Self-terminating barium ion laser at 614.2 nm

Highlights

• Detailed study of self-terminating Ba⁺ laser is presented.

• Application of eptron significantly improves Ba⁺ laser output characteristics.

• Hydrogen effect for metal vapor ion lasers is demonstrated.

Abstract

Experimental results of the laser operating on the barium ion self-terminating transition $6p{}^2{P}_{3/2}^o$ - $5d{}^2{D}_{5/2}$ at $\lambda =614.2$ nm investigations are presented. Through a series of tests, we demonstrated a significant improvement of the output characteristics of this laser by employing (i) a novel high-voltage nanosecond switch capable to operate at high pulse repetition frequency (more than 50 kHz), and (ii) hydrogen addition to the active medium. Achieved steady-state average output power in the burst-mode operation amounted to 125 mW with laser pulse duration of $\approx 5$ ns at half-maximum.

Introduction

The laser being investigated belongs to a class of self-terminating lasers [1] operating only in pulse mode due to intrinsic transitions properties. The most powerful and efficient self-terminating lasers oscillate on resonance-metastable (RM) transitions of neutral atoms, such as copper, gold, and strontium vapor lasers (VL). Applications of such lasers include fields like micromachining [2], medicine [3], [4], laser isotope separation [5], [6], [7], active optical systems [8], [9], [10], [11], etc. Improvement of the output parameters via investigating various operating conditions of RM lasers remains a key focus of this field [12], [13], [14], [15], [16].

Another selected direction of research of RM lasers is driven by transitions in ions. The latter allows to achieve values of average output power $P_{\mathit{av}}$ and lasing efficiency $\eta$ close to those of some RM transitions in neutral atoms, but can also operate in UV spectrum where the ions have a large number of appropriate transition lines [17].

For example, numerical simulations of RM transitions $4p{}^2{P}_{3/2}^o$ - $3d{}^2{D}_{5/2}$ ($\lambda =854.2$ nm) and $4p{}^2{P}_{1/2}^o$ - $3d{}^2{D}_{3/2}$ ($\lambda$ = 866.2 nm) in ${\mathit{Ca}}^{+}$ show that under the conditions of saturated power approximation the values of $P_{\mathit{av}}$ and $\eta$ are comparable to those of RM transition $4s4p{}^1{P}_1^o$ - $4s3d{}^1{D}_2$ ($\lambda$ = 5.546 $\mu$m) in Ca [18]. Similarly, employing kinetic model [19] for ${\mathit{Cu}}^{+}$ RM laser operating on $3d^{9}4p{}^1{P}_1$ - $3d^{9}4s{}^3{D}_1$ ($\lambda$ = 201.5) and $3d^{9}4p{}^1{P}_1$ - $3d^{9}4s{}^1{D}_2$ ($\lambda$=211.2 nm) transitions shows that with a thyratron-based high-voltage circuit, one can achieve total $P_{\mathit{av}}\approx 0.3$ W and efficiency $\eta \approx 0.027\%$at a pulse repetition frequency (PRF) of 10 kHz.

In contrast to on neutral atom RM transitions, RM ion lasers have been studied to a lesser extent. Though lasing on RM transitions in several ions such as ${\mathit{Pb}}^{+},{\mathit{Hg}}^{+},{\mathit{Yb}}^{+},{\mathit{Ba}}^{+},{\mathit{Sr}}^{+},{\mathit{Eu}}^{+},{\mathit{Ca}}^{+}$ is reported [[1], [20]], obtaining oscillation on ion RM transitions is hindered by several factors.

Firstly, the predominant mechanism of excitation of the upper laser level is the reaction ${\mathit{Me}}^{+}+e\to {\mathit{Me}}^{+*}+e-\mathrm{\Delta }E$, i.e. excitation from the ion ground level. Lasing threshold associated with concentration of active medium particles in the ground level for atom transitions lies in the range $n_A=(0.5\dots 1)·{10}^{14}$ cm⁻³, and there is no reason to believe that it will be much lower for RM ion lasers. At such pre-pulse concentrations of ions $n_{i0}$ (and correspondingly of electrons $n_{e0}$), the low active resistance of the medium renders achieving population inversion technically difficult. Specifically, due to slow heating of electrons at the leading edge of the excitation pulse, there is a preferential population of the metastable states [21] which reduces the laser pulse energy with increasing $n_{e0}$. As a result, with $n_{e0}\approx {10}^{14}$ cm⁻³ conventional excitation circuits fail to achieve lasing. To mitigate this problem, the leading edge of the excitation pulse has to be shortened. In particular, this approach was proven successful with a copper vapor laser in [22], [23], [24].

Second problem is the relatively small possible concentration of active medium ions. Indeed, due to weak imprisonment of radiation the lifetime of the resonance levels only slightly increases with respect to the spontaneous decay time of their populations into the ion ground level. Also, this problem arises in RM ion lasers where excitation from the ground level of the atom occurs via reaction $\mathit{Me}+e\to {\mathit{Me}}^{+*}+2e-\mathrm{\Delta }E$, e.g. in ${\mathit{Ca}}^{+}$VL. In this setting, considerable energy is required to reach high enough $n_{i0}$ to allow imprisonment of radiation to diminish the rate of decay of the ions in resonance state. This forces operation in a mode in which high values of $n_{e0},n_{i0}$ are realized automatically. This can be implemented, for example, at high PRF at which the plasma does not have enough time to recombine during the inter-pulse interval.

However, implementing high PRF in some vapor lasers poses a third problem. Specifically, for ${\mathit{Ba}}^{+}$VL the lower laser levels $5d{}^2{D}_{3/2,5/2}$ are located relatively close to the ground level (Fig. 1). Indeed, using the Boltzmann equation, one can estimate the ratio of populations of these metastable levels to the ground one $6s{}^2{S}_{1/2}$ at operating temperature $T=774°$C to be 0.24% and 0.12% respectively. In practice, the gas temperature distribution over the tube diameter is highly inhomogeneous with a maximum at the center, and, consequently, the relative concentrations of ions in metastable states are even higher. For example, in a gas discharge tube (GDT) with bore $d=12.7$ mm and wall temperature $T=847°$C the average temperature on the tube axis during the excitation burst reached $\approx 2500°$C [25]. In these conditions, the population ratios would have increased to 15.8% and 16% respectively.

Final obstacle is a rapid radiative decay of the upper laser levels to the lower ones. Specifically, ${\mathit{Ba}}^{+}$ laser transitions with $\lambda =649.7$ and 614.2 nm are characterized by decay times of only 32 and 23 ns correspondingly [26] which is more than an order of magnitude shorter than for the copper vapor laser. Even faster decay is accomplished by promising laser RM transitions in the UV spectrum range[17]. Thus, in order to achieve population inversion in a ${\mathit{Ba}}^{+}$VL, one has to use very short excitation pulses.

Described hurdles well explain the fact that RM lasing in ions was implemented only in a few active media with lower output power compared to the RM lasers on neutral atoms. Indeed, the highest average power and efficiency was achieved for ${\mathit{Ca}}^{+}$VL at $\lambda$ = 854.2 and 866.2 nm [27]. In this exmerimental setting, a GDT of length $l=15.2$ cm and bore $d=3.8$ cm allowed to achieve average output power 0.74 W at $f=6.85$ kHz and maximum efficiency $\eta \approx$ 0.05%.

Our interest in the ${\mathit{Ba}}^{+}$VL revolves around several advantages of this system. First, this laser can be a much cheaper alternative for a gold vapor laser in active optical systems. Second, via our previous experience with BaVL [28], it was revealed that high barium adhesion to GDT structural parts facilitates good retention of barium vapor within the active volume rendering much longer lifetime of the laser in comparison to the copper and gold vapor lasers.

Finally, as indicated in [[29], [30]], the lower laser levels $5d{}^2{D}_{3/2,5/2}$ of ${\mathit{Ba}}^{+}$ are characterized by quite high relaxation rates (electron quenching rate constant $k_e=3\dots 9·{10}^{-8}$ cm³s⁻¹ and quasiresonant charge exchange reaction constant $k_{+}=(3\dots 5)·{10}^{-10}$ cm³s⁻¹) which makes it possible to utilize PRF in the order of several hundred kHz.

To our knowledge, there have been no detailed experimental studies of ${\mathit{Ba}}^{+}$VL on RM $6p{}^2{P}_{3/2}^o$ - $5d{}^2{D}_{5/2}$ ($\lambda$ = 614.2 nm) and $6p{}^2{P}_{1/2}^o$ - $5d{}^2{D}_{3/2}$ ($\lambda$ = 649.7 nm) transitions (Fig. 1). Measurements carried out in [31] suggest that ${\mathit{Ba}}^{+}$VL in GDT with $d=8$ mm and $l=50$ cm yields peak power at several tens of watts with a pulse duration of $\approx 10$ ns and a PRF up to 15 kHz. Also, lasing on these transitions has been observed in [32].

Based on the above, the study of the ${\mathit{Ba}}^{+}$VL is undertaken on the one hand as a transitional stage of the creation of RM ion lasers in the UV spectrum range, on the other hand, in order to assess the prospects of its application.

However, the development of a high-voltage excitation circuits capable to produce pulses with a short rise time ($\approx$ 1–5 ns) and a high PRF (up to $\approx 100$ kHz) is a rather difficult task, mainly due to various limitations of switching technology. Nevertheless, one way to solve this problem is employing a new type of switch based on a capillary discharge with plasma cathode - an eptron [[33], [34]].