1. Introduction

High harmonic generation (HHG) using ultrashort pulses is a most versatile technique for generation of coherent radiation in the extreme ultraviolet (XUV) range. Therefore, the search for more efficient schemes for HHG in different isotropic media is of special interest. Originally, gases and laser-produced plasmas (LPP) were the main subjects of studies for conversion of ultra short near-infrared laser pulses towards the XUV region [1,2]. Another emerging field is HHG in solid bulk media, for example in ZnO or ZnSe [3]. However, the harmonics in that case are located in the near-infrared and visible ranges due to strong absorption of shorter-wavelength radiation in solids, thus restricting the applicability of this technique for generating coherent XUV radiation. HHG in solids could also come from nanoparticles. In this case, clusters, quantum dots and large nanoparticles may act as an intermediate material between atoms and solids. The relatively low absorbance from these small particles in XUV allows stronger harmonic yield from these species compared to commonly-used gaseous media.

Earlier, the advanced properties of LPP over gas media have been demonstrated in the case of carbon plasma [4–6]. The efficiencies of high-harmonic generation in carbon-containing plasmas and gases were compared under closely matched experimental conditions [4]. Those measurements have show that, in the 14–25 eV photon energy range, carbon plumes are more efficient than argon suggesting that the high-order nonlinear response of carbon nanoparticles is significantly enhanced. The studies have demonstrated higher conversion efficiency for the low-order harmonics generated in graphite plasma compared with argon gas in the case of 10-ns ablating pulses. Among the main factors influencing HHG conversion efficiency in medium the density and length play an important role. The other experimental conditions of HHG (confocal parameter of driving radiation, pulse duration and energy, etc.) were maintained equal for both media. In both cases weak focusing conditions, with the confocal parameter much longer than the medium, were used.

Carbon-contained plasmas have been proven to be most effective medium for lower-order harmonic generation in the 40 to 100 nm range of XUV. Both ablated bulk graphite and different carbon-contained clusters and nanoparticles, such as fullerenes, grapheme, carbon nanotubes, nanofibers and nanoparticles, allowed strong 9th to 21st orders of 800 nm driving field. The ten-fold enhancement of carbon HHG over argon HHG [5] was attributed to those harmonics. Meanwhile, carbon plasma provided approximately same cutoff for maximally enhanced harmonics as most frequently used argon gas.

In the past, the comparative studies of carbon plasma and argon gas have shown the advantages of the former medium once one became aimed in stronger harmonic emission in the 50–100 nm spectral range of XUV. One limitation of the intense carbon harmonic source driven by a 0.8 µm wavelength Ti:sapphire laser has been the low cutoff around ∼32 eV. The application of relatively longer-wavelength pulses allows generation of harmonics from carbon plasma up to ∼70 eV by increasing the driving laser wavelength to 1.71 µm [7]. As it was revealed in [7], the carbon harmonic intensity is found to be high despite the long wavelength driving laser. Experiments have shown only ∼30% decrease in the harmonic intensity when changing the driving laser wavelength from 0.8 µm to 1.71 µm. Meanwhile, notice that, at the shortest wavelength region, the HHG conversion efficiency from these lasers was notably low, which may diminish the advantages of the application of different concepts of harmonic enhancement, like quasi-phase matching (QPM), in the case of this plasma even using the near infrared sources. Such conclusion has been revealed during QPM studies of the harmonics generated from the tunable (1100–2400 nm) femtosecond source [8]. A search for the suitable plasma media showing preferable features compared with gases thus became an interesting topics of studies. To analyze the potentially efficient plasma media we have chosen some of them taking into account earlier reported data on their ability in generating higher order harmonics of 800 nm driving laser, compared with the carbon plasma.

The comparison of harmonic yields from argon and other plasma media has yet been reported. Meanwhile, many samples to be ablated have shown the relatively strong harmonics in the XUV region shorter than 40 nm. Among them, silver and silver sulfide plasmas are distinguished with strongest yield down to the cutoff region at around of 20 nm and beyond. Another advantage in application of these and other plasmas for comparison with gas-induced harmonics is the analysis of the behavior of XUV yield in the case of different geometry of experiments. In particular, the tuning of LPP with regard to gas jet allows defining the optimal conditions of joint influence of two sources of harmonics. LPP being installed before gas jet can provide the harmonics distinguished by divergence, cutoff and yield with regard to those produced from the plasmas coincided with the gas jet or installed after gas jet.

Silver has earlier shown the excellent properties of harmonic generation. The calibration of XUV registration system using different approaches has demonstrated large conversion efficiency of the harmonics in the range of H20 – H40 (∼10⁻⁵ [9,10]). However, there have never been back-to-back measurements that compare the efficiencies of gas and plasma harmonics, using the same pump laser and under the same conditions. Such comparison is important, since the harmonic emission in the case of HHG in LPPs is extremely sensitive to various experimental conditions, including the heating pulse intensity and the distance of the driving laser from the target surface.

In this paper, we analyze and compare the harmonic yields from the argon gas, silver plasma, silver sulfide plasma and stainless steel plasma. The single- and two-color pumps of gas and plasma media allows enhancement of odd and appearance of strong even harmonics during propagation through the LPPs and gases, even at very small ratio between second and first waves (∼1:30). We demonstrate the variation of the divergence of harmonics at different positions of LPP with regard to gas jet as well as resonance enhancement of single harmonic. These studies demonstrate 3- to 20-fold enhancement of harmonics from LPPs compared with gas medium at similar concentrations and lengths of gas and plasma.

2. Experimental arrangements

HHG experiments were carried out using 800-nm, 35-fs, 1-kHz, 1.5-mJ laser pulses (Spitfire Ace, SpectraPhysics), which were focused using 400-mm focal length spherical lens in the vacuum chamber contained gas jet and LPP (Fig. 1a). The gas jet (diameter of nozzle 0.25 mm) was placed at the focal plane of the focused radiation and operated with a continuous Ar flow. LPP was produced using either 35 fs, 800 nm pulses or 200 ps, 800 nm pulses from the uncompressed radiation of the same laser on the surfaces of bulk silver sulfide and silver targets placed at different positions with regard to the gas jet. The plasma was also formed on the surface of stainless steel nozzle in the vicinity of gas flow. The debris from the ablation coats the window from which the ablating pulse enters the vacuum chamber, thus reducing the fluence of heating radiation on the target with time. To exclude this effect, we used the protective microscope slide, which was installed between the target and the entrance window and could be easily replaced from time to time.

 

Fig. 1. (a) HHG setup. DB, driving beam; HB, heating beam; FL, focusing lens; NC, nonlinear crystal (BBO); Ar, argon gas inlet; TC, target chamber; LP, laser plasma; G, gas flow; XUVS, extreme ultraviolet spectrometer; CM, cylindrical mirror; FFG, flat field grating; MCP, microchannel plate; CCD, CCD camera. (b) Area of laser-gas-plasma interaction. N, target for ablation; HB, heating beam; P, laser-produced plasma; N, nozzle of gas tube; GF, gas flow; IA, interaction area of spreading gas and plasma, though which the driving femtosecond pulse propagated in the direction normal to the picture. (c) Harmonic intensities distribution of harmonics in the cases of single-color pump (800 nm, dotted curve) and two-color pump (800 nm + 400 nm, solid curve) of Ar gas.

Download Full Size | PDF

The targets were installed either at the focal plane of focused radiation, or in front of and after of the gas jet at a distance of ∼5 mm. Thus the positions of LPPs were within the confocal parameter of the focused driving beam (b = 14 mm). The maximal used intensity of femtosecond pulses in the focal plane was 5×10¹⁴ W cm⁻². In most cases, the single-color pump (SCP) of gas or plasma was used for HHG when we used 800 nm pulses. To generate harmonics using two-color pump (TCP), the 0.2-mm thick barium borate (BBO) crystal was inserted in the vacuum chamber on the path of the focused 800 nm radiation. The conversion efficiency of 400 nm radiation was 3%.

The concentration of gas particles was estimated to be 5×10¹⁷ cm⁻³, while plasma concentration was estimated to be 3×10¹⁷ cm⁻³. The diameter of the nozzle providing gas flow was 0.25 mm, while the sizes of the heating pulse on the target surface were ∼0.3 mm. The driving beam propagated at the distance of ∼0.5 mm from the nozzle and target surfaces. Thus one can assume a similarity in the concentrations of harmonic emitters and the lengths of gas and plasma media through which the 35 fs pulse was propagated to generate harmonics. Particularly, the active lengths of the media at the distances of 0.5 mm from the target and nozzle were assumed to be 0.3 and 0.4 mm for the gas and plasma medium, respectively.

Gas and plasma jets were intercepted with each other once the ablating target was placed close to the nozzle of gas jet (Fig. 1b). In most experiments, the LPP was positioned either in front of or behind the gas jet.

Harmonic emission was analyzed using XUV spectrometer contained flat field grating (1200 grooves/mm, Hitachi) and microchannel plate (Hamamatsu). The harmonic emission was directed towards the grating using either cylindrical or flat gold-coated mirror placed at 30 with regard to the propagation of laser radiation. Most of experiments were carried out using the collection of harmonic emission along the vertical axis using the cylindrical gold-coated mirror. The flat mirror was used once we analyzed the divergence of harmonics. The images of dispersed harmonic emission were collected using a CCD camera coupled to a computer to acquire the harmonic spectra. In both media, comparable divergences and Gaussian-like beam profiles of the harmonics were observed.

In earlier studies [5,6], a calibrated silicon photodiode was used to measure the harmonic energy, thus establishing the absolute measurements of conversion efficiencies and harmonic fluencies. In present studies we did not performed the absolute calibration of our registration system, but rather carried out the comparison of the harmonic yields from gas and plasmas at identical laser and medium conditions.

3. Results and discussion

In most of cases, we used strong first field (800 nm) and weak second field (400 nm) to pump the gas plasma plume. These studies demonstrated that, even at 30:1 ratio between these two fields, dramatic changes in the spectrum of harmonics occur, which point out the decisive influence of a weak second field on the whole process of HHG. Similar influence could be expected once one considers the variation of cut-off for TCP versus SCP schemes. The decrease of cutoff did not follow the quadratic dependence on the wavelength, though we observed the almost two-fold decrease of maximally generated harmonic order and stronger odd harmonics in the case of TCP of both gas and plasma media.

Strong harmonic generation in the case of TCP is possible due to the formation of a quasi-linear field, selection of a short quantum path component, which has a denser electron wave packet, and higher ionization rate compared with SCP [11]. The orthogonally polarized second field also participates in the modification of the trajectory of accelerated electron from being two-dimensional to three-dimensional, which may lead to removal of the medium symmetry. With suitable control of the relative phase between the fundamental and the second harmonic radiation, the latter field enhances the short path contribution, resulting in a clean spectrum of harmonics.

This TCP-induced enhancement of harmonics has earlier been realized through generation of 400-nm radiation in a separate channel with further mixing with 800-nm radiation in gases, as well as through direct generation of second harmonic (H2) in thin BBO crystals followed with focusing of two co-propagating beams in gases [12,13]. The following studies of TCP in LPPs [14–16] have also shown the advantages of the TCP of LPP with regard to the SCP using both 800 nm and 1300 nm fundamental sources and their H2.

Figure 1c shows the SCP- and TCP-induced HHG spectra from the argon gas. The efficiency of HHG was larger in the case of TCP with regard to SCP. Also the strong even harmonics were observed. Even harmonics were two times stronger than odd harmonics, while the cutoff was decreased in the case of TCP with regard o SCP (H27 and H22 respectively).

Harmonic generation in gas and silver plasma jets at different positions of LPP was carried out using SCP at similar intensity of driving pulse and approximately equal concentrations of gas and plasma media. The procedure of these experiments was as follows. Firstly, we achieved best harmonic yield from gas jet and then scanned the heating pulse either on the target placed close to the needle, through which gas flowed to the chamber, or targets placed prior to or after the needle.

Upper panel of Fig. 2a show the harmonic spectrum from argon. The H23 cutoff was observed in the case of gas emitters. Second third and fourth panels of this figure show the HHG in the cases of plasmas produced in front of gas jet, at the position of gas jet, and behind the gas jet, respectively. Correspondingly, the distance between LPP and gas jets along Z axis was -5, 0, and + 5 mm with regard to the focal plane of driving beam, which corresponded to the position of plasma within the confocal parameter of focused radiation. Figure 1a shows the interaction of gas and plasma jets in the case of their coincidence along the Z axis. Gas jet was placed in the focal plane of the focusing lens. One can see both the growth of harmonic yield and cutoff extension in the case of gas + plasma configuration. Strongest harmonics were observed in the case of coincidence of gas and plasma jets (third panel from the top of Fig. 2a).

 

Fig. 2. Comparison of harmonic spectra generating in pure Ar gas and mixture of Ar gas and Ag plasma in the case of (a) SCP and (b) TCP. (a) LPP was placed before gas jet (second panel from top), after gas jet (fourth panel), as well as was coincided with the gas jet (third panel). (b) LPP was placed after gas jet. The energies of driving and heating femtosecond pulses 0.9 and 0.4 mJ. The delay between pulses was maintained at 73 ns. (c) Images of harmonic spectra from Ar + Ag medium observed by CCD camera. Panel 1 corresponds to the position of LPP before gas jet. Panel 2 corresponds to the position of LPP after gas jet. Panel 3 shows the harmonic spectrum during shorter collection time to distinguish the details of the divergence of harmonics and the appearance of interference pattern along the vertical axis.

Download Full Size | PDF

Comparison of harmonic spectra generating in pure Ar gas and combination of Ar gas and Ag plasma in the case of TCP is shown in Fig. 2b. LPP in that case was placed at 5 mm after gas jet. The application of thin (0.2 mm) BBO crystal allowed generation of approximately equal odd and even harmonics from the gas + plasma medium. Three- to five-fold growth of even harmonic yield in the case of gas + plasma compared with gas medium at similar experimental conditions was observed. The enhancement factor was larger in the case of odd harmonics (compare the yields of H15, H17, and H19 in the cases of gas and gas + plasma media).

The raw images of harmonic spectra from different combinations of generating media in the case of replacement of cylindrical mirror by the plane one are presented in Fig. 2c. The goal of these studies was the qualitative definition of the conditions of different gas and plasma formations on the variation of harmonic divergence and harmonic distribution in the XUV range, rather than the quantitative measurements of the gain and conversion efficiency of HHG. Visual presentation of spectra and observation of the difference between the divergences of harmonics in the cases of plasma + gas and gas + plasma have motivated us to present the images of generating spectra appearing on the screen of monitor. The saturated images of harmonics are intentionally chosen to present the spectra for better viewing. In the meantime, for the line-outs of the HHG spectra we used the unsaturated images.

Figure 2(c) demonstrates the raw images of harmonic spectra obtained using SCP of plasma + gas (panel 1) and gas + plasma (panel 2) configurations. We also show the harmonic spectrum generated in the argon gas and silver plasma using smaller collection time of CCD camera (panel 3). One can see the difference in the divergences of harmonics generation in the case of two different positions of plasma jet with regard to gas jet. The divergence of harmonics arising from plasma + gas configuration was significantly smaller compared with the case of gas + plasma. Moreover, in the latter case we observed the irregularity in the divergence resembling the interference pattern along the vertical axis for most of harmonics (panel 3). This interference-like distribution of divergence can be attributed to the interaction of harmonics originated from gas jet with those from plasma jet.

In the case of gas HHG, the installation of gas jet prior to the focal plane causes the preferential involvement of short trajectory of accelerated electron in harmonic generation leading to narrower divergence of coherent XUV emission. Opposite case (gas after focal plane) increased the role of long trajectory of electron thus increasing the divergence of generated harmonics. Similar difference in the divergences of harmonics was observed during plasma HHG studies.

Below we address our observations of the double media configuration for HHG. One can assume that weak harmonics from the first medium (gas jet) can further be amplified in second medium (plasma jet). Similar concept of amplification has been analyzed in [17] when the first experimental demonstration of the parametric amplification of attosecond pulse trains at around 11 nm was demonstrated in the helium amplifier driven by intense laser pulses and seeded by high-order harmonics generated in a neon gas jet. Their measurements have suggested that amplification can take place only if the seed pulse-trains are perfectly synchronized in time with the driving laser field in the amplifier. In our case, the distance between gas and plasma media was maintained at ∼5 mm. It is difficult to assume the amplification of weak harmonics from gas jet in the following plasma jet at our experimental conditions. The exclusion of gas harmonics, by closing the nozzle, led to almost similar harmonic spectrum generated in the plasma medium. One of possible scenarios of interaction of gas harmonics and plasma harmonics in the case of gas + plasma configuration could be some interference between these two sources of harmonic emitters. We observed a weak interference pattern in that case (Fig. 2(c), panel 3). Nevertheless, it would be an important step towards designing amplifiers for realization of strong XUV pulses, provided the determination of the optimal conditions for enhancement from the first medium generating seeding pulses towards the second medium amplifying those XUV pulses. These studies are under consideration in our laboratory. This method of harmonic amplification can also be useful for determination of harmonic pulse duration, similarly to the technique demonstrated in [18]. Varying the delay with sub-10-as temporal resolution, they were able to resolve electric field evolution within the attosecond pulse train, which allowed them estimating the duration of the individual XUV pulses within the train to be about 0.2 fs.

Analogous studies of silver sulfide plasma were carried out by similar manner to clarify SCP and TCP of gas and plasma media at different conditions of experiment. Figure 3a shows the HHG using SCP in the case of plasma + gas configuration. One can see that in the case of LPP placed prior to gas jet the harmonic distribution was almost similar to the case when gas nozzle was closed. The presence of gas in these experiments was insignificant. Another situation was observed when LPP was installed after the gas jet (Fig. 3b). The harmonic yield was notably stronger compared with the former case. The comparison of two HHG spectra (with and without gas, Fig. 3b) shows that the influence of gas harmonics can be attributed to the enhancement of lower-order harmonics. Similarly, stronger odd and even harmonics in the case of TCP of two configurations (plasma + gas and gas + plasma) revealed the preferable position of plasma jet behind the gas jet (Fig. 3c).

 

Fig. 3. Harmonic spectra generated in the Ar gas + Ag₂S LPP medium in the case of presence (dotted curves) and absence (solid curve) of gas. (a) LPP was placed before gas jet. (b) LPP was placed after gas jet. (c) Two-color pump of argon gas and silver sulfide plasma. Upper panel: gas without plasma. Middle panel: gas after plasma. Bottom panel: gas before plasma.

Download Full Size | PDF

Finally, the experiments were carried out by ablation of the nozzle of gas jet. Figure 4 shows the harmonic distribution in the in the case of the argon gas and stainless steel plasma. The stainless steel needle was ablated in the area of the nozzle by picosecond pulses. Without the ablating nozzle by picosecond pulses, the harmonic spectrum represented the ordinary spectrum of gradually decreasing harmonics from gas jet (red solid curve). The ablation of nozzle led to appearance of the components of stainless steel ablation in the axis of driving beam propagation. These components drastically changed the HHG spectrum (blue dotted curve). The harmonics from H13 to H25 generated in the mixture of gas and plasma were five times stronger than those generated in the pure argon gas. Most intrigued pattern was observed in the area of harmonic cutoff. One can see that the 27th harmonic almost disappeared compared to the neighboring ones. Meanwhile, H29 was notably stronger that H25 and equal to H23.

 

Fig. 4. Harmonic spectra distributions in the cases of argon gas (solid curves) and argon as + stainless steel LPP. Insets show the CCD images of harmonics spectra and elemental distribution of used stainless steel, which contained ∼12% of chromium.

Download Full Size | PDF

Stainless steel contains a number of constituents that define its structure, hardness, and other physical and chemical properties. The elemental contain of used sample is shown in the inset in Fig. 4. One can see that one of the important constituents was the chromium (∼12%). The role of Cr in the SS harmonic spectrum generated in stainless steel plasma has been early studied by using pure Cr and pure Fe as the targets for HHG experiments [18]. The former plasma has shown a considerable enhancement of the intensity of 29th harmonic. Those observations clearly pointed out that the increase of the intensity of the 29th harmonic in the case of ablated stainless steel, although not as pronounced as in ablated Cr, was caused by the presence of the latter component in our sample.

Previous studies of the photoabsorption and photoionization spectra of Cr plasma in the range of 41–42 eV [19] have demonstrated the presence of strong transitions, which could be responsible for a suppressed pattern of the harmonic spectrum in the wavelength region of 29.5–31 nm. The region of ‘giant’ 3p → 3d resonances (44–45 eV) of Cr II spectra was analyzed in [19,20] and strong transitions that could enhance the nonlinear optical response of the plume were revealed. The harmonic wavelength can be resonant with the transition between the ground and the autoionizing state (the excited state embedded in the continuum) of the generating ion and that this transition exhibits high oscillator strength. In particular, in the four-step model describing the enhancement of resonance harmonics [21] the ionized and laser-accelerated electron is captured into the autoionizing state of the parent ion, and, in the final step, the radiative relaxation of this state to the ground state, a harmonic photon is emitted.

4. Conclusions

A search for suitable schemes for HHG in different isotropic media is of special interest of laser community. Originally, gases and laser-produced plasmas were the main subject of studies for conversion of ultra short near-infrared laser pulses towards XUV region. The comparative studies of carbon plasma and argon gas have shown in the past the advantages of the former medium once one became aimed in stronger harmonic emission in the 40–100 nm spectral region of XUV. Carbon plasma did not allow generation of strong harmonic in the shorter wavelength region. A search of suitable plasma media showing preferable features compared with gases thus became an interesting topics of studies. To analyze the potentially efficient plasma media we have chosen some of them taking into account earlier reported data on their ability in generating higher order harmonics, compared with carbon plasma.

In this paper, we have demonstrated stronger HHG in the case of different plasmas (Ag, Ag₂S, stainless steel) with regard to the harmonics generating in argon gas. We have compared the harmonic yields from gas and gas + plasma media under closely matched experimental conditions. Three- to five-fold growth of harmonic yield in the case of LPP compared with gas medium at similar experimental conditions was achieved. Our measurements have shown that, in the spectral range of 15–50 nm, the used plumes are more efficient than argon. We have also analyzed different schemes of HHG, such as single- and two-color pumps, tunable positions of LPP with regard to gas jet, which allowed determining best configuration of gas + plasma applications for harmonic generation.