1629373397-425d4de58b7aea127ffc7c337418ea8d (846389), страница 65
Текст из файла (страница 65)
10.57 Schematic of direct-drive laser fusiongrow into large gouges, the laser light is carefully smoothed out by random phaseplates. The layer of frozen DT is serendipitously smoothed out by the small amountof heat from the slow decay of tritium into 3He, an electron, and a neutrino. In theU.S., this approach is followed at the Laboratory for Laser Energetics in Rochester,NY.In indirect-drive fusion, the capsule is compressed by X-rays inside a goldcylinder called a hohlraum (“empty space” in German).
Figure 10.58 shows ahohlraum with laser beams entering holes at the ends. The beams must not hit thesides of the hole, or they would create plasma that will refract or reflect the beams.Once inside the hohlraum, the beams strike the walls, creating a bath of X-rays,which will then heat the capsule and compress it. This is a miniature version of anH-bomb. Several views of hohlraums are shown in Fig. 10.59.In the U.S., indirect-drive fusion is explored at the NIF at Livermore.
Thisfacility is a marvel of engineering. Light from a single oscillator is preamplifiedand then split into 192 beams. The beams are amplified in Nd-glass plates lit byLED flashlamps powered by 422 MJ of capacitors. Faraday rotators and polarizersare inserted to prevent reflected light from being amplified backward anddestroying the front end. Since parametric instabilities are less important at shorterwavelengths, the 1.05 μm Nd-YAG (neodymium-doped yttrium aluminum garnet)beams are frequency tripled by KDP (potassium dihydrogen phosphate) crystals to351 nm in the green. A single dust particle on one of the glass plates will absorbenough laser light to damage the plate, which must be replaced.
Refrigerator-sizedreplacement units are stored below beam lines and can be inserted into place. Anearly picture of the NIF laser bay is shown in Fig. 10.60.The laser beams are mirrored in a switchyard to enter into a spherical targetchamber, shown in Fig. 10.61. They must all reach the target within a fewpicoseconds. Light travels only 0.3 mm in a picosecond. Hence the lengths in thissystem the size of a football field have to be kept constant with temperature control.10.4Inertial Fusion395Fig. 10.58 Drawing of a hohlraum [courtesy of Lawrence Livermore Laboratory via GoogleImages]Fig. 10.59 Hohlraums: (a) a hohlraum in its holder; (b) size of a hohlraum; (c) a “rugby”hohlraum with an ellipsoidal cavity; (d) suspension of a capsule in a hohlraum with thin plasticmembranes or “tents”39610Plasma ApplicationsFig. 10.60 The laser bay of the National Ignition Facility [courtesy of Lawrence LivermoreLaboratory via Google Images]Fig.
10.61 The NIF target chamber, seen from inside [courtesy of Lawrence Livermore Laboratory via Google Images]. Note the figure of a man above center10.4Inertial Fusion397The beams carry 15 MJ of 351 nm light, giving a peak power of 430 TW. The totalelectricity generation in the U.S. is about 0.5 TW on average.Impressive as these numbers are, fusion with glass lasers is not an energy source.The goal of ignition, in which the alpha particles from the DT reaction keeps theplasma hot, has not been achieved. The most that has been achieved by 2015 is fuelgain, in which the fusion energy out is larger than the laser energy in, with a laserefficiency of about 1%.
Furthermore, the NIF can pulse only twice a day to allowthe glass to cool. A fusion reactor would require pulsing at 10 Hz. Each shotdestroys a target such as that shown in Fig. 10.59. The hohlraum cannot simplybe dropped from the top of the chamber; the DT ice would melt. The target has to beshot into the lasers’ focus in exactly the right orientation. After each shot, the debrishas to be cleared so as not to interfere with the next laser shot.
Heavy products likegold take a long time to reach the wall. Liquid Li walls or FLiBe (fluorine, lithium,beryllium) waterfalls have been suggested to catch the gamma rays and otherproducts but they cannot clear the whole volume rapidly. The NIF cannot lead toa fusion reactor. It is funded by the Nuclear Weapons Stockpile Stewardshipprogram in addition to the Department of Energy and serves as a safe way tostudy such topics as the equation of state under extreme conditions without usingunderground nuclear explosions.In addition to accurate focusing, the timing of the laser pulses is crucial.
Asimple intense laser pulse can drive the fuel into a center hotspot until the ρrcriterion is exceeded, but one can do much better than that. Here are two ways. Infast ignition, the compression and heating phases follow each other, just as in agasoline engine. First, one laser pulse compresses the fuel to the required density;then a second high-intensity, ultrashort pulse heats the core. The compression pulsecan have less energy and uniformity than usual, so that the energy gain is 10–20times higher than normal (Fig.
10.62).In shock ignition, the general principle is to launch a series of shocks whichovertake one another at the central spot. In NIF, this is done in just two steps with asingle laser. Figure 10.63a shows the timing compared with that of a conventionalsingle pulse. First, there is a low-velocity compression, which entrains more fuelmass and saves energy compared the normal procedure. This is followed by a strongFig.
10.62 Stages of fast ignition [modified from Google Images; originally from Los AlamosNational Laboratory’s Dense Plasma Theory]39810Plasma ApplicationsFig. 10.63 (a) Time sequence of a shock ignition pulse, compared to a conventional pulse. Thetotal time has to be short enough to avoid heating the hohlraum [L.J. Perkins et al., Phys. Rev. Lett.103, 045004 (2009)]. (b) Design of a capsule for shock ignitionFig. 10.64 Computedfusion energy from shockignition compared withnormal compression [L.J.Perkins et al., loc. cit.]shock pulse timed to arrive when the compression reaches stagnation at a velocityof about 4 107 cm/s.
The shock contains 20–30 % of the energy of the pulse.The hope is to generate 120–150 MJ of fusion power using only 1–1.6 MJ of laserpower. Figure 10.63b is a possible design of a capsule suitable for shock ignition,and Fig. 10.64 shows the expected fusion energy vs. laser energy.To improve the twice-a-day repetition rate, plans are to construct a HighRepetition Rate Advanced Petawatt Laser System (HAPLS), designed in collaboration with the Czech Institute of Physics. The laser would produce 1 PW (1015 W)in 30 fs pulses at 10 Hz.
In collaboration which Czech scientists, this would beinstalled at the Extreme Light Infrastructure Beamlines facility (ELI-Beamlines),under construction in Dolnı́ Breźańy near Prague in the Czech Republic. Such highrep-rate lasers are already available in KrF (krypton-fluoride) lasers.10.5Semiconductor Etching399InputLaserPulsedPowerRecirculatorBzCathodeElectronFoilBeam SupportLaser Cell(Kr+F2)AmplifierWindowFig. 10.65 Diagram of a krypton-fluoride laser amplifier [J.D. Sethian et al., Phys.
Plasmas10, 2142 (2003)]10.4.2 KrF LasersKrypton fluoride lasers, such as the Electra laser at the Naval Research Laboratoryin the U.S., can pulse at 5 Hz at 400–700 J to give almost steady-state power. Thetechnology can be extrapolated to 106 shots at 50–150 kJ for laser fusion. In thediagram of Fig. 10.65, an electron beam produced by pulsed power (Sect. 10.2.1.1)passes through a 1-mil (0.001 in.) Ti foil to ionize a cell containing Kr and F2.
Therecirculator reflects the electrons to use them a second time. The KrF wavelength of248 nm is even shorter than the 3ω, 351 nm wavelength of Nd-YAG lasers and issuitable for minimizing parametric instabilities. Though KrF lasers have muchlower power than glass lasers, they can provide the rep rate needed for fusion.10.5Semiconductor EtchingAll the computers, iPhones, iPads, and tablets are based on a semiconductor chipabout 1 cm2 in size and containing billions of transistors. The doubling of thetransistor count every 2 years is known as Moore’s Law, shown in Fig. 10.66. A“chip” can contain several cores, or Central Processing Units (CPUs).
The progressis so rapid that in 1989 the largest Intel chip contained 1.18 million transistors witha critical dimension (CD) of 1 μm for its smallest feature. In 2014, the countincreased to 5.56 billion transistors with a CD of 22 nm (0.022 μm).Chips cannot be made without plasmas. The transistors lie on several layers, andconnections are made between layers with copper or aluminum conductors40010Plasma Applications10,000,000,000Years to doubleNo.
of transistors per chip1,000,000,000Itanium 2 Dual1.82.02.2100,000,000Core 2 QuadItanium 2TukwilaAMD RV770Core 2 DuoPentium M10,000,000Pentium1,000,000486386100,000286808610,00080801,000197019801990200020102020YearFig. 10.66 Moore’s law of transistors on a chipFig. 10.67 Cross section ofa transistor on apolycrystalline substratedeposited into channels called vias. The transistors themselves are made byrepeated plasma etching and deposition steps.
Some transistors, called FinFETs,are themselves three-dimensional. Hundreds of chips are made at once on a300-mm wafer of single-crystal silicon. The plasma must be uniform over thewafer. Chips near the boundary are usually not as good as those inside and couldbe sold as lower-frequency chips. The industry is pursuing the next step of 450-mmwafers, which will greatly increase the yield but would require an even largeruniform plasma.The structure of a transistor is shown in Fig. 10.67. Its function is to control anelectrical current with a gate that requires very little power, just as a dam cancontrol a large amount of hydro power by opening and closing an actual gate. In10.5Semiconductor Etching401Fig.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
