A Few Practical Laser Examples

This is a continuation from the previous tutorial - laser output-beam properties.

Let us look at just a few practical examples of real lasers that illustrate some of the points we have been discussing, notably the ruby solid-state laser, and the helium-neon gas laser.

The Ruby Laser

The first laser of any type ever to be operated was in fact the flash-pumped ruby laser demonstrated by T. H. Maiman at the Hughes Research Laboratory in early 1960.

We have already shown in Figure 1.10 [refer to the atomic energy levels and spontaneous emission tutorial] the quantum energy levels associated with the unfilled \(3d\) inner shell of a Cr³⁺ ion when this ion replaces one of the Al³⁺ ions in the sapphire or AI₂O₃ lattice. Up to ~ 1% of such replacements can be made in the sapphire lattice to create pink ruby.

By placing such a ruby rod shaped roughly like a slightly overweight cigarette inside a spiral flashlamp filled with a few hundred Torr of xenon (Figure 1.50), and then discharging a high-voltage capacitor bank through this lamp, Maiman was able to use the blue and green wavelengths from this lamp to optically pump atoms from the \(^4A_2\) ground level of the Cr³⁺ ions in the lattice into the broad \(^4F_2\) and \(^4F_1\) bands of excited levels. In ruby, atoms excited into these levels will relax very rapidly, and with close to 100% quantum efficiency, down into the comparatively very sharp \(^2E\) levels, or \(R_1\) and \(R_2\) levels, lying ~14,400 cm or 694 nm (~1.8 eV) above the ground level.

The ruby laser is, however, a three-level laser system, in which the lower laser level is also the ground energy level. By pumping hard enough, we can nonetheless cycle more than half of the Cr³⁺ ions from the ground level up through the pumping bands and into the highly metastable upper laser level, with its fluorescent lifetime of \(\tau\approx\) 4.3 msec. Thus, even though ruby is a three-level system rather than a four-level system, which is usually very unfavorable, with sufficiently hard pumping Maiman was able to produce a powerful burst of laser action from the ruby rod.

In a small flash-pumped laser such as ruby, or others, the flashlamp may be connected to a capacitor bank of perhaps 10 to 100 microfarads charged to a prebreakdown voltage of perhaps 1,000 to 1,500 volts, corresponding to ~5 to 50 J of stored energy.

The lamp itself is then triggered or ionized by a high-voltage pulse, so that it becomes conducting. The capacitor energy then discharges through the lamp with a typical pulse length of perhaps 200 μsec, peak currents of up to a few hundred amperes, and peak electrical power input of 25 to 250 kW.

The laser rod may convert the pump light in a typical solid-state laser into laser energy with ~ 1% efficiency, leading to laser output energies of 50 mJ to 0.5 J per shot, and average powers during the pulse of 2.5 to 25 kW. (We will discuss later the technique of "Q-switching," which can extract the same laser energy in a very much shorter pulse with very much higher peak power.)

The laser action in ruby actually occurs not as a clean and continuous laser action during the pulse, but as a series of short "spikes" or relaxation-oscillation bursts during the entire pumping time (see Figure 1.51). We will discuss this spiking behavior in more detail in a later tutorial.

Other Solid-State Lasers

There are many such solid-state lasers besides ruby (though unfortunately not many in the visible region). The most common of these are the rare-earth ions in crystals or glasses, with by far the most widely used examples being Nd³⁺ lasers using Nd:YAG (Nd³⁺ ions in yttrium aluminum garnet) and Nd:glass materials.

The spiral flashlamp and diffusely reflecting pump enclosure used in Maiman's first ruby laser is now almost always replaced by one or more straight lamps placed parallel to the rod along the axes of an elliptical pump cavity (Figure 1.52).

In the first ruby lasers, partially transparent metallic silver mirrors were evaporated directly onto the polished ends of the laser rod (though such metallic mirrors are quite sensitive to optical damage at higher powers).

Later solid-state lasers quickly shifted to the use of external dielectric-coated mirrors, just as in gas lasers. The round-trip gains in ruby and other solid-state lasers are often much higher than in gas lasers—up to round-trip power gains of 10X and higher—so that mirrors with much lower reflectivity or higher transmission output can be employed.

Pulsed solid-state lasers are used for a variety of smaller-scale laser cutting, drilling, and marking applications; as military rangefinders and target designators; and in an enormous variety of scientific and technological experiments.

By taking advantage of improved lamp efficiencies and laser materials, as well as the fact that most other materials are four-level lasers, we can also operate several solid-state lasers continuously at cw power outputs in the 1-100 W range with efficiencies of ~ 1% or slightly higher, using electrical inputs of 100 W to 10 kW into xenon or krypton-filled arc lamps. (Both laser rod and lamps must, of course, be carefully water-cooled.) Even ruby can, with some difficulty, be made to oscillate on a cw basis. We will discuss the very useful Nd³⁺ laser system in detail in later tutorials.

The Helium-Neon Laser

Another of the most common and familiar types of laser is the helium-neon gas laser developed at the Bell Telephone Laboratories in 1960 and 1961. The laser tube in a He-Ne laser consists of a few Torr of helium combined with approximately one-tenth that pressure of neon inside a quartz plasma discharge tube, which is usually provided with an aluminum cold cathode and an anode, as in Figure 1.53.

This discharge tube may be 10 to 50 cm long and a few mm in diameter in a typical small laser. To avoid broadening of the laser transition by isotope shifts (and for other more complex reasons), a mixture of single-isotope He³ and Ne²⁰ is usually employed; and it is found empirically that the optimum pressure-diameter product \(pd\) in such a laser is a few Torr-mm and that the optimum gain per unit length varies inversely with tube diameter \(d\).

This tube is then excited with a dc discharge voltage typically of order 1,000 to 1,500 vdc, producing a dc current typically of order ~10 mA from a special high-purity aluminum cold cathode. (Radio-frequency excitation through external electrodes was also employed in many early lasers, but has been found to be generally less convenient.)

Because a dc glow discharge in this pressure range has a negative-resistance I-V curve (Figure 1.54), a ballast resistance in series with the dc voltage supply is necessary to stabilize the discharge; and an initial higher-voltage spike must be supplied to ionize the gas and break down the gas discharge each time the tube is turned on. 

The discharge tubes in many gas lasers (especially with longer lasers, or lasers for research purposes) may be provided with Brewster-angle end windows which transmit light of the proper linear polarization with essentially zero reflection loss at either face. (Because of the very low gain in the He-Ne system, reflection losses of several percent at each of the air-dielectric interfaces would be totally intolerable.)

In many small inexpensive internal-mirror He-Ne lasers, however, the end mirrors are sealed directly onto the discharge tube, as part of the laser structure (Figure 1.55). Extreme cleanliness and purity of the laser gas fill is vital in the inherently low-gain He-Ne system; the tube envelope must be very carefully outgassed during fabrication, and a special aluminum cathode employed, at least in long-lived sealed-off lasers.

The end mirrors themselves are carefully polished flat or curved mirrors with multilayer evaporated dielectric coatings, having as many as 21 carefully designed and evaporated layers to give power reflectivities in excess of 99.5% in some cases.

The pumping mechanism in the He-Ne laser is slightly more complex than those we have discussed so far. The helium gas, as the majority component, dominates the discharge properties of the He-Ne laser tube.

Helium atoms have in fact two very long-lived or metastable energy levels, generally referred to as the \(2^1S\) ("2-singlet-S") and \(2^3S\) ("2-triplet-S") metastable levels, located ~20 eV above the helium ground level.

Free electrons that are accelerated by the axial voltage in the laser tube and that collide with ground-state neutral helium atoms in the laser tube then can excite helium atoms up into these metastable levels, where they remain for long times.

There is then a fortuitous—and very fortunate-near coincidence in energy between each of these helium metastable levels and certain sublevels within the so-called \(2s\) and \(3s\) groups of excited levels of the neutral neon atoms, as shown in Figure 1.56. (The atomic energy levels in neon, as in other gases, are commonly labeled by means of several different forms of spectroscopic notation of various degrees of obscurity.)

When an excited He atom in one of the metastable levels collides with a ground-state Ne atom, the excited He atom may drop down and give up its energy, while the Ne atom simultaneously takes up almost exactly the same amount of energy and is thus excited upward to its near-coincident energy level.

This important type of collision and energy-exchange process between the He and Ne atoms is commonly referred to as a "collision of the second kind." Any small energy defect in the process is taken up by small changes in the kinetic energy of motion of one or the other atom.

This process thus amounts to a selective pumping process, carried out via the helium atoms, which efficiently pumps neon atoms into certain specified excited energy levels. As Figure 1.56 shows, laser action is then potentially possible from these levels into various lower energy levels in the so-called \(2p\) and \(3p\) groups.

The first successful laser action in any gas laser was in fact accomplished by A. Javan and co-workers at Bell Labs in late 1960 on the \(2s_2\rightarrow2p_4\) transition of helium-neon at 1.1523 microns in the near infrared.

Shortly thereafter A. D. White and J. D. Rigden discovered that the same system would lase on the familiar and very useful \(3s_2\rightarrow2p_4\) visible red transition at 633 nm (or 6328 Å), as well as on a much stronger and quite high-gain set of \(3s\rightarrow3p\) transitions near 3.39 microns.

(A half-dozen or so different nearby transitions within each of these groups can actually be made to lase, with the strongest transition in each group being determined in part by the relative pumping efficiencies into each sublevel and in part by the relative transition strengths of the different transitions.)

Characteristics of Gas Lasers

The laser gain in the He-Ne 633 nm system is quite low, with perhaps \(2\alpha_m\approx0.02\) to \(0.1\) \(\text{cm}^{-1}\) (often expressed as "2% to 10% gain per meter"); and the typical power output from a small He-Ne laser may be 0.5 to 2.0 mW. With a dc power input of ~10 W, this corresponds to an efficiency of ~0.01%.

Several manufacturers supply inexpensive self-contained laser tubes of this type for about $100 retail and considerably less in volume production. Such lasers are very useful as alignment tools in surveying, for industrial and scientific alignment purposes, supermarket scanners, video disk players, laser printers, and the like.

(The dominance of the He-Ne laser in such applications may soon be ended by even cheaper and simpler semiconductor injection lasers.) Larger He-Ne lasers with lengths of 1 to 2 meters that can yield up to 100 mW output at comparable efficiencies are also available.

There are also scores of other gas lasers that are excited by using electrical glow discharges, higher-current arc discharges, hollow-cathode discharges, and transverse arc discharges.

One notable family of such lasers are the rare-gas ion lasers, including argon, krypton, and xenon ion lasers, in which much larger electron discharge currents passing through, for example, a He-Ar mixture can directly excite very high-lying argon levels to produce laser action in both singly ionized Ar⁺ and doubly ionized Ar⁺⁺ ions.

Such ion lasers are generally larger than the He-Ne lasers, and even less efficient, but when heavily driven can produce from hundreds of milliwatts to watts of cw oscillation at various wavelengths in the near infrared, visible, and near ultraviolet. Longer-wavelength molecular lasers, such as the CO₂ laser, and shorter-wavelength excimer lasers are other examples of important gas laser systems.

Other Properties of Real Lasers

Practical lasers in fact come in a great variety of forms and types, using many different kinds of atoms, molecules, and ions, in the form of gases, liquids, crystals, glasses, plastics, and semiconductors. These systems oscillate at a great many different wavelengths, using many different pumping mechanisms. Nearly all real lasers have, however, certain useful properties in common.

Temporal and Spatial Coherence

As we have discussed in some detail in earlier sections, nearly all lasers can be:

(a) Very monochromatic.

Real laser oscillators can in certain near-ideal situations oscillate in a single, essentially discrete oscillation frequency, exactly like a coherent single-frequency electronic oscillator in more-familiar frequency ranges.

This oscillation will, as with any other real oscillator, still have some very small residual frequency or phase modulation and drift, because of mechanical vibrations and thermal expansion of the laser structure and other noise effects, as well as small amplitude fluctuations due to power supply ripple and the like. Such a high-quality laser can still be, however, one of the most spectrally pure oscillators available in any frequency range.

More typically, a real laser device will oscillate in some number of discrete frequencies, ranging from perhaps 5 or 10 simultaneous discrete axial modes in narrower-line lasers up to a few thousand discrete and closely spaced frequencies in less well-behaved lasers with wider atomic linewidths.

Real lasers will also in many cases jump more or less randomly from one oscillation frequency to another, and the amplitudes and phases of individual modes will fluctuate randomly, because of mode competition combined with the kinds of unavoidable mechanical and electronic perturbations mentioned above.

Nonetheless, the degree of temporal coherence in even a rather bad laser will generally be much higher than in any purely thermal or incoherent light source, and especially in any thermal source providing anywhere near the same power output as the laser's oscillation output power. 

(b) Very directional.

The output beam from a typical real laser will also be very directional and spatially coherent. This occurs because, with properly designed mirrors, many lasers can oscillate in a cavity resonance mode which is essentially a single transverse mode; and this mode can approximate a more or less ideal quasi-plane wave bouncing back and forth between carefully aligned end mirrors.

As we discussed in the preceding section, the resulting output beam from the laser can then be a highly collimated or highly directional beam, which can also be focused to a very tiny spot. Such a beam can be projected for long distances with the minimum amount of diffraction spreading allowed by electromagnetic theory. It can also be focused to a spot only a few wavelengths in diameter, permitting all the power in the laser beam to be focused onto an extremely small area.

Even lasers with nonideal spatial properties (perhaps because of distorted laser mirrors or, more commonly, because of optical aberrations and distortions in the laser medium or in other elements inside the laser cavity) will typically oscillate in only some moderate number of transverse modes, representing some lowest-order transverse mode and a number of more complicated higher-order transverse modes.

Note again that the longitudinal-mode or frequency properties and the transverse-
mode or spatial properties of most laser oscillators are more or less independent, so that, for example, even wide-line or multifrequency lasers can very often have well-controlled transverse mode properties and can oscillate in a nearly ideal single transverse mode.

Other Real Laser Properties

Besides these two basic properties, specific individual lasers can be:

(c) Very powerful.

Continuous powers of kilowatts or even hundreds of kilowatts are obtained from some lasers, and peak pulse powers exceeding 10¹³ Watts are generated by other lasers. (It is interesting to note that this peak power is an order of magnitude more than the total electrical power-generating capacity of the United States—but of course for a very short time only.)

(d) Very frequency-stable.

Both the spectral purity and the absolute frequency stability of certain lasers can equal or surpass that of any other electronic oscillator; so these lasers can provide an absolute wavelength standard with an accuracy exceeding that of any other presently known technique.

(e) Very widely tunable.

Although most common lasers are limited to fairly sharply defined discrete frequencies, those of the spectral lines of the specific atoms employed in certain lasers (e.g., organic dye lasers and to a lesser extent semiconductor lasers) can be tuned over enormous wavelength ranges, and so are extremely useful for spectroscopic and chemical applications.

(f) Very broadband.

Many laser transitions, though very narrowband in fractional terms, have extremely wide linewidths compared to those of conventional radio or microwave frequencies. Hence, such lasers can provide very broadband amplification and, more important, can generate and amplify extraordinarily short optical pulses. Laser pulses as short as a few picoseconds in duration are relatively commonplace, and some mode-locked lasers can generate light pulses as short as 30 femtoseconds (30 x 10^-15 seconds) in length.

(g) Very efficient.

The efficiency of most common lasers is smaller than designers would like, ranging from ~ 0.001 to 0.1% in many gas lasers, up to typically 1 or 2% in optically pumped solid-state lasers. A few selected lasers, such as the CO₂ laser and the semiconductor injection lasers, can have efficiencies as high as 50% to 70% in converting electrical power directly into coherent radiation.

Examples of Practical Laser Systems

It is impossible to catalog all the laser devices that have been demonstrated to date, especially since the variety of laser materials, laser pumping methods, and laser experimental techniques is almost endless.

Some laser systems that are particularly well-known, particularly useful for practical applications, or particularly interesting for other reasons will be discussed in more detail later in this tutorial series.

The overall situation at present is that at least 10⁵, and up to 10⁶, distinct laser transitions that have been demonstrated, at wavelengths ranging from \(\lambda\ge\) 600 μm (0.6 mm) in the far infrared to the present short-wavelength record of \(\lambda\) = 1160Å from the pulsed-discharge H₂ laser in the near ultraviolet.

At still longer wavelengths, besides more familiar vacuum tubes and semiconductor devices, there are several varieties of millimeter-wave and microwave masers, including molecular beam masers, solid-state electron paramagnetic resonance masers, and nuclear magnetic resonance (NMR) masers.

These last devices in fact carry the stimulated-emission principle down to frequencies below 100 Hz. Lasers operating in the X-ray region do not yet seem to have been successfully demonstrated, though several candidates in the soft X-ray region (100-200Å) appear very promising.

A critical study by Bennett in 1979 identified a total of 1,329 distinct laser wavelengths coming from 51 different elements, considering only neutral atoms and ions in gases (no molecules). This data was culled from a source file of 30,000 (!) literature articles.

There were, for example, 203 identified laser lines from neutral neon alone, grouped into the clusters of transitions shown in Figure 1.57. Another study identified 270 new lasing lines on various vibrational-rotational transitions in a limited wavelength range for the CO molecule alone.

When we consider the enormous diversity of potential molecular species, and the very large number of distinct rotational-vibrational transitions in any one such molecule, it is not impossible that the number of potential distinct molecular laser lines could exceed one million.

Laser action has been obtained thus far in atoms, molecules, and ions in vapor (gas) phase; in atoms, ions, or molecules in crystals, glasses, and liquid solutions; in organic dye molecules in liquids, vapors, gels, and plastics; in semiconductors of several varieties; and in molecules and molecular radicals in planetary atmospheres and in interstellar space.

Commercially Available Lasers

Of the laser systems mentioned so far, those that are now readily available in routine commercial production include the He-Ne 633 nm laser; many different sizes of both cw and TEA CO₂ lasers at 9 to 11 μm; various argon, krypton, and other noble-gas ion lasers in the visible and UV; the pulsed N₂ laser at 377 nm; the blue cadmium ion laser; the Nd:YAG laser (including many Q-switched, mode-locked, and wavelength-doubled versions); similar ruby, Nd:glass, and alexandrite solid-state lasers; various KrF and other excimer lasers; several varieties of flash-pumped, N₂-laser-pumped, YAG-laser-pumped, and cw-argon-laser-pumped tunable dye lasers; and of course many versions of the GaAs injection laser.

In addition there is much development work in government, industrial, and university laboratories on large Nd:glass laser systems and on the atomic iodine laser for laser fusion systems, and on various chemical lasers (HF, DF, CO, CO₂) for military applications.

There are also development efforts to a lesser extent on the copper vapor laser, various hollow-cathode visible gas lasers, and a few others. Most of the other known laser systems are available only as (expensive) custom prototypes, or by constructing one's own "home-built" version. (Many chemists, biologists, solid-state physicists, and spectroscopists have now become expert amateur laser builders.) 

Commercial development of many other lasers has been rather slow, because the expensive engineering effort to develop a commercially engineered product cannot be justified until a market has been clearly identified. At the same time, commercially significant applications for certain lasers cannot be easily developed if the lasers are not available in commercially developed form. 

Laser-Pumping Methods

The list of successful laser-pumping methods that have been demonstrated to date includes the following. 

• Gas discharges, both dc, rf,.and pulsed, including glow discharges, hollow cathodes, arc discharges, and many kinds of pulsed axial and transverse discharges, and involving both direct electron excitation and two-stage collision pumping.
• Optical pumping, using flashlamps, arc lamps (pulsed or dc), tungsten lamps, semiconductor LEDs, explosions and exploding wires, other lasers, and even gas flames and direct sunlight.
• Chemical reactions, including chemical mixing, flash photolysis, and direct laser action in flames. It is instructive to realize that the combustion of one kg of fuel can produce enough excited molecules to yield several hundred kilojoules of laser output. A chemical laser burning one kg per second, especially if combined with a supersonic expansion nozzle, can thus provide several hundred kW of cw laser output from what becomes essentially a small "jet-engine laser."
• Direct electrical pumping, including high-voltage electron beams directed into high-pressure gas cells, and direct current injection into semiconductor injection lasers.
• Nuclear pumping of gases by nuclear-fission fragments, when a gas laser tube is placed in close proximity to a nuclear reactor.
• Supersonic expansion of gases, usually preheated by chemical reaction or electrical discharge, through supersonic expansion nozzles, to create the so-called gasdynamic lasers.
• Plasma pumping in hot dense plasmas, created by plasma pinches, focused high-power laser pulses, or electrical pulses. There are also widely believed rumors that X-ray laser action has in fact been demonstrated in a rod of some laser material pumped by the ultimate high-energy pump source, the explosion of a nuclear bomb.

In general, any nonequilibrium situation that involves intense enough energy deposition is reasonably likely to produce laser action, given the right conditions. Schawlow's Law asserts in fact that anything will lase if you hit it hard enough.

Schawlow himself has attempted to illustrate this by building, and then consuming, the first edible laser— a fluorescein dye in Knox gelatine, "prepared in accordance with the directions on the package" and then pumped with a pulsed N₂ laser.

The fumes of Scotch whiskeys are also rumored to give molecular laser action in the far infrared when pumped with CO₂ radiation at 10.6 μm; and Israeli ingenuity has demonstrated a gasoline-fueled chemical laser which is ignited by an automobile spark plug (kilojoules per gallon and resulting pollution problems not identified).

Laser Performance Records

Much ingenuity as well as much sophisticated physics and engineering have thus far gone into laser research and development. As a result of this, plus the enormous flexibility of the stimulated-emission principle, in nearly every performance characteristic that we can define, the world record for any type of electronic device can be claimed by some laser device or laser system (generally a different laser for each characteristic). Different lasers can claim the current performance records in the following areas.

(a) Instantaneous peak power. A rather modest amplified mode-locked solid-state laser system can generate a peak instantaneous power in excess of ~ \(10^{13}\) W—or several times the total installed-electrical generating capacity of the United States—though only for a few picoseconds.

(b) Continuous average power. The unclassified power outputs from certain infrared chemical lasers are in the range of several hundred kilowatts to one-half megawatt of continuous power output. The classified figures for cw power output are, at a guess, probably several megawatts cw or greater.

(c) Absolute frequency stability. The short-term spectral purity of a highly stabilized cw laser oscillator can be at least as good as \(1:10^{13}\). The absolute reproducibility of, for example, a He-Ne 3.39 μm laser stabilized against a methane absorption line will exceed 1 part in \(10^{10}\), and may become much better. The absolute standard of time at present is already an atomic stimulated absorption device, the cesium atomic clock. This may be replaced in the future as an absolute standard for both frequency and time by a very stable laser, stabilized against an IR or visible absorption line.

(d) Short pulsewidth. Mode-locked laser pulses shorter than 1 ps (\(10^{-12}\) sec) in duration are now fairly routine. The current record is in fact a mode-locked and then compressed dye laser pulse with duration (full width at half maximum) of \(\tau_p\approx\) 12 femtoseconds, or \(1.2\times10^{-14}\) seconds. Since this corresponds to a burst of light only ~6 optical cycles in duration, further sizeable improvements may be difficult.

(e) Instantaneous bandwidth and tuning range. Most common lasers are limited to sharply defined discrete frequencies of operation that depend on the transitions of the specific atoms employed in the laser, and to fairly narrow tuning ranges that depend on the linewidths of these atomic transitions. Both organic dye lasers in the visible and semiconductor lasers in the near infrared can offer, however, instantaneous amplification bandwidths of order \(\Delta\lambda\approx200\) Å. This corresponds, for the former, to a frequency bandwidth \(\Delta{f}\approx24\times10^{12}\) Hz, or 24,000 GHz, or about one telephone channel for every person on Earth. 

(f) Antenna beamwidths. The diffraction-limited beamwidth of a visible laser beam coming from a telescope 10 cm in diameter is considered easy to obtain. In order to obtain such a beamwidth at even a high microwave frequency of 30 GHz (\(\lambda\) = 1 cm), we would have to use diffraction-limited microwave antenna two kilometers in diameter. 

(g) Noise figure. Laser amplifiers actually do riot offer particularly good noise-figure performance in the usual sense of this term, because of the unavoidable added noise that comes from spontaneous emission in the laser medium. (It is simply not possible to have an inverted laser population without also having spontaneous emission from the upper level.)

This comparatively poor noise performance is, however, really an inherent limitation of the optical-frequency range rather than of the laser principle. That is, it can be shown that no coherent or linear phase-preserving amplifier of any kind can be a highly sensitive receiver or detector at optical frequencies, because "quantum noise" imposes a rather poor noise limitation, equivalent to an input noise of one photon per inverse amplifier bandwidth, on any such optical amplifier, no matter how it operates.

Spontaneous emission is the putative source of this noise in a laser device, but any other conceivable optical amplifier with the same performance characteristics will have some equivalent noise source. (This noise limitation can be viewed as representing, if you like, the quantum uncertainty principle appearing in another guise.) Real lasers can, however, operate very close to this quantum noise limit.

Maser amplifiers can, in any case, provide noise figures in the microwave and radio-frequency ranges that are lower than those for any other electronic types of amplifiers at the same frequencies (though both cooled parametric amplifiers and even microwave traveling-wave tubes can come very close to the same values).

Natural Masers and Lasers

It is also very challenging to realize that naturally occurring molecular masers and lasers with truly enormous power outputs have been oscillating for eons in interstellar space, on comets, and in planetary atmospheres in our own solar system.

Naturally occurring maser action was first identified from observations that certain discrete molecular lines in the radio emission coming from interstellar clouds had enormously large intensities (equivalent to blackbody radiation temperatures of \(10^{12}\) to \(10^{15}\) K), but at the same time had very narrow doppler linewidths, corresponding to kinetic temperatures below 100 K. The radiation was also found to be sometimes strongly polarized, and to occur only on a very few discrete lines in the complex spectra of these molecules.

The only reasonable explanation is that these emissions represent naturally occurring microwave maser action on these particular molecular transitions. Such astronomical maser amplification has been seen on certain discrete vibrational and rotational transitions of molecules, such as the hydroxyl radical (\(\text{OH}^-\), 1,600 to 1,700 MHz), water vapor (\(\text{H}_2\text{O}\), ~ 22 GHz), silicon monoxide (SiO, mm wave region), and a few others. The pumping mechanism responsible for producing inversion is still uncertain, but may involve either radiative pumping by IR or UV radiation from nearby stellar sources or collision pumping by energetic particles. There is of course no feedback; so the observed radiation represents highly amplified spontaneous emission or "ASE" rather than true coherent oscillation.

More recently, amplified spontaneous-emission lines corresponding to population inversion on known CO₂ laser transitions near 10.4 and 9.4 μm have similarly been observed coming from the planetary atmospheres, or mesospheres, of the planets Mars and Venus. The pumping mechanism is believed to be absorption of sunlight by the CO₂ molecules. The net gains through the atmospheric layers are remarkably small (\(\le\) 10%) but the total powers involved quite large, because of  the large volumes involved in these "natural lasers."

The next tutorial discusses in detail about Keplerian Afocal Lenses.