When using a nuclear-optical converter (NOC) to pump NPLs, immediate excita­tion of the laser medium is done using a well-known method: optical pumping. In this case, the kinetic energy of the nuclear radiation is initially converted by the NOC into light (luminescent) emission. (Sometimes NOCs are called “nuclear” lamps). The possibility of using NOCs for laser pumping was apparently examined for the first time in a U. S. patent [38] where various crystal and plastic scintillators and liquid and gas luminescent media were proposed as NOCs, and dielectric
materials doped with neodymium and chromium and other condensed and gas laser media known at that time were proposed as laser elements.

Pulsed lasing at electron-vibration transitions of the first negative system of the molecular ion N2+ (X = 427.8 nm) with pumping of a high-pressure He-N2 mixture with an electron beam with a duration of 20 ns was obtained in 1974 [146]. Later on, with pumping of the mixture He-N2-H2 (X = 391.4; 427.8 ns) with an electron beam with a duration of 0.7 qs, it was possible to realize a quasi-CW operating mode using hydrogen to quench low laser levels: N2+(X) + H2! N2H+ + H [147]. The efficiency of the laser was ~1 % at a specific power deposition q = 5-10 kW/cm3.

The possibility of pumping this laser with nuclear radiation was examined in study [148]. As a result of calculating the kinetics of plasma processes in the mixture He-N2-H2, it was concluded that it was possible to achieve lasing at X = 391.4 and 427.8 nm for q = 1-3 kW/cm3, which is quite attainable in experi­ments on fast reactors EBR-L and SPR III.

Soon afterward, the same authors used an EBR-L setup to obtain lasing at the 391.4-nm line when the He-N2-H2 mixture was excited by uranium fission frag­ments with a output power of ~10 W and nI~0.01 % [149]. This laser became the first NPL radiating in the UV range of the spectrum. Later on, in experiments with the EBR-L reactor, lasing was also registered at X = 427.8 nm of the ion N2+ [150]. For an optimal mixture of He-N2-H2 at pressures of 3.5-5 atm (partial pressures of nitrogen and hydrogen 3-10 Torr), at lines 391.4 nm and 427.8 nm, laser efficiencies of 0.3 % and 0.2 %, respectively, were obtained. Confirmation of these results was obtained in experiments [151] with a BARS-6 pulsed reactor, in which a He-N2-H2 mixture was excited by uranium fission fragments. Lasing was observed at the lines 391.4 and 427.8 nm for substantially lower specific threshold power depositions of 50-60 W/cm3 (nI ~0.1-0.2 %).

As the results of experiments and calculations showed (see Chap. 8), it is necessary to fulfill two basic conditions to obtain lasing in stationary mode: provide the necessary specific power deposition of the laser medium, and organize flowing of the gas mixture in the laser cell so as to eliminate its overheating and reduce the optical non-conformities. Taking into account these conditions, VNIIEF designed and built laser module LM-4 [1—6] in 1994 for joint operation with the powerful BIGR pulsed reactor [7], which can, in one of its operating modes, generate neutron pulses with a duration of >1 s. The experimental complex was supposed to demonstrate the possibility of creating reactor laser-type units with a stationary operating mode.

The LM-4 laser module [1—6] is a stationary-operating gas NPL functioning in gas-flowing mode (Figs. 6.1 and 6.2). The module and device for circulation of the gas were placed on a cart (see Fig. 6.1) and moved to the reactor core on it. The LM-4 module consists of four identical laser channels (Fig. 6.3). Over the length of the uranium layers, the average thermal neutron fluence reached values of (5-7) x 1014 cm~2 with a duration of 0.5-1.7 s.

Laser channels of the module are included in a single gas loop and are separated from one another by heat exchangers (radiators), which cool the gas when it is passed through. The radiators are packages 100-cm long, comprised of thin

© Springer Science+Business Media New York 2015 183

S. P. Melnikov et al., Lasers with Nuclear Pumping,

DOI 10.1007/978-3-319-08882-2_6

aluminum fins around 0.3-mm thick, and the same gaps between them. To equalize the gas dynamic perturbations of the gas flow, the edges of the fins were sharpened. The radiators do not have forced cooling, so their functionality is limited by their own heat capacity.

The active length of each channel, determined by the length of the uranium layers along the optical axis, is equal to 100 cm. The dimension of the channel in the direction perpendicular to the gas flow is 2 cm; in the direction parallel to the flow it is 6 cm. Layers of metallic 235U about 5 mg/cm2-thick are applied to 100 x 6 cm2 plates. At their ends, the channels are closed by windows arranged at a Brewster angle. The experiments primarily used stable optical cavities consisting of a 100 % spherical mirror with a radius of curvature of 20 m and a flat semitransparent dielectric mirror. The distance between the mirrors is about 170 cm.

The gas-flowing system provides flow of the gas mixture sequentially through all the laser channels in the transverse direction relative to the channel axis at rates of 4-10 m/s. At a gas flow rate of 10 m/s, the duration of the stationary flow is 3-5 s. The gas-flowing device is a circular, single-blade force-pump capable of creating an overpressure of 80 Torr. The internal diameter of this unit is 80 cm, the blade length is 109 cm, and the angle of the full rotation of the blade is 270°.

To ensure the functionality of the LM-4, the following devices were developed: a transport system, which accomplishes and controls the module movement from the measurement area to the reactor hall, and fixes it close to the reactor core [1]; a system for degassing, filling, and flowing of the gas [1, 4]; the systems for remote adjustment of the cavity mirrors and withdrawal of laser radiation from the reactor hall to the measurement area [6, 8, 9]; an automated system for registration of parameters of the laser radiation, neutron flux, and gas mixture rate [1, 6,10]; and a technological monitoring system for communications between the BIGR reactor and the module [1, 6].

The optical circuit scheme of the experiments is shown in Fig. 6.4. The route for withdrawal of laser radiation from the reactor hall includes aluminum 100 %

mirrors (parts 2-6) and a beam splitter (7). A lens system (10) was used to reduce the transverse cross section of the beam and construct a reduced image of the output mirrors of the laser channels at the input of the registration unit. Magnetostrictive motors with a feed pitch of around 1 ^m were used to rotate some of the mirrors.

Depending on the experimental tasks, the registration unit included various instruments and methods: the photoresistors FSA-G1 and avalanche photodiodes LFD-2 for registration of the lasing pulse shape, laser pulse energy meters (IMO-2N and J50-710 of the Molectron Company), methods based on the Ragulsky wedge to measure radiation divergence, a lateral-shift interferometer with a Pulnix camera by the Spiricon Company to study the optical non-uniformities of the laser media, etc.

The first series of experiments from 1994 to 1995 demonstrated the possibility of NPL operation in stationary mode with a lasing duration of up to 1.5 s [2—5]. The choice of composition and pressure of the mixtures (He, Ne)-Ar-Xe (A = 1.73, 2.03 ^m) was made on the basis of experimental data obtained using the LUNA — 2M setup (see Chap. 3, Sect. 3.1). Oscillograms of one of the experiments are shown in Fig. 6.5. Gas flowing was started roughly 0.5 s before the start of the neutron pulse. In the absence of gas flowing through the gas channels, there was no

lasing. The lasing threshold for different mixtures was achieved with thermal neutron flux densities of (1-5) x 1014 cm-2 s-1. When the lasing threshold was exceeded by a factor of 2-3, the lasing and pumping (neutron pulse) pulse shapes were roughly the same. The maximal lasing power was 15-20 W for each channel.

Before the second series of experiments (2001), the systems for module control, registration of radiation, and calibration of the optical scheme were improved. Experiments were aimed at studying the possibility of serial addition of laser channels (see Chap. 10, Sect. 10.5) and development of methods based on the lateral-shift interferometer to determine the optical non-uniformities of the laser medium [5, 6, 11, 12]. As a result of the experiments, the possibility of a serial scheme for adding laser channels was demonstrated. For example, for an Ar-Xe (100:1) mixture at a pressure of 0.5 atm, the output power at the 1.73 pm line when two channels were added proved to be 2.4 times more than for a single channel. The lasing duration at this time increased from 0.65 to 1.2 s.

Interferometric investigations of optical non-uniformities showed the correct­ness of conception about gas-dynamic and thermophysical processes in NPLs with transverse gas flowing (see Chap. 9, Sect. 9.3). Thus it follows from the interfero — grams obtained that an optical wedge forms in the longitudinal direction with respect to the gas flow, while in the transverse direction there is a positive distrib­uted lens. The data of interferometric measurements can also be used to determine the energy deposition to the gas medium [13]. A comparison of the specific power depositions determined by the interferometric method and as a result of calculations using the measured number of fissions in uranium layers showed their approximate agreement, with a precision of 10-20 %.

Possible causes of insufficiently high energy parameters obtained using the LM-4 module include a number of design defects, in particular the presence of windows arranged at the Brewster angle in the laser channels, as well as the

Fig. 6.6 Transverse section of LM-8 laser module: (1) cover of housing; (2) gas pipeline; (3) body of module’s work zone; (4) shielding made of borated polypropylene; (5) radiator; (6) laser channel; (7) supplying gas pipeline; (8) graphite neutron reflector; (9) neutron moderator; (10) assemblage with four laser channels; (11) deferent gas pipeline; (12) force-pump housing; (13) force-pump blade

placement of the Plexiglas fast neutron moderator inside the sealed space to be evacuated. The latter circumstance can lead to a reduction in the purity of the gas mixture owing to its contamination by impurities adsorbed by the moderator. To eliminate these defects, the LM-8 eight-channel laser module [5, 14] was devel­oped, in which the polyethylene moderator and graphite reflector are arranged outside the active laser volume. A diagram of the transverse section of the LM-8 module is shown in Fig. 6.6.

An optical scheme of the first series of experiments with the LM-8 laser module is shown in Fig. 6.7. The module allows some experiments with serial addition of two and four laser channels to be carried out. Channels 1,8, and 3-6 are combined by optical cavities using reflecting mirrors. The numbering of the channels follows the direction of gas mixture flowing. The individual channels 2 and 7 are “control” channels.

In the first experiments [14] with the LM-8 module on the BIGR reactor, it was possible to serially add the radiation from four laser channels, and to obtain a relatively high efficiency of laser conversion (n ~ 0.7 %) at low neutron fluences of ~1013 cm~2, which corresponds to a power deposition of just 0.4-0.6 W/cm3.

All of the examined NPL designs and LM-4 and LM-8 modules are intended basically for operation close to the reactor core. The development of a laser module that is an elementary cell of the reactor core is the next phase in the path of development of the reactor laser. One such version, examined in study [15], was the 16-channel LM-16 module, surrounded by a graphite neutron moderator. Laser channels used in the LM-4 and LM-8 devices formed the basis of the LM-16 module. When such a module is used in the reactor core of a continuous stationary reactor, it is possible to resolve problems associated with the compatibility of the laser cell with the reactor core, and certain “laser” problems, for example addition of radiation of the laser channels and its withdrawal from the reactor core [15].

At the VNIIEF, a concept was proposed for a system with transverse gas flowing (relative to the optical axis) [22—24]. This system consists of a set of laser channels with a rectangular cross-section that have planar uranium layers with a width of ~10 cm that are deposited to the inner surface of the channel side walls parallel to the gas flow direction. In order to cool the mixture heated in a channel, a radiator is placed at its outlet. The most convenient and simplest type of radiator consists of a set of thin plates that are positioned parallel to the gas flow and perpendicular to the uranium layers. An identical radiator is placed at the channel inlet which, by virtue of gas heat exchange with the plates, makes it possible to receive the requisite inlet temperature of the working mixture along the entire length of the laser and to equalize its velocity. The laser channel cross-section is shown in Fig. 9.1. The next laser channel can be installed behind the outlet radiator, for which this radiator will also be the inlet radiator, etc. Thus, a chain of laser channels can be built that are merged into a common gas circuit.

These experiments used three various laser devices, which consisted of two vol­umes connected optically to each other by a small coupling aperture [22]. The first cylindrical volume, which was located near the y-radiation source, formed the field of the radiation that was synchronized with the pumping wave. Subsequent radia­tion proceeded through the coupling aperture and entered the second (conical) volume and was amplified to the necessary energy. As a result, a radiation wave front, which was almost spherical, formed at the output of the conical amplifier, and the equivalent beam divergence was determined by the ratio of the diameter of the coupling aperture to the cone length.

The cylindrical part of these three laser devices was 100-cm long with a diameter of 10 cm. The conical portion of two from these devices was 2-m long, and the third device which had target chamber, was 14-m long and 0.8-m base diameter. As the active medium, a Ne-Xe-NF3 (1,000:26.2:6.2) mixture was used at 5 atm. The dependence of the specific power deposition into the active medium on the distance to a coupling aperture 3 mm in diameter is shown in Fig. 12.4. The scheme of the laser device with a 30 l vacuum target chamber which was placed 12.5 m from the top of the cone is shown in Fig. 12.5. A spherical mirror 80-cm in diameter with a 2.1 m focal distance was situated at a distance of 15 m from the top of the cone. The mirror focused the laser radiation inside the chamber on the target through a quartz window 36-cm in diameter. The target was shielded from the direct influence of the y-radiation by a 50-cm thick steel screen. The experiments used a target with a reversed corona [24], which is a thin-walled polyethylene cone 30-mm high with an open base and an apex angle of 30°.

The experiments measured laser parameters and the characteristics of the X-ray emitted by the plasma that formed when the target was irradiated with the laser radiation. The transmittance of the target chamber window at the laser wavelength

Fig. 12.4 The dependence of the specific power deposition of the active medium on the distance to the coupling aperture placed at the top of the cone [23]

can, in principle, be reduced by y-radiation. Therefore, this parameter was controlled.

The laser device with a target chamber had the following laser characteristics [13, 21—23]: the energy (power) of the laser radiation was 700 J (1.4 x 1011 W), half-amplitude pulse duration of the laser radiation was 7 ns (Fig. 12.6), laser energy efficiency was 3 %, and the contrast of the output radiation was >104. The experiments showed that the transparency of the quartz window did not change during a pump pulse. The characteristics of the recorded X-ray radiation agree with the calculated results for the laser energy at the target—about 1 kJ. This testifies that a spherical light wave formed in the conical amplifier since only in this case is it possible to transfer the laser energy to the target without losses using a focusing system.

The reported energy parameters were not the maximum. In subsequent experi­ments, a cylindrical XeF-laser with a larger active medium volume showed a laser energy of 6 kJ at an intrinsic efficiency of about 4 % [13].

Lasers

2.1 Specifics of Experiments on Pulsed Reactors

Virtually all of the experimental investigations to seek and study the characteristics of NPLs excited by the products of nuclear reactions were carried out on pulsed nuclear reactors. Pulsed nuclear reactors are distinguished for the composition and structure of their core, the duration and fluence of the neutron pulse, the volume and configuration of the space for the irradiation, and the pulse repetition frequency. Some aspects of the use of pulsed reactors for laser pumping were examined in the survey [1]. This section uses some of the data from study [1], supplemented by information from other sources.

By now—in Russia and elsewhere—more than 10 pulse reactors and a number of modifications have been developed and are in operation. More information about pulse reactors and their characteristics is contained in monograph [2]. The fuel for such reactors includes: metallic highly enriched 235U, uranium-molybdenum alloys, a triple alloy of uranium-zirconium-hydrogen, uranium-aluminum alloys, aqueous solutions of uranium salts, etc. Starting in the 1950s, pulsed reactors were inten­sively utilized in various fields of science and technology as powerful sources of neutron and у radiation. The duration of neutron pulses varied from 30 ps to ~0.1 s at neutron flux densities in experimental channels of 1017-1019 cm-2 x s-1 and neutron fluences of 1013-1017 cm-2.

The most complete information about pulsed reactors that operate (or operated) in various countries, the operating principles of these devices, and their character­istics and application, is contained in the monograph [2]. In Russia, pulsed reactors are developed and operated primarily at VNIIEF and VNIITF [2, 3]. At various times in Russia, the pulsed reactors VIR-1, VIR-2, TIBR, BR-1, BIGR (VNIIEF), IIN-3 (IAE), EBR-L (VNIITF), BARS-6 (FEI) have been used, and outside Russia—pulsed reactors such as the Godiva, SPR, TRIGA, ACRR, APRF, etc., which have similar characteristics. In 2003 reports were issued about experimental NPL investigations in China on the CFBR-II reactor [4]. Table 2.1 shows the

© Springer Science+Business Media New York 2015 S. P. Melnikov et al., Lasers with Nuclear Pumping, DOI 10.1007/978-3-319-08882-2_2

characteristics of certain of these reactors and experiments with NPLs conducted on them (the results of experiments are discussed in Chap. 3).

Initially the pulsed reactors were developed for testing various materials and electronic devices inside or close to the reactor core, so by no means were all of them adapted for conducting complex laser experiments. The reactor EBR-L [2,3,31], the dual-core reactor BARS-6 [3, 32], as well as the water reactor VIR-2 M [2, 26, 33], originally designed for general research uses and later adapted for experiments with NPLs, were perhaps the exceptions.

In experiments to find laser media for NPLs, fast neutron reactors were used as a general rule. Examples are the BR-1, EBR, Godiva, and SPR using small reactor cores (~30 cm) made of metallic 235U or its alloys, with a duration of reactor pulse of 50-100 qs. This selection may be explained by the fact that such reactors assure a maximal neutron flux density, and accordingly, maximal specific power depositions in laser media, thus facilitating attainment of the lasing threshold in NPLs. The reactor core usually is set at a height of 1.5-2 m from the floor of the casemate, the thick walls of which serve as biological shielding. The fluxes of neutrons and у quanta are maximum at the center of the reactor core and decrease roughly by an order of magnitude at its outside surface. Among the drawbacks of using such reactors include the large spatial non-uniformity of the neutron flux, which limits the length of the laser cells to ~50 cm. The exception is the two-core BARS-6 reactor, which can be used to uniformly irradiate laser cells up to 150 cm in length.

Different versions of the placement of NPLs in experiments using fast neutron — pulsed reactors are shown in Fig. 2.1. The version of Fig. 2.1a is the most widespread, because in this case the influence of NPLs on the parameters of the reactor is not significant. The maximal specific power deposition of the gas media

of up to 5 x 103 W/cm3 is implemented with the arrangement of the laser cells inside the reactor core (Fig. 2.1b, d).

The active length and volume of the NPLs can be increased using a multipass laser cell. Two versions of such devices are shown in Fig. 2.2. In the first case [18, 34], up to six ceramic tubes were placed inside a rectangular stainless-steel chamber; layers of 235U3O8 were deposited to the interior surfaces of these tubes having a diameter of 2.54 cm. The full active length of such a laser reached 240 cm [34]. Flat gold-coated mirrors were used as the deflecting mirrors. Another version

of the design [35], intended for excitation of gas media with nuclear reaction products 3He(n, p)3H, differs in that there are no tubes with uranium layers, and rectangular plates with a gold or silver coating were used as the deflecting mirrors. In this case, when the angle between the cavity mirrors is changed, it is possible to vary the number of passes in the cell and consequently the active length of the laser.

Laser cells usually are surrounded by a layer of moderator (polyethylene, Plexiglas, water) 3-5-cm thick for softening the neutron spectrum, making it possible to increase the laser power deposition by a factor of 10-100 when using the nuclear reactions given in Table 1.4 (Chap. 1, Sect. 1.2). When pulsed reactors with a short pulse duration are used (50-100 qs), the time dependencies of the flux of fast neutrons (reactor power) and the pumping power of the laser medium do not coincide, which is related to the process of thermalization of the neutrons inside the moderator. Figure 2.3 shows the results of calculations [36] of the time dependence of the power deposition (flux density of the slowed neutrons) inside the moderator under the conditions of experiments [21] (thickness of the cylindrical polyethylene moderator 5 cm, duration of reactor pulse t1/2 = 50 qs). The use of a moderator spreads the pumping pulse duration to ~150 qs, and shifts the pulse maximum position on ~30 qs. The results of calculations agree with experimental data.

It is more convenient to optimize the design of the NPLs (composition and parameters of laser medium, design of cavity, methods of removing surplus heat) for the purpose of selecting the optimal variation of the cell of the stationary or quasi-stationary nuclear-laser facility using pulsed nuclear reactors based on ther­mal or intermediate neutrons, with the moderator in the reactor core: the VIR-2 M reactor with reactor core from a solution of 235U salts in water, and the TRIGA pool-type reactor and its modification, the ACRR, the fuel elements of which are made from a uranium-zirconium-hydrogen alloy. These reactors are characterized by a long-duration neutron pulse (1-50 ms), with a maximal neutron fluence of ~5 x 1014 cm~2, as well as the possibility of experimentation with laser cells up to 200-cm long.

The above experimental layouts were used primarily to search for active NPL media and study their characteristics (laser spectrum, energy characteristics, and laser thresholds). To study multichannel NPLs, VNIIEF created the experimental complex LM-4/BIGR [15, 16] and the IKAR nuclear-laser facility, which consti­tutes a model of the reactor-laser [37]. VNIITF and FEI are developing multichannel LIRA [38] and Stand B [39] facilities. Among the studies outside of Russia with which we are familiar, one can note the experiments on the ACRR reactor (Sandia Laboratories, United States) using the large-scale laser setup ALEC (Advanced Laser Excitation Cavity) [19]. The designs of such complex multichannel facilities and the basic results obtained from studying them are cited in Chap. 6.

Following the initial stage, a sequence of plasmochemical processes occur in the plasma, tending to return it to its equilibrium state. An important place among such processes is held by charged particle recombination reactions, as a result of which plasma neutralization and the formation of charged atoms occur.

Kinetics of Plasma Processes for a Single-Component Mixture

For a single-component gas medium A at atmospheric pressure, the basic recombi­nation processes are:

1. collisional-radiative recombination

A+ + e + e! A * +e, (4.6)

2. three-body recombination

A++ e + A! A *+A, (4.7)

3. dissociative recombination

A++ e! A *+A. (4.8)

Molecular ions A+ are formed as a result of ion conversion:

A++ 2A! A++ A, (4.9)

and the rate constants of this process differ insignificantly for all ions of rare gases and amount to kic = (0.6-3.5) x 10~31 cm6/s at a gas temperature of Tg = 300 K [44].

As a result of recombination processes (4.6) and (4.7), initially highly excited atom states are formed, which are stabilized through collisions with a third particle (electron or atom). Further relaxation of excited states occurs as a result of colli­sions with plasma electrons and gas atoms, and through spontaneous decay. A special place among the recombination processes is held by the dissociative recom­bination reaction (4.8), which at a high pressure often is not only the basic process of neutralization of charged particles, but also one of the chief channels for the formation of excited particles, including populating of the upper lasing levels of NPLs [45]. In the case of dissociative recombination, the binding energy of the recombination electron is converted into the kinetic energy of recession of the atoms.

Values of the rate constants of recombination processes differ greatly and depend first of all on the electron temperature Te. For recombination processes (4.6) and (4.7) of atomic ions A+, the rate constants equal:

• for collisional-radiative recombination kcr = 4 x 10~9Te~45cm6/s (Te, K) for any atomic ions of rare gases [46];

• for three-body recombination ktrк(0.5-30) x 10~22Te~25cm6/s (Te, K) for A = Xe, Kr, Ar, Ne, He [47].

Rate constants of the dissociative recombination (4.8) for molecular ions of rare gases are shown in Table 4.8 [48]. The dependence on electron temperature for kdr is significantly weaker than for kcr and ktr, so with an increase in Te, the influence of dissociative recombination grows sharply.

For clarity, let us examine an argon plasma and estimate the characteristic times of plasma processes for q к 0.01-5 kW/cm3, which occur during operation of gas NPLs. Given an atmospheric pressure of the argon, Teк0.5-2.0 eV; ne к 5 x 1012-

2 x 1014 cm-3 and Z ~ 2 x 10-7-7 x 10-5. A comparison of the characteristic times of the recombination process Tcr = (kcrne2)-1« 1 s, Ttr = (ktrne[Ar])-1« 0.3 s ([Ar] = 2.7 x 1019 cm-3, the concentration of Ar atoms), the process of ion conversion Tic = (kic[Ar]2)-1« 5 x 10-9 s and dissociative recombination t^ = (kdrne)-1 « 1 x 10-7-2 x 10-6 s shows that in argon plasma, neutralization of charged particles occurs exclusively by the channel of dissociative recombination of molecular ions.

A similar conclusion can be drawn for plasma based on Ne, Kr, and Xe. First of all, for helium plasma, the constant kdr is two or three orders of magnitude lower; secondly, the process of electron thermalization occurs more efficiently, and accordingly, Te is significantly less. In this case, one cannot entirely neglect the recombination processes (4.6) and (4.7) with the participation of atomic He+ ions.

Balance equations of charged particles in plasma for the conditions examined above, without taking into account recombination processes (4.6) and (4.7), can be written as follows

XF1 = f+- к’сИ+1М2.

ІИ = ксИ+ЦД]2 — kdr [A+ ]ne,

ne = [A+] + [A+],

where [A+] and [A2+] are the concentrations of atomic and molecular ions, f+ = q/wi is the rate of ion formation (w, is the energy cost of forming an ion-electron pair). For Eq. (4.10), the solution has the simple form [2, 49]:

[A+] = f+Tic [1 — exp( tjTic)], Tic = k;c [A]2 , (4.13)

which was obtained given the condition that [A+] = 0 for t = 0.

In the plasma of gas NPLs, the characteristic times of all the main plasma processes are considerably less than the duration of the excitation pulse (minimal duration of reactor pulse is ~50 ps), so that for q « 0.01-5 kW/cm3, a quasi-stationary mode is established in the plasma over a time on the order of (f+kdr)-1/2 « 5 x 10-8­2 x 10-6 s. The concentration of electrons and ions can be found from the solution to Eqs. (4.10)-(4.12), if we equate the values of the derivatives with zero:

ne = f-T-c (vS+T + 1); [A+] = f+Tic;

[Д+]f (vffiT-1),S=f+rpd;. (4Л4)

The characteristics of the plasma being examined depend significantly on the dimensionless parameter S, the insertion of which makes it possible to carry out a
convenient classification of the excitation conditions. With so-called “weak” exci­tation, the condition S > 1 is fulfilled, and from the expression (4.14) we obtain:

n — [A+] = f >>[A+]. (4.15)

In the case of “weak” excitation, the non-stationary system of Eqs. (4.10)-(4.12) has an analytical solution [49]:

In the case of “strong” excitation (S ^ 1) we have:

If we assume S = 1, it is possible to obtain the expression for the degree of plasma ionization, Zws, which separates the “weak” and “strong” excitation regions:

For example, for argon at atmospheric pressure, — 5 x 10~4 (Te — 0.5-2 eV). In experiments with pulsed reactors, q < 5 kW/cm3 and Z < 10~5 ^ Zws, so that the estimate of parameters of the plasma of gas NPLs can be made in an approximation of the “weak” excitation.

For the conditions in which atmospheric pressure gas NPLs were studied with using of pulsed reactors, there is “weak” excitation (S > 103), so that for evaluative calculations of the plasma characteristics, one can use the formula (4.15). The validity of Eq. (4.15) is confirmed, for example, in experiment [50], where the electron concentration was measured for excitation of neon (PNe = 0.24 atm) by a proton beam with an energy of 20 MeV in a range q — 2 x 10~5-5 x 10~2 W/cm3 (S > 104).

The balance Eqs. (4.10)-(4.12) are written on the assumption that neutralization of the charged particles occurs exclusively owing to the dissociation recombination of molecular ions. This assumption is based on estimates of the rates of recombi­nation processes, and it was assumed that the values Te are known, and for gas NPLs are 0.5-2.0 eV.

To determine Te, it is necessary to consider the energy balance of plasma electrons. Studies [2, 49] analyzed the basic processes leading to establishment of some average energy of plasma electrons e. There are four such processes: (1) for­mation, in the interval 0 < £e < Im, of electrons with an average energy 7m/2, exceeding ee; (2) dissociative recombination resulting in the disappearance of the

slowest electrons; (3) inelastic electron-atom processes of atom excitation; (4) elas­tic electron-atom collisions. As a result of the first two processes, electrons are heated, and as a result of the latter two they are cooled.

For the region of “weak” excitation (S ^ 1), the electron energy balance equa­tion has the form [2]:

0

Table 4.9.

As a result of transformations, from expression (4.19) it is possible to obtain a rather simple transcendent equation:

Te = 300 K). The dimensionless constant C is equal to 0.015, 0.03, 0.053, and 0.06, respectively, for Ne, Ar, Kr, and Xe. For large values Z ^ 1, Eq. (4.20) is simpli­fied, since in this case F(Z) ^ 1 (Table 4.9):

e 1/g = Cg,0.5 — g,.

From solving of the approximate Eq. (4.21) given the condition that ne = [A^] (S ^ 1), it follows that the values g = Te/1m for Ne, Ar, Kr, and Xe differ insignif­icantly and are found in the range of 0.14-0.17. In particular, the calculation for Ar (C = 0.03; 1m « 11.5 eV) yields g « 0.16 and Te« 2.2 x 104 K.

From the cited results, it follows that in the case of “weak” excitation for Z ^ 10-4, the electron energy balance and Te scarcely depend on Z. In the opinion of the authors of study [2], this is explained by the fact that with an increase in the gas pressure, there is a sharp drop in the energy distribution function in the region of єе > 1m and consequently a reduction in the percentage of plasma electrons expending their energy on the excitation of atoms.

In studies [30, 31], the identical method was used to determine the energy deposi­tion to the gas. After irradiation of the laser cell with a short neutron pulse, damping pressure oscillations arose in the gas. From a comparison of the amplitude of the first peak of dependence of the pressure on the time with analogous amplitudes of dependencies obtained by means of computer simulation of gas dynamics (here the values of the energy transported to the gas are varied), the share of the energy of fission fragments transmitted to the gas was determined. In both studies, this percentage was roughly half the value obtained from direct calculation of the fragment energy deposition by the described methods. It is noted in [30], however, that direct measurements of the number of fission fragments emitted to the gas agree with the calculated values.

The conditions of conduct of experiments in these studies had some differences. For instance, in [30] a rectangular cell with plane-parallel arrangement of the layers containing uranium was used; the half-width of the neutron pulse was t1/2~0.15 ms; only argon was studied, with three values of initial pressure (0.7; 2.8; 4.14 atm). In [31], a cylindrical cell was used, with a uranium layer deposited on its internal surface (on a section with length 0.65 that of the total length). The half-width of the neutron pulse was t1/2 ~ 0.4 ms. Experiments were conducted with helium and neon with several values of pressure, in a range from 0.5 to 5 atm.

In study [31], the energy deposition was also evaluated from the pressure established after damping of the fluctuations. From the oscillograph curve shown in [31], it follows that the time of pressure establishment is ~10 ms. In the opinion of the authors [31], corrections allowing for the cooling of the gas through heat conduction were not great.

It should be noted that the calculations of [19, 20] (see also Chap. 8) of thermal and gas-dynamic processes in laser cells with the same internal diameter as in [31] (2.8 cm) showed that a relative reduction of pressure of ~20-35 % occurred in the gas due to heat removal from the walls of the cell after ~10 ms (from the start of the irradiating pulse with half-height duration of t1/2 ~ 3 ms). We also make mention of analogous calculations [34] for a rectangular cell with dimensions of 1 x 7 x 60 cm, from which it follows that in ~4 ms, around 10 % of the deposited energy departs owing to thermal conduction through the cell walls.

As one possible reason for the discrepancy of data between theory and experi­ment, it is proposed that a significant portion of the deposited energy does not have time to be thermalized, and is carried to the cell walls in the form of luminescent radiation [30], in particular through transformation into UV radiation of excimer molecules [31]. Another reason for the reduction in the energy deposition might be the low quality of fabrication of the active layers. Unfortunately, study [31] pro­vides no data on the specific energy deposition to the gas, which hampers analysis of the results of this study.

One of the main problems that arise when creating NPLs and RLs is the develop­ment of uranium layers that are stable relative to different influencing factors. These layers serve a dual function in RLs: they serve as reactor fuel and provide fission fragments to excite the laser medium.

Experience with NPLs using different types of uranium layers makes it possible to define their main performance requirements.

1. The uranium layer must be deposited on substrate made of a material that absorbs neutrons poorly, for example, aluminum or zirconium. The optimum thickness of the layer is about half the path length of the fission fragments in the layer. The density, thickness, and 235U concentration of the layer should be as uniform as possible.

2. The layer should be mechanically durable and thermally stable and have high adhesive bond to the substrate. Localized bond decrease between the layer and substrate disrupts heat removal and leads to the mechanical deterioration of the layer. This effect may arise with long-term storage of the layer and as a result of thermocyclic tension.

3. The “substrate-layer” boundary should be resistant to the thermal and radiation — stimulated diffusion of the layer material into the substrate.

4. To increase the operational lifetime of the NPL and RL, layers must be devel­oped with a minimum sputtering factor of the layer surface during output of the fission fragments to the gas. The sputtering factor (the average number of dispersed atoms from the surface per one emergent fragment) in a complex manner depends on the microstructure of the layer surface and changes within a wide range 1-104 atoms/fragment [20, 21].

The first experiments at VNIIEF on reactor VIR-2 (see Chap. 2, Sect. 2.3) used uranium oxide-protoxide layers manufactured using chemical deposition from solutions of uranium salts onto aluminum substrates. These layers have a loosely coupled structure and poor durability. This led to contamination of the laser channel, including the cavity mirror, with uranium dust.

To study NPLs, VNIIEF associates used uranium layers produced using the following technologies:

• Electric precipitation of uranium from a solution onto an aluminum substrate (see, for example, [22]), resulting in the formation of a layer that is a mixture of uranium oxide (UO2) and uranium protoxide (U3O8).

• Deposition of uranium from gas-phase compound onto a heated aluminum substrate (see, for example, [23]) with the formation of UO2.

• Magnetron ion-beam sputtering in a vacuum [24] resulting in the formation of metallic layers (a-uranium) or intermetallic compounds (UAl3 and UAl4). To protect the metallic uranium from corrosion and reduce the sputtering factor, the surface of the metallic uranium layer may be covered by a thin aluminum film with a thickness of about 0.5 pm.

The main characteristics of layers produced using the above two technologies are given in Table 10.4. Table 10.5 shows the comparative efficiency of different layers with 235U thicknesses of about 4.5 mg/cm2 (the efficiency of the layer was determined as the ratio of the power output by the fragments into the gas to the power released in the layer).

Each technology has its advantages. For example, magnetron sputtering yields metallic layers (a-uranium) with maximum efficiency. However, to prevent corro­sion of these layers, layer protection, which reduces efficiency, is necessary. Magnetron sputtering allows layers to be deposited in the intermetallic phase of uranium compounds (UA13 and UAl4), which have exceptionally high strength properties, but lower efficiency.

The adhesive strength of the uranium oxide layers is somewhat lower than that of the metallic layers. However, not one destruction has been noted in almost 20 years. Over this period, the VIR-2M/LUNA-2M experimental setup (see Chap. 2, Sect. 2.4) generated about 2,000 pulses with a total neutron fluence of ~1016/cm2. Moreover, no deterioration in NPL characteristics was observed. Tests [25] of layers made of metallic a-uranium showed that the adhesion of the layers does not diminish when the layer is kept for 10 years and after 150 heating cycles from 20 to 300 °C with a 5 min cycle time.

Oxide layers have a grain texture that is coarser than metallic layers. This is an advantage from the point of view of the suppression effect of sputtering. Metallic layers were studied by VNIIEF associates on the VIR-2M pulse reactor using mass spectrometry, electron microscopy and a — and y-spectrometry. These investigations established the following: sputtering of the layer mainly occurs by inelastic energy transfer from the fragment to the target atoms [20] and, for oxide layers, the sputtering factor is 103 atoms per fragment [25]. This factor may be reduced by about tenfold using special thermal processing [26]. Deposition of thin films (for example, 0.5 pm thick Al) to the surface of the uranium layer reduces the sputtering factor to several units.

Investigations of radiation-induced diffusion have shown [25] that this effect arises at a neutron fluence of about 2 x 1018/cm2. In this case, the uranium layer “descends” into the substrate to a depth of about 0.5 pm. This reduces its efficiency by 7 %.

Life tests of the metallic layers, which were performed on the RPT-6 reactor (NIIAR, Dimitrovgrad) up to a neutron fluence of ~1020/cm2, have shown that the uranium layers on the whole retain their initial texture to a 2 x 1019/cm2 fluence of thermal neutrons [27]; the operational life of metal layers with a protective alumi­num layer exceeds 0.1 % with respect to burn-up and, for intermetallic layers, 0.4 %. With this, the main characteristics of the layer remained highly stable. Potential methods for increasing the operational life of layers up to 1-2 % with respect to burn-up have been examined in study [28].

The problems of operational life and durability of uranium layers, discussed in this section do not belong among the principal physical problems determining the potential for creating RLs. However, they broadly impact the appearance of the RL and its possible applications.

Already in the first experiments performed with the VIR-2 reactor in 1972, the output power of the xenon laser with an optimal pressure and composition of the He-Xe mixture was 25 W with nl~0.5 % [1]. In subsequent experiments using the TIBR-1M, VIR-2, and VIR-2M reactors, lasing was obtained in a spectral range of 1.15-3.65 pm at 24 transitions of the Xe, Kr, and Ar atoms, with excitation of active

© Springer Science+Business Media New York 2015 53

S. P. Melnikov et al., Lasers with Nuclear Pumping,

DOI 10.1007/978-3-319-08882-2_3

.М3/: I

5*45/2],

5*45/2];

Fig. 3.1 (continued)

media by uranium fission fragments. Their spectral, energy, and threshold charac­teristics were studied in detail. In part these results are cited in the reviews [1—4]. Table 3.1 shows the basic data for all the lasers studied at VNIIEF operating on transitions of Xe, Kr, and Ar atoms with references to the original studies. Experimental units and methods are examined in Chap. 2, Sects. 2.3 and 2.4.

The majority of laser lines belong to the transitions nd-(n + 1)p of Xe, Kr, and Ar atoms (see Fig. 3.1). Apart from that, lasing was observed at 3.65 pm (transition 6p — 6s XeI); 3.07 and 2.86 pm (transitions 5p-5s KrI); 1.87 pm (transition 4d’-5p’ KrI); 2.87 pm (transition 5p’-5s’ ArI); 1.27 pm (transition 3d’-4p’ ArI) and 1.15 pm (transition 4p-4s ArI). We note that nearly all of the laser transitions shown in Table 3.1 were observed early in low-pressure gas discharge lasers [18] except for the 2.81 pm line of the Хе atom and the 1.87 pm line of the Kr atom. The mechanism of creating the population inversion on transitions of Xe, Kr, and Ar atoms is reviewed in Sect. 5.2 of Chap. 5.

The laser operating on transitions of the Xe atom has the maximal energy parameters. The active medium of this laser is mixtures of He-Xe, Ar-Xe, Kr-Xe, He-Ar-Xe, Ne-Ar-Xe and pure Xe at pressures <6 atm. Helium, neon, argon, and krypton are buffer gases, while the concentration of xenon is 1-10 %. Depending on the conditions of the experiment (pressure and composition of mixtures, power deposition, properties of the cavity mirrors), lasing was observed at different lines in a spectral range of 1.7-3.5 pm.

The maximal energy parameters (output power Wout < 2 kW, ni < 2.5 %) were registered at the 1.73; 2.03 and 2.65 pm lines of the Xe atom, which originate from the level 5d[3/2]10. For these same lines, the minimal laser thresholds Ф^ < 1013 cm-2 s-1 were observed. The lowest laser threshold, at Ф^ = 1.5 x 1012 cm-2 s-1 (specific power deposition at laser threshold qth~ 0.1 W/cm3) was registered for mixtures of Ar-Xe (A = 2.03 pm) [6, 7]. Such low thresholds make it possible to use not only neutron radiation of stationary nuclear reactors to pump an Ar-Xe laser, but also radioisotope sources [7].

NPLs operating on transitions of Kr and Ar atoms have lower energy parame­ters—in pumping mixtures of He(Ne)-Ar and He-(Ne)-Kr at atmospheric pressure with uranium fission fragments, щ < 1 % was obtained. Among NPLs using Ar atom transitions, one should note the He-Ar laser (A = 1.15 pm) [8, 16], in which quasi-CW lasing occurs at transition 4p[1l/2]1-4s'[1/2]1° as a result of collision “quenching” of the lower metastable level 4s'[1/2]1° by atoms of the buffer gas, helium.

In experiments using an Ar-Xe mixture carried out on the LUNA-2M setup in 1985, a competition effect was observed among the laser lines 1.73; 2.03 and 2.65 pm of the Xe atom, which have a common upper laser level. Figure 3.2 shows oscillograms of pulses of neutron and laser radiation for the mixtures He-Xe, Ar-Xe, and Ar-Xe-He [6]. In the He-Xe mixture, lasing occurs only at A = 2.03 pm (Fig. 3.2а). In the mixture Ar-Xe, lasing first occurs at A = 2.03 pm. With an increase in the pumping power, the line 1.73 pm appears, and there is stopping of lasing at A = 2.03 pm. The 2.03-pm line arises again at the end of the pumping pulse after cessation of lasing at A = 1.73 pm. Small additions of helium

LA

VO

Note: P pressure of active medium, W power of laser radiation (output power),power efficiency (ratio of output power to power deposition), thermal — neutron flux density averaged over active medium length, at which the laser threshold is achieved (threshold thermal-neutron flux density), Ф„ил maximum thermal-neutron flux density averaged over active medium length, at the reactor pulse maximum

Fig. 3.2 Oscillograms of thermal neutron pulse (a) and of laser pulses: (b) mixture of He-Xe (1,000:1), Р = 2 atm; (c) mixture of Ar-Xe (100:1), Р = 0.5 atm; (d) mixture of Ar-Xe-He (100:1:100), Р = 1 atm [6]

(0.25-0.5 atm), which insignificantly affect the power deposition, lead to elimina­tion of the 1.73-qm line (Fig. 3.2d).

Competition of laser lines having the common upper laser level 5d[3/2]10 may be explained by the differences in probabilities of radiative decay and the rate con­stants of the processes of collision “quenching” of the lower laser levels by atoms of the buffer gas. An analogous effect was observed in the Ar-Xe mixture for two other lines of the Xe atom (2.65 and 1.73 qm; 2.03 and 2.65 qm) [6, 7], in the Kr-Xe mixture for the 2.63 and 2.81 qm lines of the Xe atom [9], as well as in the mixture Ne-Ar for the 2.31 and 2.21 qm lines of the Ar atom [15]. Competition of lines was also observed in excitation of the mixture Ar-Xe by an electron beam [19, 20]. The most complete analysis of the different variants of competition of laser lines having a common upper or lower laser level was done in studies [21—23].

For a more detailed study of xenon NPLs using mixtures of He-Xe (1.73; 2.03 and 2.65 qm), Ar-Xe (1.73 qm), and He-Ar-Xe (2.03 and 2.65 qm), a series of

experiments was carried out on the LUNA-2M setup to measure the energy and threshold characteristics as a function of the partial pressure of the xenon with various transmission coefficients of the cavity mirrors [12, 13, 24]. Narrow-band dielectric mirrors were used in the experiments, which made it possible to preclude simultaneous lasing at several lines. The basic goal of the experiments was to obtain information about the parameters of the active media of xenon NPLs (small-signal gain a0, saturation intensity Is, and coefficient of distributed losses P) as a function of the partial pressure of the xenon. This information, obtained as a result of processing of experimental data using an approximation by Rigrod’s formula [25], is shown in Table 3.2 for q = 40 W/cm3 (maximum of pumping pulse) and optimal partial pressures of xenon.

To study the parameters of the active medium of a xenon NPL, a direct method of measurement by the “oscillator-amplifier” scheme was also used [26]. Experi­ments were carried out on the LUNA-2M setup for the mixture He-Ar-Xe (380:380:1) at a pressure of 1 atm at the 2.03 pm line. The small-signal gain increases approximately linearly with an increase in the specific power deposition, and at a maximal value of q = 40 W/cm3 reaches 1.2 x 10~2, while the saturation intensity is virtually independent of q and is 70-90 W/cm2. The output laser power at the pumping pulse maximum at the outlet from the amplification channel (790 W) proved to be virtually twice as high as at the input (400 W).

In the experiments considered above, simple, stable two-mirror cavities were used. The presence of two identical laser channels in the LUNA-2M setup makes it possible to study more complex schemes for combining the NPL channels and the radiation withdrawal, which is of interest for forming output radiation of multichannel laser facilities, particularly to reduce the number of light beams (see Chap. 10). VNIIEF proposed two methods of combining the NPL channels, which by analogy with the parallel and serial connection of elements in electrical circuits, can be called “serial” and “parallel” combining of laser channels (see Fig. 10.7). The results of computational and experimental investigations in this area are shown in studies [27—30].

The basic results of investigations on combining the radiation of laser channels on the LUNA-2M setup are shown in Table 3.3. The active laser media used were mixtures of He-Ar-Xe (380:380:1) at a pressure of 1 atm (A = 2.03 pm) and Ar-Xe

(380:1) at a pressure of 0.5 atm (A = 1.73 qm). The active volumes of the laser channels were identical and equal to 1.9 l. It is clear from Table 3.3 that the output power of two serially combined channels (A = 2.03 qm) and parallel combined channels (A = 1.73 qm) is virtually twice the output power of a single channel. It should be noted that optical non-uniformities arising in the laser medium in the process of the pumping pulse have a significant effect on the laser parameters of the double laser channel with serial combining. This is particularly distinctly demon­strated for the mixture Ar-Xe. In the case of the parallel combining, the shape of the laser pulse of several combined channels does not depend on the number of channels, but the output power and energy are proportional to the number of combined channels. This is explained by the fact that with parallel combining (in contrast to serial), there is no accumulation of optical non-uniformities. The parallel scheme is therefore very promising for use in multichannel NPLs.