Efficient Excitation of Gain-Saturated Sub-9-nm-Wavelength Tabletop Soft-X-Ray Lasers and Lasing Down to 7.36 nm

We have demonstrated the efficient generation of sub-9-nm-wavelength picosecond laser pulses of microjoule energy at 1-Hz repetition rate with a tabletop laser. Gain-saturated lasing was obtained at $\lambda =8.85\mathrm{nm}$ in nickel-like lanthanum ions excited by collisional electron-impact excitation in a precreated plasma column heated by a picosecond optical laser pulse of 4-J energy. Furthermore, isoelectronic scaling along the lanthanide series resulted in lasing at wavelengths as short as $\lambda =7.36\mathrm{nm}$. Simulations show that the collisionally broadened atomic transitions in these dense plasmas can support the amplification of subpicosecond soft-x-ray laser pulses.

Popular Summary

Intense laser pulses in the soft-x-ray wavelength range below 10 nm are in high demand in applications that include high-resolution microscopy, materials spectroscopies, and materials modification in the nanoscale. The few national single-user free-electron laser facilities available worldwide cannot meet the strength of the demand, nor can they be easily reproduced in large numbers. Developing much more compact and widely accessible soft-x-ray lasers becomes a necessary scientific challenge. In this paper, we demonstrate the efficient generation of intense laser pulses of sub-9-nm wavelengths, picosecond durations, microjoule energies, and high-repetition rates using a tabletop, plasma-based laser.

Significant progress has been achieved in the past several years in the development of compact, plasma-based soft-x-ray lasers. They have enabled tabletop applications at wavelengths as short as 13.2 nm. But, extending compact plasma-based lasers to shorter wavelengths is very challenging due to a steep increase in the required pump energy. The shorter the desired wavelength is, the larger the pump energy required, which, in turn, limits the repetition rates. We tackle this challenge with a new pump-laser-focusing geometry that allows us to heat a uniform plasma column of optimized density at nearly the speed of light with a picosecond optical laser pulse of only 4 J. The rapid and uniform heating leads to the excitation of efficient laser transitions in lanthanide ions excited through collisions with fast electrons in the plasma. In our experiments we have obtained gain-saturated lasing at 8.85-nm wavelength in 29-times-ionized lanthanum atoms. Furthermore, scaling of the result along the next several heavier elements of the lanthanide series has resulted in lasing at several shorter wavelengths, down to 7.36 nm.

We have also conducted simulations to show that in these dense plasmas the collision-broadened laser transitions can support the amplification of sub-picosecond soft-x-ray pulses. This opens a path for the generation of bright femtosecond soft-x-ray laser pulses on a tabletop. The short wavelengths, microjoule pulse energies, ultrashort pulse durations, and high-repetition rates of these lasers will enable new applications, for example, time-resolved imaging of ultrafast nanoscale phenomena in materials, to be realized in small laboratory settings.

Figure 1

Schematic representation of the setup used in the sub-9-nm SXRL experiments. (a) A pair of cylindrical mirrors focuses the short-pulse beam at 35° grazing incidence on the target, creating a line focus of uniform width. A reflection echelon is used to obtain quasi-traveling-wave excitation. (b) Ray-trace simulation of the line focus on target. All rays fall within the linewidth defined by diffraction (red lines).

Figure 2

(a) End-on spectra of a line-focus lanthanum plasma column showing saturated amplification in the $\lambda =8.85\mathrm{nm}$ line of Ni-like La. (b) Intensity of the $\lambda =8.85\mathrm{nm}$ laser line as a function of the plasma-column length. The line is a fit of the data that yields a gain coefficient of $33{\mathrm{cm}}^{-1}$ and a gain-length product of 14.6. The points are the average of 10 laser shots and the error bar represents 1 standard deviation.

Figure 3

Characteristics of the $\lambda =8.85\mathrm{nm}$ lanthanum laser. (a) Measured near-field pattern. The center of the laser beam at the amplifier exit is $14\mu \mathrm{m}$ from the target. (b) Measured far-field pattern at a distance of 0.83 m from the source.

Figure 4

(a) Measured laser intensity as a function of the delay between the peak of the prepulse and the short pulse. The points are the average of 4–15 laser shots and the error bar represents 1 standard deviation. (b) Measured shot-to-shot intensity variation of the $\lambda =8.85\mathrm{nm}$ laser operating at a 1-Hz repetition rate.

Figure 5

Computed variation of the FWHM pulse duration of the $\lambda =8.85\mathrm{nm}$ ASE laser (triangles), and amplified high-harmonic seed pulse (circles) as a function of the lanthanum plasma column length. The seed-pulse duration initially increases due to linewidth narrowing and later, by saturation broadening. The high-harmonic seed pulse is assumed to have an energy of 0.1 nJ and the gain and plasma conditions are those resulting from the simulations described in the text.

Figure 6

End-on spectra showing lasing at progressively shorter wavelengths in the $4d{}^{1}S_0\to 4p{}^{1}P_1$ line of nickel-like lanthanide ions, down to $\lambda =7.36\mathrm{nm}$ in nickel-like samarium.