephasing also means that BSRS
-20
is not impaired by coherent Raman gain suppression [15] (see
Supplement 1).
Normalized spectral power density (dB/nm)
-30
(a) 4
Pump
backward interaction length means that the backward Stokes sig-
nal is unable to generate its own Stokes lines [see Fig. 5(b)], with
2
the result that the Stokes energy is concentrated within a narrow
t (ns)
0 spectral band, and also has the consequence that fiber losses play
-2
only a minor role, unlike for forward SRS.
-4
(b) 4
Forward
6. CONCLUSIONS
Stokes
In conclusion, backward Stokes light can be efficiently generated
t (ns)
2 8
Intensity (TW/m2)
in a short length of gas-filled HC-PCF in a simple arrangement
0 6
-1 involving only a monolithic gas cell. The system allowed observ-
(c) 4 4 ing, for the first time, unambiguously noise-seeded BSRS in a
Backward
Stokes 2 hydrogen-filled HC-PCF, with quantum efficiency of 41%.
t (ns)
2
0
When the BSRS is self-seeded by back-reflected forward Stokes
0 light, the quantum efficiencies can be as high as 65%, even
-1
-200 -100 0 4 8 12 100 200 though the forward Raman gain of the gas is higher than the back-
Position (cm) ward gain. The efficiency of the effect will be even higher for
Fig. 7. Simulated spatio-temporal evolution of (a) the pump, (b) for- pumping in the ultraviolet, when the Raman gain is much
ward Stokes, and (c) backward Stokes signals in FIFO configuration. The stronger [20]. The backward Stokes light is spectrally very narrow
simulation parameters correspond to those used in Fig. 2(c). The dashed and has a high-quality LP01 -like mode profile, while the backward
horizontal lines mark the time when the backward Stokes signal attains its Raman coherence is concentrated close to the input face of the
peak intensity. Note that the spatial scale inside the fiber is magnified fiber, preventing the generation of higher-order Stokes sidebands.
for clarity. Together with recent developments in hollow-core fiber technol-
ogy [25], the results pave the way to a new generation of fiber-
based Raman lasers and amplifiers, ideal for operation in other-
the pump by BSRS close to the fiber input reduces the pump wise difficult-to-access spectral regions such as the ultraviolet
power available for FSRS along the rest of the fiber. Finally, [26,27] and the mid-infrared [28–30]. The HC-PCF system also
FSRS depletes the residual pump light, gradually weakening provides a novel platform for ultrafast pulse compression and am-
the BSRS gain along the fiber. plification in counter-propagating geometries [31].
The remarkable agreement between the experimental mea-
surements and numerical simulations makes it possible to obtain See Supplement 1 for supporting content.
additional information about the system by examining the com-
peting dynamics of the forward and backward coherences.
REFERENCES
Figure 7 shows the simulated spatio-temporal evolution of the
pump, and forward and backward Stokes signals for the param- 1. R. R. Jacobs, J. Goldhar, D. Eimerl, S. B. Brown, and J. R. Murray, “High‐
eters used in the FIFO configuration in Fig. 2(c). The causal se- efficiency energy extraction in backward‐wave Raman scattering,” Appl.
Phys. Lett. 37, 264–266 (1980).
quence of scattering events may be traced clearly. Conversion to 2. J. R. Murray, J. Goldhar, D. Eimerl, and A. Szoke, “Raman pulse com-
the backward Stokes signal takes place within a very narrow space- pression of excimer lasers for application to laser fusion,” IEEE J.
time window, close to the entrance face of the fiber where the Quantum Electron. 15, 342–368 (1979).
backward coherence wave is strongest (see Fig. 8). In contrast, 3. M. Maier, W. Kaiser, and J. A. Giordmaine, “Backward stimulated Raman
scattering,” Phys. Rev. 177, 580–599 (1969).
the forward coherence (Fig. 8) peaks in the second half of the 4. J. O. White, “High-efficiency backward Stokes Raman conversion in
fiber, extending from ∼8 cm to ∼14 cm (the secondary forward deuterium,” J. Opt. Soc. Am. B 7, 785–789 (1990).
coherence peak, centered at ∼1.5 ns, is caused by reflection of 5. V. V. Akulinichev, V. A. Gorbunov, and E. G. Pivinskii, “Competition
the backward Stokes signal at the input of the fiber). The short between forward and backward stimulated Raman scattering in gases,”
Quantum Electron. 27, 427–432 (1997).
6. K. Sentrayan, A. Michael, and V. Kushawaha, “Intense backward Raman
lasers in CH4 and H2,” Appl. Opt. 32, 930–934 (1993).
Coherence (x100) 7. G. I. Kachen and W. H. Lowdermilk, “Relaxation oscillations in stimulated
0.4 0.8 1.2
0 1.6 Raman scattering,” Phys. Rev. A 16, 1657–1664 (1977).
4 8. P. Rabinowitz, A. Kaldor, R. Brickman, and W. Schmidt, “Waveguide H2
Raman laser,” Appl. Opt. 15, 2005–2006 (1976).
9. A. J. Berry, D. C. Hanna, and D. B. Hearn, “Low threshold operation of a
2 waveguide H2 Raman laser,” Opt. Commun. 43, 229–232 (1982).
t (ns)
waves in a H2-filled hollow-core photonic crystal fiber,” Optica 2, 536–539 23. A. V. Sokolov, D. D. Yavuz, and S. E. Harris, “Subfemtosecond pulse
(2015). generation by rotational molecular modulation,” Opt. Lett. 24, 557–559
14. A. Abdolvand, A. Nazarkin, A. V. Chugreev, C. F. Kaminski, and P. St.J. (1999).
Russell, “Solitary pulse generation by backward Raman scattering in 24. M. K. Mridha, D. Novoa, S. T. Bauerschmidt, A. Abdolvand, and P. St.J.
H2-filled photonic crystal fibers,” Phys. Rev. Lett. 103, 183902 (2009). Russell, “Generation of a vacuum ultraviolet to visible Raman frequency
15. S. T. Bauerschmidt, D. Novoa, and P. St.J. Russell, “Dramatic Raman comb in H2-filled kagomé photonic crystal fiber,” Opt. Lett. 41, 2811–2814
gain suppression in the vicinity of the zero dispersion point in gas-filled (2016).
hollow-core photonic crystal fiber,” Phys. Rev. Lett. 115, 243901 25. C. Wei, R. J. Weiblen, C. R. Menyuk, and J. Hu, “Negative curvature
(2015). fibers,” Adv. Opt. Photon. 9, 504–561 (2017).
16. M. Ziemienczuk, A. M. Walser, A. Abdolvand, and P. St.J. Russell, 26. M. Cassataro, D. Novoa, M. C. Günendi, N. N. Edavalath, M. H. Frosz,
“Intermodal stimulated Raman scattering in hydrogen-filled hollow- J. C. Travers, and P. St.J. Russell, “Generation of broadband mid-IR and
core photonic crystal fiber,” J. Opt. Soc. Am. B 29, 1563–1568 UV light in gas-filled single-ring hollow-core PCF,” Opt. Express 25,
(2012). 7637–7644 (2017).
17. I. V. Tomov, R. Fedosejevs, D. C. D. McKen, C. Domier, and A. A. 27. F. Belli, A. Abdolvand, W. Chang, J. C. Travers, and P. St.J. Russell,
Offenberger, “Phase conjugation and pulse compression of KrF-laser “Vacuum-ultraviolet to infrared supercontinuum in hydrogen-filled pho-
radiation by stimulated Raman scattering,” Opt. Lett. 8, 9–11 (1983). tonic crystal fiber,” Optica 2, 292–300 (2015).
18. J. R. Murray and A. Javan, “Effects of collisions on Raman line profiles of 28. A. V. Gladyshev, A. F. Kosolapov, A. N. Kolyadin, M. S. Astapovich, A. D.
hydrogen and deuterium gas,” J. Mol. Spectrosc. 42, 1–26 (1972). Pryamikov, M. E. Likhachev, and I. A. Bufetov, “Mid-IR hollow-core silica
19. W. K. Bischel and M. J. Dyer, “Temperature dependence of the Raman fibre Raman lasers,” Quantum Electron. 47, 1078–1082 (2017).
linewidth and line shift for the Q(1) and Q(0) transitions in normal and 29. A. V. Gladyshev, A. F. Kosolapov, M. M. Khudyakov, Y. P. Yatsenko,
para-H2,” Phys. Rev. A 33, 3113–3123 (1986). A. N. Kolyadin, A. A. Krylov, A. D. Pryamikov, A. S. Biriukov, M. E.
20. W. K. Bischel and M. J. Dyer, “Wavelength dependence of the absolute Likhachev, I. A. Bufetov, and E. M. Dianov, “4.4-μm Raman laser based
Raman gain coefficient for the Q(1) transition in H2,” J. Opt. Soc. Am. B 3, on hollow-core silica fibre,” Quantum Electron. 47, 491–494 (2017).
677–682 (1986). 30. A. Benoît, B. Beaudou, M. Alharbi, B. Debord, F. Gérôme, F. Salin, and
21. K. Hakuta, M. Suzuki, M. Katsuragawa, and J. Z. Li, “Self-induced phase F. Benabid, “Over-five octaves wide Raman combs in high-power
matching in parametric anti-Stokes stimulated Raman scattering,” Phys. picosecond-laser pumped H2-filled inhibited coupling Kagome fiber,”
Rev. Lett. 79, 209–212 (1997). Opt. Express 23, 14002–14009 (2015).
22. S. E. Harris and A. V. Sokolov, “Subfemtosecond pulse generation by 31. N. J. Fisch and V. M. Malkin, “Generation of ultrahigh intensity laser
molecular modulation,” Phys. Rev. Lett. 81, 2894–2897 (1998). pulses,” Phys. Plasmas 10, 2056–2063 (2003).