When methanol microdroplets of 15mm size are irradiated by intense femtosecond laser pulses of moderate intensities (<2×1016W cm-2), we observe a ‘red-light flash’ from the microplasma. We report on the presence of a large ‘blue shoulder’ (that extends to about 200 nm from the incident laser wavelength) in the scattered spectra that corresponds to the red-light flash. A prepulse is found to be essential for producing the large blue shift, which is attributed to the rapid subsequent ionization of a near-critical density preplasma when the main pulse is incident.
© 2006 Optical Society of America
Matter irradiated by intense laser pulses is rapidly ionized when the laser electric field exceeds the Coulomb field experienced by the valence electrons. Laser interactions with solids produce near solid-density plasmas with temperatures of up to tens of keV. Such hot plasmas form a bright source of x-rays, energetic electrons and high energy ions. Advances in laser technology over the past two decades have enabled vastly increased intensities to be delivered, resulting in an enormous increase in the temperatures that can be generated in laboratory plasmas . New schemes for controlling the plasma characteristics and optimizing the photon or particle emission respectively have received much attention lately [2, 3]. Alongside the rapid technological developments  in laser-matter interaction research, much effort has also been devoted to tailoring the ‘matter’ that absorbs the laser energy . Use of novel targets, rather than atoms in the gas phase or plain solid slabs, has recently come under scrutiny. Nanoclusters, for example, have a solid-like density but their spatial extent is much smaller than the wavelength of the incident light, resulting in interaction dynamics very different from intense laser interactions with bulk matter. For example, light absorption is very efficient and production of energetic particles can be much more effective than in bulk solid targets at similar intensities. The questions of primary interest are: how small the target material should be to behave differently from bulk matter, and how the advantages provided by such mesoscopic targets can be best exploited.
Microdroplets have already attracted some interest in this context , but the physics of the light-matter interaction is still largely unexplored with these targets. Applications based on these systems to produce bright and relatively debris-free sources of EUV (extreme ultra violet radiation) and x-rays are only just beginning to be examined . In our own recent studies we have demonstrated that the droplets can be very efficient in producing hard x-rays as compared to those obtained from the irradiation of plain slab targets [8, 9]. For example, moderate intensities of 1014–1015 W cm-2 are adequate for copious generation of the hard x-rays from 15µm methanol droplets. Not only is hot electron generation possible at much lower intensities, the yield or the brightness of the source can be a few orders of magnitude larger than that observed with solid target under similar laser irradiance. The hot electron generation depends sensitively on the presence of a appropriate preplasma produced by a weak pulse that arrives a few 10’s of ns ahead of the main pulse .
In these droplet experiments, the x-ray generation is always accompanied by a ‘red light flash’ that is generated when the main laser pulse in incident on the droplet plasma. In experimental measurements, this red light flash can be used to adjust the laser focusing conditions and the alignment of the laser beam on the droplet. In this paper, we report on the spectral measurements of this red flash and show that very large spectral blue shifts in the scattered light are responsible for its generation. A prepulse generates a preplasma. Measurements  show that the droplet is expanded to 30 µm diamter, when the main pulse arrives. The resulting near-critical-density plasma sphere very efficiently absorbs the main pulse leading to rapid further ionization. A small fraction of the laser light that is scattered through this microplasma exhibits very large blue shifting. Spectral blue shifting is much more common in the studies of optical field ionization of gases. Earlier studies on the ionization induced blue shifting through gas plasmas at 1016 W cm-2 have shown peaks that are shifted by about 20–30 nm from the incident wavelength [10, 11]. Similar blue shifts have also been observed in cluster plasmas [12, 13]. As we demonstrate in this report, at an order of magnitude lower intensity with the droplet target, we observe blue shifts that extend to more than 200 nm. Our experiments also show that the blue shifting of the laser and the hot electron generation are correlated. Rapid and efficient ionization of the droplet material appears to precede hot electron generation.
The experimental details are elaborated in our recent work  and only the salient features are presented here. The droplets are generated by pressurizing methanol (≃20 atm backing pressure) through a 10 µm capillary that is attached to a piezoelectric crystal, which is dithered at 1 MHz, to produce steady stream of 15µm droplets. The inset in Fig. 1(a) shows an image of the droplet stream. The droplet stream is injected into a differentially pumped vacuum chamber that is maintained at 10-3 Torr. The laser used is a CPA (chirp pulse amplification) system that generates 10mJ pulses of about 40fs duration at a 10 Hz repetition rate. The laser pulses are focused on to the droplet with a 30cm lens to achieve intensities up to 2×1016 W cm-2. In the prepulse experiments, the prepulse is generated using a beam-splitter and delaying the intense main pulse by about 10 ns from the weak pulse. The x-ray measurements from the droplets are carried out using NaI scintillation detector coupled to the conventional electronics for pulse height analysis and computerized data acquisition. The spectrum of the scattered light is measured with a CCD based grating spectrometer (WaveStar V570-1070 nm).
3. Results and discussion
When 15 µm methanol microdroplets are exposed to intensities ≤2×1016 W cm-2, we find that they generate copious hard x-rays. Even at these moderate intensities, hard x-rays up to 300 keV were observed and the measured bremsstrahlung spectra could be fitted with a Maxwellian electron energy distribution of about 52 keV in temperature, as shown in Fig. 1(a). Our investigation further revealed that a prepulse generated by the laser system critically alters the generation of the x-rays. If there is no prepulse about 10ns ahead of the main pulse, there is no measurable hard x-ray yield . Moreover, the x-ray generation is always accompanied by what appears as a bright flash of red light. Prepulse intensities of less than 1% of the main pulse resulted in the loss of both the red light flash and the hard x-rays. A prepulse that is only a few ps ahead of the main pulse does not show similar features at the intensities probed in our experiments: at least a few ns delay between the prepulse and the main pulse is necessary for these observations. Figure 1(b). shows the spectrum of the ‘red light flash’ that appears when measured at an angle of 30 degrees to the propagation direction of the incident main pulse. A similar spectrum is also observed at other angles to the propagation direction. Also shown is the spectrum of the faint main pulse that is measured from the scattered light in the absence of the prepulse. As can be seen from Fig. 1(b), the spectrum has a structure on the lower wavelength region or a large blue shoulder that extends to about 250nm from the incident 800nm pulse. The fine-structure on the broad blue shifted spectra varies from shot-to-shot. It is also very noticeable that no additional (red-shifted) peaks appear at wavelengths larger than the incident pulse.
The spectral profiles have a resemblance to previous observations made when 100 femtosecond pulses of 1016W cm-2 were sent through low dense gas targets [10, 11]. Those experimental measurements demonstrated that while there are no spectral components generated in the higher wavelength region, a broad blue shoulder appears that extends till about 30nm from the incident laser wavelength. The spectral shift increases as the intensity of the laser pulse is increased and is larger with more easily ionizable Xenon than with He or Ne at the same intensity. Rapid ionization gives rise to a free-electron plasma via a spatio-temporal change in the index of refraction, leading to a spectral shift of:
where ω 0 is the frequency of the incident light, x is the distance traveled through the plasma region of length z and n is the refractive index of the plasma medium, given by the Drude model: n=(1- /)1/2, where ωp is the plasma frequency.
The spectral shifts observed here have very similar behavior: (i) While there is no spectral components generated in the higher wavelength region, there is a broad structure in the lower wavelength region, except that structure extends to more than 200nm in our case. (ii) The blue-shifted spectral profile in our measurements depends strongly on the intensity of the main pulse that follows the prepulse. Figure 2(a) shows the variation in the intensity of the blue-shifted profile as a function of the main pulse energy, at a fixed prepulse intensity. This behavior is similar to the previous measurements with gas jets, where an increase in laser intensity enhanced the blue shifted spectral components. Larger intensities lead to increased ionization rates and a steeper gradients in the refractive index, giving larger blue shifts and yields of the shifted spectral components. Figure 2(b) shows the shift in the spectral component measured as a function of the prepulse energy at a given main pulse intensity. Ω laser refers to the frequency of the incident laser light and Ω light refers to frequency at the maxima of the blue shifted spectrum.
We should stress that there are significant differences with previous experiments due to the nature of the target and the materials that we are using. The microdroplets plasma are much more inhomogeneous than gaseous plasmas. The thickness of the target varies according to where the light traverses the sphere. The different spatial regions thus produce different shifts along the transverse Gaussian laser profile interacting with spherical droplet. The ionization of the molecules, and the ionization of the fragments that result, add additional complexity to the ionization dynamics. This may account for the fine structure observed in the ‘blue shoulder’. On a shot to shot basis, the droplet shows a small jitter about the focal waist and therefore the peak structure on the blue shoulder varies, though the latter appears for every pulse incident on the droplet. The spherical nature of the droplet plasma also leads to refraction of the scattered light. So the blue shifted spectral components are also scattered from the droplet at large angles outside the focal cone, more than that possible with the gaseous targets.
Using novel microdroplet targets to absorb intense ultra short pulses leads to a very different interaction dynamics compared to a conventional target. The target has densities comparable to a solid slab target, but has a very different geometry and is mass-limited. Unlike a solid slab target, where material with super-critical density always exists and the laser absorption is pre-dominantly in the critical density region, the interaction physics for the droplet is qualitatively different. Ionization and the subsequent plasma expansion of a droplet due to the prepulse leads to a large scale-length plasma that can be sub critical, unlike in solids. The spherical geometry of the droplet, with a size that is more than an order of magnitude larger than the wavelength of light, can cause scattering of the electric fields inside the droplet. Even if the incident intensity is not large enough to cause sufficient ionization on the surface of the droplet, the transmitted light can be focused inside the droplet via Mie scattering. A prepulse with an intensity up to 1014 W cm-2 can still undergo Mie scattering and focusing inside the droplet, giving rise to ‘hot-spots’.
Formation of such hot-spots has been demonstrated experimentally . The hot-spots in and around the droplet lead to faster droplet expansion and we have shown that in about 10ns, the droplet is expanded to twice the initial diameter . The main pulse now interacts with the large scale-length and large volume plasma that is at best close to the critical density. In this near-critical, large volume plasma prepared by the prepulse, the main pulse can be efficiently absorbed, leading to further rapid ionization and the large spectral shifts observed. Detailed modeling of the dynamics of the spectral broadening for our target is beyond the scope of the present work, and would involve a self-consistent calculation of the laser propagation and ionization through an inhomogeneous droplet. However, a simple estimate based on Eq. (1) shows that for a shift of about 200 nm and an effective interaction length for the leading edge of the pulse of 30 µm (the droplet preplasma diameter), a density change of 9×1020 cm-3 or 60% of the critical density for 800 nm light would be required in this case, which is quite plausible for our methanol droplet plasma.
The spectral shifts are larger here than those observed in the previous experiments because of both the larger length of interaction (a interaction length of about 10 µm was inferred in the previous measurement [10, 11] as compared to 30µm expected in our droplet experiments) and also because of the larger density of the material (more than an order of magnitude larger).
In the case of the droplets, the spectral shifts also seemed to be accompanied by efficient hard x-ray generation from the droplet plasma. Simultaneous measurement of the hard x-ray yield (50-300keV) demonstrate that both the x-ray emission and the intensity of the scattered light show similar variation with the prepulse energy. In Fig. 3 a plot of the change in hard x-ray yield with prepulse and yield of the blue shifted red-light is shown. The main pulse in these experiments has a intensity of 9×1014 W cm-2. It is clear from the figure that the increase in the red light is also accompanied by the generation of hard x-rays. This seem to indicate that the preparation of the optimal density preplasma in the droplet case not only results in rapid ionization (and blue shift) but also in the production of hot electrons. This correlation is seen by the simultaneous increase in the x-ray yield along with the blue shifted light. This provides an important observation for future modeling of the dynamics of interaction of the intense laser light in the droplet plasmas.
In summary, we have demonstrated that a prepulse prepares an optimum large scale-length spherical plasma when microdroplet targets are used in femtosecond laser-matter interactions. A main pulse with intensities ≤1016 W cm-2 induces rapid ionization and is subjected to blue shifting of about 200 nm, which is much larger than the shifts measured with gaseous targets at order-of-magnitude higher intensities. The preparation of the optimal length plasma is a prerequisite not only for the blue shift but also for the x-ray generation, which both exhibit a similar dependence on the laser parameters.
References and links
1. H. Schwoerer, P. Gibbon, S. Düsterer, R. Behrens, C. Ziener, C. Reich, and R. Sauerbrey, “MeV X Rays and Photoneutrons from Femtosecond Laser-Produced Plasmas,” Phys. Rev. Lett. 86, 2317–2320 (2001). [CrossRef]
2. J. Faure, Y. Glinec, A. Pukhov, S. Kiselev, S. Gordienko, E. Lefebvre, J.-P. Rousseau, F. Burgy, and V. Malka, “A laser-plasma accelerator producing monoenergetic electron beams,” Nature 431, 541–544 (2004). [CrossRef]
3. H.C. Kapteyn, M. M. Murnane, and I.P. Christov, “Extreme Nonlinear Optics: Coherent X Rays from Lasers,” Physics Today , 58, 39–44 (2005). [CrossRef]
4. J. Hein, S. Podleska, M. Siebold, M. Hellwing, R. Bödefeld, G. Quednau, R. Sauerbrey, D. Ehrt, and W. Wintzer,“Diode-pumped chirped pulse amplification to the joule level,” Appl. Phys. B 79, 419–422 (2004). [CrossRef]
5. T. Nishikawa, H. Nakano, K. Oguri, N. Uesugi, K. Nishio, and H. Masuda, “Nanohole-array size dependence of soft x-ray generation enhancement from femtosecond-laser-produced plasma,” J. App. Phys. 96, 7537–7534 (2004) and references therein. [CrossRef]
6. S. Düsterer, H. Schwoerer, W. Ziegler, D. Salzmann, and R. Sauerbrey, “Effects of a prepulse on laser-induced EUV radiation conversion efficiency,” Appl. Phys. B. 76, 17–21, (2003). and references therein. [CrossRef]
7. L. Malmqvist, L. Rymell, and H. M. Hertz, “Droplet-target laser-plasma source for proximity x-ray lithography,” Appl. Phys. Lett. 68, 2627–2629 (1996). [CrossRef]
8. M. Anand, C.P. Safvan, and M. Krishnamurthy, “Hard X-ray generation from microdroplets in intense laser fields,” Appl. Phys. B 81, 469–477 (2005). [CrossRef]
9. M. Anand, A. S. Sandhu, S. Kahaly, G. Ravindra Kumar, M. Krishnamurthy, and P. Gibbon, “Enhanced hard X-ray emission from microdroplet preplasma,” Appl. Phys. Lett. (in press).
10. W. M. Wood, C. W. Siders, and M. C. Downer,“Measurement of femtosecond ionization dynamics of atmospheric density gases by spectral blueshifting,” Phys. Rev. Lett. 67, 3523–3526 (1991) [CrossRef]
11. B. M. Penetrante, J. N. Bardsley, Wm. M. Wood, C. W. Siders, and M. C. Downer, “Ionization-induced frequency shifts in intense femtosecond laser pulses,” J. Opt. Soc. Am. B 92032–2040, (1992). [CrossRef]
12. H. Singhal, V. Arora, P. A. Naik, and P. D. Gupta, “Spectral blueshifts in laser light scattered from argon-gascluster plasmas,” Phys. Rev. A 72, 043201 (2005). [CrossRef]
13. K. Y. Kim, I. Alexeev, T. M. Antonsen, A. Gupta, V. Kumarappan, and H. M. Milchberg,“Spectral redshifts in the intense laser-cluster interaction,” Phys. Rev. A 71, 011201 (2005). [CrossRef]
14. C. Farve, V. Boutou, S.C. Hill, W. Zimmer, M. Krenz, H. Lambrecht, J. Yu, R.K. Chang, L. Woeste, and J.P. Wolf, “White-Light Nanosource with Directional Emission,” Phys. Rev Lett. 89, 035002 (2002). [CrossRef]