A strong anti-Stokes Raman signal, from the vibrational Q(1) transition of hydrogen, is generated in gas-filled hollow-core photonic crystal fiber. To be efficient, this process requires phase-matching, which is not automatically provided since the group velocity dispersion is typically non-zero and—inside a fiber—cannot be compensated for using a crossed-beam geometry. Phase-matching can however be arranged by exploiting the different dispersion profiles of higher-order modes. We demonstrate the generation of first and second anti-Stokes signals in higher-order modes by pumping with an appropriate mixture of fundamental and a higher-order modes, synthesized using a spatial light modulator. Conversion efficiencies as high as 5.3% are achieved from the pump to the first anti-Stokes band.
© 2013 Optical Society of America
In stimulated Raman scattering (SRS), pump photons of frequency are scattered into Stokes photons of frequency where is the Raman frequency shift. This process is accompanied by the excitation of molecules—the scatterers—into a higher rotational or vibrational state. The synchronously oscillating scatterers form a coherence wave with frequency and phase velocity , where and and are the propagation constants of the pump and Stokes waves.
If an excited molecule scatters a pump photon while decaying to the ground state, a blue-shifted anti-Stokes (AS) photon of frequency is generated. This phenomenon can for example be used to generate narrowband radiation in the deep and vacuum ultraviolet spectral region . Furthermore, CARS spectroscopy  and Raman frequency combs—which can span several octaves —also rely on the generation of an AS signal. Phase-coherent frequency combs even provide a method for generating trains of ultra-short pulses [4, 5], which can be especially interesting when pumping with a continuous wave laser [6, 7].
Efficient AS generation—which is crucial for all these applications—in general requires phase-matching, i.e., the wavevector mismatch must be zero. This condition is not automatically satisfied since the group velocity dispersion is typically non-zero. In free space, this mismatch is commonly eliminated using oblique beams . Other techniques include using dispersion-compensating mirrors in a cavity , or birefringent waveguides .
Gas-filled photonic crystal fibers (PCF) provide huge advantages for SRS applications compared to free-space systems, including exceptionally low thresholds , and pressure-controllable dispersion . If one wishes, however, to establish a system based on single-pass propagation through a fiber, neither oblique beams nor dispersive mirrors can be used to achieve phase-matched AS generation.
Here we present a fiber-based method for obtaining phase-matching for efficient AS generation, making use of higher-order modes (HOMs) in a hollow-core photonic crystal fiber (HC-PCF) [11, 12]. This method works because the dispersion profiles of HOMs differ considerably from that of the fundamental LP01 mode. The technique involves arranging a combination of modes (or modal mixtures) in the pump, Stokes and AS signals that approximately satisfies phase-matching, and then fine-tuning for efficient AS generation by adjusting the gas pressure.
2. Experimental set-up and techniques
2.1 Basic set-up
A schematic of the experimental set-up is shown in Fig. 1. The pump source is a 1064 nm microchip laser delivering pulses of duration 1.8 ns (FWHM) at a repetition rate of 1 kHz. Its Gaussian beam is projected onto a spatial light modulator (SLM) , and then coupled into a hydrogen-filled kagomé-PCF. By adjusting the SLM, a controllable mixture of fundamental and HOMs is excited in the fiber. Depending on the experimental conditions (gas pressure, pump energy and pump mode) SRS-induced first and second vibrational AS signals are observed in different HOMs at the fiber output. After passing through appropriate bandpass filters for spectral selection, the signal power is measured and its near-field pattern imaged by a CCD camera.
2.2 Properties of gas-filled kagomé-PCF
The fiber used in the experiment is a 1.2 m long, broadband-guiding kagomé HC-PCF filled with hydrogen gas. A scanning electron micrograph of its structure is shown in Fig. 2. The effective refractive index of the LPpm mode guided in a kagomé-PCF can be written to good accuracy as [10, 14]:10], and ngas the refractive index of the core material. The modal dispersion can be conveniently adjusted by varying the pressure of the filling gas—in this case hydrogen, for which dispersion data is available in the literature . Figure 5 shows the calculated dispersion curves for several of the HOMs in an evacuated kagomé-PCF (frequency versus wavevector difference k0 – βpm, where βpm = k0npm is the modal wavevector and k0 = 2π/λ is the vacuum wavevector).
We make use of the vibrational Raman transition Q(1) of ortho-H2, which provides a Raman frequency shift of (4155 cm−1) . When pumped at 1064 nm this results in a Stokes signal at 1912 nm, a first AS signal at 737 nm and a second AS signal at 564 nm. The process begins once the first Stokes signal has reached threshold. In order to optimize generation of the AS signal, the hydrogen pressure, the pump energy and the pump mode mixture are varied. The highest pressure available in the gas-filling system was 25 bar, and the maximum pump pulse energy coupled into the fiber was 50 µJ.
2.3 Generating modal mixtures with SLM
The electronically-driven liquid crystal array in an SLM changes the local phase of the beam pixel-by-pixel. In order to excite a certain mode mixture in the fiber, its phase pattern is imprinted on the incoming Gaussian beam by means of the SLM [Fig. 3]. However, as the amplitude of the beam cannot be changed with the phase-only SLM used in the experiment, the desired mode mixture is not exactly reproduced in the fiber. Even so, since the field distribution reflected off the SLM is known, the mode mixture excited at the fiber input face can be calculated. This is done by calculating the complex amplitude distribution of the focused beam at the fiber input face using the Fraunhofer (far-field) approximation. Then the overlap between this field and the LPpm fiber mode is determined numerically [10, 14], and the fraction fpm of energy launched into the LPpm mode determined. Note that, when calculating the SLM phase pattern of the mixture of LP01 and LP11 modes, we ignore the radial dependence, retaining only the correct azimuthal phase function. This has the advantage of making the SLM phase pattern independent of the beam radius of the incoming Gaussian beam [Fig. 3(a)].
3. Experimental results
3.1 Basic observations
When the pump light was launched purely in the fundamental mode, no significant AS signal was seen. The situation changed dramatically, however, when the pump beam contained a mixture of fundamental and LP11 modes [Fig. 4(a)]. In this case, a first AS signal was generated in the LP11 mode [Fig. 4(b)], appearing either as a double-lobed mode or a doughnut mode. A second AS signal was observed in a clean four-lobed LP21 mode at pressures below ~20 bar [Fig. 4(c)], gradually switching to the LP11 mode at higher pressures. At 25 bar (the highest pressure available in the gas system) the second AS signal was in a distorted LP11 mode [Fig. 4(d)], which we attribute to the presence of some LP21 mode. As explained later (Section 3.3), at even higher pressures we expect to see a clean LP11 mode. When the pump was changed to a mixture of LP01 and LP02 modes, the AS signal switched to the LP03 mode, as shown in Fig. 4(e).
3.2 Analysis of results
Figures 5(a)–5(c) explain the generation of the AS signals when the pump light is in a mixture of fundamental and LP11 modes. The frequency conversion takes place in three stages. First, a stimulated Stokes signal is generated in the LP01 mode, mediated by the coherence wave marked with a green arrow in Fig. 5(a). This signal then seeds the generation of a second coherence wave [purple arrow in Fig. 5(a)] that converts LP11 pump photons to Stokes photons in the LP01 mode. In the second stage, the first AS is generated [Fig. 5(b)]. No coherence wave is available to drive the red dashed transition towards the first AS in LP01; neither the green nor the purple coherence wave fits. An AS signal can however be generated in the LP11 mode using either the purple coherence wave in conjunction with the pump in the LP01 mode or the green coherence wave in conjunction with the pump in the LP11 mode. Note that this is only possible if the pump signal is simultaneously available in the two modes. If not, none of the two transitions will work because of the lack of either a suitable coherence wave or pump light in an appropriate mode. In the third stage a second AS signal can be generated in the LP21 mode by re-using the purple coherence wave, or in the LP11 by exploiting the green coherence wave [Fig. 5(c)]. Depending on the pressure, conversion to either the LP21 or the LP11 mode is preferred, as explained in Section 3.3.
The appearance of the first AS signal in the LP03 mode can be explained in a similar manner by means of the dispersion curves for modes of higher radial order [Fig. 5(d)]. The pump is a mixture of LP01 and LP02 modes. The Stokes is again generated in the LP01 mode, but this time a coherence wave is generated by scattering from the LP02 pump mode to the LP01 Stokes mode (yellow arrow). This coherence wave is phase-matched for conversion of the LP02 pump to an AS signal in the LP03 mode.
Generation of the AS signals is affected by the core refractive index, which changes with pressure [Eq. (1)]. This results, for example, in switching of the second AS signal from the LP21 mode at low pressure to the LP11 mode at high pressure. In Fig. 6 the phase-mismatch calculated using the dispersion curves is plotted versus the pressure. Efficient AS generation is not only possible when the phase-mismatch is zero, but can also occur for small amounts of phase-mismatch by nonlinear phase-locking [17–19]. This allows the phase-mismatch to be compensated by SRS-induced nonlinear phase-shifts of pump, Stokes and anti-Stokes signals. It is even possible that the AS signal will be stronger in the presence of a slight phase-mismatch because parametric gain suppression can occur at the exact phase-matching point  (gain suppression occurs when the “pump+Stokes” and “anti-Stokes+pump” coherence waves destructively interfere, resulting in loss of Stokes and anti-Stokes signals). As the phase-mismatch increases, however, AS generation via phase-locking becomes more and more inefficient, eventually fading away. Consequently, efficient AS generation is only possible close to the phase-matching pressure.
Note, however, that pressure changes influence not only the dispersion but also the gain of the Raman medium and the dephasing time T2 of the coherence waves . This makes the pressure dependence of the AS signal highly complicated. We now present a qualitative explanation of the observations.
Over the pressure range used in the experiments (1 to 25 bar) the phase-mismatch for the first AS signal [Fig. 6(a)] is considerably larger when it is in the LP01 (blue line) than when it is in the LP03 (red line) or the LP11 mode (green line). This explains the experimental observation that the AS signal appears in the two HOMs. Indeed, the phase-mismatch is small enough for both LP11 and LP03 modes to allow nonlinear phase-locking (as mentioned above), and thus efficient AS generation. Note that in the presence of phase-locking, pressure and phase-mismatch play only secondary roles. Which of the two HOMs is preferentially excited can be controlled by altering the modal mixture of the pump light. If the pump contains a mixture of LP01 and LP02 modes, the AS signal is generated in the LP03 mode. If however it contains the LP01 and the LP11 modes, the AS signal appears in the LP11 mode.
In contrast, the phase-matching conditions for second AS generation [Fig. 6(b)] change considerably with pressure, as does the observed mode. At lower pressures, the phase-mismatch is smaller for the LP21 mode (cyan line) and a four-lobed mode is observed. The higher the pressure, the better is the phase-matching for the second AS in the LP11 mode (purple line). As a result, competition between the LP21 and the LP11 modes is observed at pressures above ~20 bar in the experiments.
3.4 Optimizing the anti-Stokes signal
According to the explanation given in the paragraphs above, it is clear that efficient generation of the first AS signal in the LP11 mode depends strongly on the presence of both LP01 and the LP11 modes at the pump frequency. In order to investigate this further, we used the SLM to vary the ratio of the two pump modes while keeping the total coupled pump energy constant. We define the conversion efficiency as the AS output energy divided by the pump output energy in the absence of SRS; this takes account of coupling and fiber losses at the pump wavelength. The conversion efficiency is plotted in Fig. 7 against the fraction f11 of pump energy in the LP11 mode.
It can be seen in Fig. 7 that conversion to the AS is efficient only for an appropriate pump mode mixture, approaching zero when the pump light is in a pure LP01 or LP11 mode (the non-zero experimental values for f11 = 0 or f11 = 1 can be explained by coupling of a small fraction of the pump energy into undesired modes). The AS conversion efficiency drops drastically at values of f11 greater than 0.6. Since in this range the pump is mainly in the LP11 mode, the gain for the LP01 Stokes signal is considerably reduced . As a consequence the threshold for the generation of Stokes in the LP01 mode is no longer reached, resulting in suppression of the AS signal.
As the pressure is increased from 15 bar (blue curve) to 25 bar (red curve), the conversion efficiency increases considerably—this was explained in Section 3.3. The maximum AS conversion efficiency reached for a fiber length of 1.2 m is 5.3%, obtained for a pump energy of 11 µJ, a pressure of 25 bar and f11 = 0.14. The trend indicates that even higher AS conversion efficiencies should be attainable at higher pressures.
Launching mixtures of fundamental and higher-order modes offers an elegant and efficient means of achieving phase-matched anti-Stokes generation by SRS in gas-filled HC-PCF. First and second AS signals, carried by HOMs, can be generated with conversion efficiencies as high as 5.3% for the first AS. By changing the pressure and the modal mixture of the pump, the first and second AS signals can be manipulated and optimized. The technique is highly versatile and can be used with almost any combination of pump wavelength, Raman-active gas and HC-PCF. It may prove useful for highly sensitive CARS spectroscopy of trace gases.
We would like to thank Dr. Andreas Walser for useful discussions.
References and links
1. V. Wilke and W. Schmidt, “Tunable UV-radiation by stimulated Raman scattering in hydrogen,” Appl. Phys. (Berl.) 16(2), 151–154 (1978). [CrossRef]
2. R. F. Begley, A. B. Harvey, and R. L. Byer, “Coherent anti-Stokes Raman spectroscopy,” Appl. Phys. Lett. 25(7), 387–390 (1974). [CrossRef]
3. F. Couny, F. Benabid, P. J. Roberts, P. S. Light, and M. G. Raymer, “Generation and photonic guidance of multi-octave optical-frequency combs,” Science 318(5853), 1118–1121 (2007). [CrossRef] [PubMed]
4. 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(6021), 1165–1168 (2011). [CrossRef] [PubMed]
5. Y. Y. Wang, C. Wu, F. Couny, M. G. Raymer, and F. Benabid, “Quantum-fluctuation-initiated coherence in multioctave Raman optical frequency combs,” Phys. Rev. Lett. 105(12), 123603 (2010). [CrossRef] [PubMed]
6. J. T. Green, J. J. Weber, and D. D. Yavuz, “Continuous-wave light modulation at molecular frequencies,” Phys. Rev. A 82(1), 011805 (2010). [CrossRef]
7. S. Zaitsu and T. Imasaka, “Phase-matched generation of higher-order continuous-wave coherent Raman sidebands,” Opt. Commun. 285(3), 347–351 (2012). [CrossRef]
8. D. Dimitropoulos, V. Raghunathan, R. Claps, and B. Jalali, “Phase-matching and nonlinear optical processes in silicon waveguides,” in Integrated Photonics Research, paper IThE3 (2004).
9. F. Benabid, J. C. Knight, G. Antonopoulos, and P. St. J. Russell, “Stimulated Raman scattering in hydrogen-filled hollow-core photonic crystal fiber,” Science 298(5592), 399–402 (2002). [CrossRef] [PubMed]
10. J. Nold, P. Hölzer, N. Y. Joly, G. K. L. Wong, A. Nazarkin, A. Podlipensky, M. Scharrer, and P. St. J. Russell, “Pressure-controlled phase matching to third harmonic in Ar-filled hollow-core photonic crystal fiber,” Opt. Lett. 35(17), 2922–2924 (2010). [CrossRef] [PubMed]
11. M. Ziemienczuk, A. M. Walser, A. Abdolvand, A. Nazarkin, and P. St.J. Russell, “Intermodal stimulated Raman scattering in hydrogen-filled hollow-core photonic crystal fiber,” J. Opt. Soc. Am. B 29(7), 1563–1568 (2012). [CrossRef]
12. B. M. Trabold, A. Abdolvand, T. G. Euser, A. M. Walser, and P. St. J. Russell, “Amplification of higher-order modes by stimulated Raman scattering in H2-filled hollow-core photonic crystal fiber,” Opt. Lett. 38(5), 600–602 (2013). [CrossRef] [PubMed]
13. T. G. Euser, G. Whyte, M. Scharrer, J. S. Y. Chen, A. Abdolvand, J. Nold, C. F. Kaminski, and P. St. J. Russell, “Dynamic control of higher-order modes in hollow-core photonic crystal fibers,” Opt. Express 16(22), 17972–17981 (2008). [CrossRef] [PubMed]
14. E. A. J. Marcatili and R. A. Schmeltzer, “Hollow metallic and dielectric wave-guides for long distance optical transmission and lasers,” Bell Syst. Tech. J. 43(4), 1783–1809 (1964). [CrossRef]
15. E. R. Peck and S. Huang, “Refractivity and dispersion of hydrogen in the visible and near infrared,” J. Opt. Soc. Am. 67(11), 1550–1554 (1977). [CrossRef]
16. J. F. Reintjes, “Stimulated Raman and Brillouin Scattering,” in Handbook of Laser Science and Technology, Supplement 2: Optical Materials, M. J. Weber, ed. (CRC Press, 1995).
17. A. Nazarkin, A. Abdolvand, and P. St. J. Russell, “Optimizing anti-Stokes Raman scattering in gas-filled hollow-core photonic crystal fibers,” Phys. Rev. A 79(3), 031805 (2009). [CrossRef]
18. V. S. Butylkin, V. G. Venkin, V. P. Protasov, P. S. Fisher, Y. G. Khronopulo, and M. F. Shalyaev, “Effect of phase locking on the dynamics of the anti-Stokes component of stimulated Raman scattering,” Sov. Phys. JETP 43, 430–435 (1976).
19. V. S. Butylkin, G. V. Venkin, V. P. Protasov, N. D. Smirnov, Y. G. Khronopulo, and M. F. Shalyaev, “Spatially-bounded phase capture and axial anti-Stokes radiation in SRS in gases,” JETP Lett. 17, 285 (1973).
21. M. G. Raymer and J. Mostowski, “Stimulated Raman scattering: Unified treatment of spontaneous initiation and spatial propagation,” Phys. Rev. A 24(4), 1980–1993 (1981). [CrossRef]