Solid State Lasers

Solid-state laser consists of a host and an active ion doped in the solid host material. The Active ion must have sharp fluorescent line, broad absorption bands and high quantum efficiency for the wavelength of interest. The host material must be strong, and fracture resistant, with high thermal conductivity and high optical quality. Glasses and crystalline materials have shown to have these characteristics, when doped with rare earth ions. Silicate glasses, phosphate glasses, crystalline material like, garnets, aluminates, metal oxides, fluorides, molybdates, tungstates, etc, are very good hosts. Important active ions are rare earth ions like, neodymium, erbium, holmium and transition metals like, chromium, titanium, nickel, etc. Some of the important solid state lasers are, Ruby, Nd:YAG, Nd:Glass, Nd:Cr:GSGG, Er:Glass, Alexandrite, Titanium: sapphire, etc.

Basic parts of a flash pumped solid-state laser are given in the adjoining figure. All the solid-state laser materials used as the active medium have their absorption bands in the visible region. Consequently, optical pumping with flash lamps having their emission spectra in the visible region is used as the excitation mechanism. Flash lamp pumped Solid state lasers are, in general, very inefficient, as only a very small region of the emission spectra is used in the absorption process, absorption band of the active ion being very narrow and rest being unutilized. Pumping using Diode lasers with precisely matching output with the absorption band of the active medium have improved the efficiency of solid state lasers considerably, some times almost touching 100%. But since the output power of the diode laser being rather low, solid-state laser output is also low. To overcome this drawback, stacks of diodes are employed to increase their total output, thus generating very high power laser giving as good as the flash lamp pumped laser systems. The real advantage of a diode pumped solid-state laser that it is very compact, light weight and small in size, with long life.

Basic parts of Solid State Laser System
Basic parts of Solid State Laser System
There are a large number of solid-state lasers and we will discuss only some of the very important solid-state lasers and their salient features.

Pumping of Solid State Lasers

Pumping of the gain media is usually performed in one of the following forms:

  • Optical pumping
  • Electrical pumping
  • Chemical pumping

So far as solid-state lasers are concerned, it is mainly the optical pumping, which is being used. Optical pumping uses either cw or pulsed light emitted by a powerful lamp or a laser beam. Optical pumping can be realized by light from powerful incoherent sources. The incoherent light is absorbed by the active medium so that the atoms are pumped to the upper laser level. This method is especially suited for solid state or liquid lasers whose absorption bands are wide enough to absorb sufficient energy from the wide band incident incoherent light sources.

Optical pumping is a resonant process; the incident photon energy hn must be equal to the energy differences between the excited states and normal states. We can express optical pumping as hn + A → A*, where A is the atom at normal state, A* is the corresponding atom at excited state. So if there are lasers whose light wavelengths are within the absorption bands of the active medium, we can use these laser lights for pumping. Since the bandwidth of laser light is very narrow, the pumping efficiency can be very high. Laser pumping is not limited to solid-state lasers, it can also be used for liquid and gas lasers. In fact diode laser pumping has become the dominant means of optical pumping for reasons discussed below.

The first ever laser, the ruby laser reported by Maiman, was pumped with a discharge lamp viz. flash lamp. Though not very efficient, still there are few advantages; for example:

  • The price per watt of generated pump power is much lower for lamps, compared with laser diodes used for diode pumping.
  • Very high pump powers (particularly peak powers) can be generated.
  • Lamps are fairly robust, e.g. quite immune to voltage or current spikes.
  • However, device lifetime, power efficiency, cooling and thermal lensing are not really important issues e.g. when a flash lamp is operated with low pulse repetition rate and low average power, as required e.g. in engraving and marking systems.

Discharge lamps used for laser pumping can be grouped in two categories: arc lamps and flash lamps. Arc lamps are usually optimized for continuous operation, whereas flash lamps find their applications in pulsed lamps.

In most of the cases, laser rod and lamp are placed within an elliptical pump chamber with reflective walls, so that a larger percentage of the generated pump light can be absorbed in the laser rod (as shown in the figure). Cooled water or an ethylene glycol mixture is circulated to remove the excess heat. In addition to rod geometries, slab lasers can also be pumped through flash lamps. Here, an array of lamps pumps a slab through its large face, possibly from both sides. The pump light may be injected through a layer of cooling fluid.

Typical Flash Lamp - Laser Rod configuration

The main disadvantages associated with Flash lamp pumping includes:

  • The lifetime of lamps is very limited - normally up to a few thousand hours.
  • Flash lamps have a broad emission spectra (see adjoining figure) whereas the absorption spectra of lasing media have more or less discreet absorption peeks. . As a result, most of the optical energy being emitted by the flash lamp goes waste.
    Typical Flash Lamp Emission Spectra
  • The wall plug efficiency of the laser (electrical to optical efficiency) is low - typically ( few percent. This results in a higher heat load, making necessary a more powerful cooling system, and the strong thermal lensing and hence a poor beam quality.
  • Electric power supplies for lamp-pumped lasers involve high electrical voltages, which raise additional safety issues.
  • The low pump brightness (compared with that achievable with diode lasers) and the broad emission wavelength range exclude many solid-state gain media.

The second technique under optical pumping is through diode lasers .The lasers based on this type of pumping are known as Diode Pumped Solid State Lasers (DPSSL) or sometimes the all-solid state lasers.

Because optical pumping is a resonant process, the wavelengths of the pumping diode lasers must be within the absorption bandwidth of the active medium to be pumped, the nearer to the absorption peak wavelength the better. The adjoining figure show the absorption spectral of Nd:YAG laser which has a peak absorption value at 810 nm. GaAs / AlGaAs quantum well (QW) diode lasers operating at about 800 nm can be used to pump this laser. Likewise, Nd:Glass has a absorption peak at 802 nm and thus can also be pumped by the same laser.

Absorption spectra of Nd:YAG crystal
Absorption spectra of Nd:YAG crystal

However, for Yb:YAG laser and Yb:glass laser, the best absorption wavelengths are 960 and 980 nm respectively, we can pump them using InGaSa/GaAs strained quantum well (QW) lasers in the 950-980 nm range.

In case of diode laser pumping, the absorption efficiency is about 0.90~0.98, whereas for flash lamp pumping, it is about 0.17. Further the energy quantum efficiency for diode laser pumping is about 1.4 times as large as flash lamp pumping, with typical value of 0.82 and 0.59 respectively. So the overall pumping efficiency of diode laser is about 7~8 times that of lamp pumping.

The advantage is quite clear. For normal pumping processes, because of the low efficiency of pumping and the required high pumping power to maintain proper power output, a large fraction of the pumping power is wasted as harmful heat. This heat has to be properly removed, i.e., the laser has to be properly cooled to maintain proper working conditions. While for diode pumped lasers, much of the absorbed power is used for final population inversion, the ratio of thermal generation from the absorbed radiation power for diode laser pumping is much less than that for lamp pumping. Thus the power required for diode pumping is far less than the lamp pumping; the absolute value of thermal burden of diode laser pumping is also strikingly small compared with lamp pumping. This makes it possible for more compact laser designs.

We can divide diode laser pumping into four types according to the degree of integration of the diode lasers: single stripe, diode array, diode bar and diode stack. Normally the pumping power increases with the integration degree.

  • Low-power lasers (up to roughly 200 mW) can be pumped with small edge-emitting laser diodes. These exhibit a diffraction limited beam quality and make it quite easy to achieve the same for the solid-state laser.
  • Broad area diodes typically generate several watts and are suitable for pumping solid-state lasers with output powers up to a few watts. Their beam quality is quite asymmetric, but normally still sufficient for achieving a diffraction-limited laser output without using complicated optics.
  • High power diode bars emit tens of watts (or even >100 W), allowing for higher output powers, particularly when several bars are combined. Their beam quality is strongly asymmetric and quite poor; as a result their radiance is much lower than that of lower-power diodes. However, beam shapers are often used to improve the beam quality and to make the beam symmetric.
  • For the highest powers, diode stacks are often used. These have a still worse beam quality and lower brightness, but can provide multiple kilowatts. We can stack the bars into a two dimensional structure, it is reported that 1 cm long bars are stacked to form an emitting area. The average power is about 100W/cm2, peak power 1kW/cm2.

There are basically two types of pump geometry, longitudinal pumping (pump beam enters the laser medium along the resonator axis) and transverse pumping (pump beam incident on the active medium from transverse directions to the resonator axis). For longitudinal pumping, the beam needs to be concentrated to a small and circular spot. These two types of pumping viz. edge pumping and side pumping are shown in the following figures.

Edge pumping and side pumping
Edge pumping and side pumping
The main advantages of diode pumping can be summarized as follows
  • The compactness of the pump source, the power supply and the cooling arrangement makes the whole laser system much smaller and easier to use.
  • A high electrical-to-optical efficiency of the pump source (order of 50%) leads to a high overall power efficiency i.e. wall plug efficiency of the laser. As a consequence, small power supplies are needed, and both the electricity consumption and the cooling demands are drastically reduced, comparing with those for lamp-pumped lasers.
  • The narrow optical bandwidth of diode lasers makes it possible to directly pump certain transitions of laser-active ions without losing power in other spectral regions. It thus also contributes to a high efficiency.
  • Although the beam quality of high power diode lasers is poor, however, end pumping of lasers provide very good overlap of laser mode and pump region, leading to high beam quality and power efficiency.
  • Diode-pumped low-power lasers can be pumped with diffraction-limited laser diodes. This allows the construction of very low power lasers with reasonable power efficiency.
  • The lifetime of laser diodes is long compared with that of discharge lamps: Further it is much easier to replace laser diodes as compared to discharge lamps.
  • Diode pumping makes it possible to use a very wide range of solid-state gain media for different wavelength regions.

The main disadvantage of diode pumping (as compared to lamp pumping) is the significantly higher cost per watt of pump power.

Ruby laser

The World's first solid-state laser, invented by Maiman in 1960, now has only a historical importance. The laser host is Aluminium oxide (Al2O3) with triply ionized chromium (Cr3+) as the active ion.

This first material used was synthetic ruby. Ruby is crystalline alumina (Al2O3) in which a small fraction of the Al3+ ions have been replaced by chromium ions, Cr3+. It is the chromium ions that give rise to the characteristic pink or red color of ruby and it is in these ions that a population inversion is set up in a ruby laser.

The two broad absorption regions centered on 400 nm and 550 nm are both used for optical pumping of the ruby. Thus most of the useful pump light for a ruby rod lies in the blue-green portion of the visible spectrum. In a ruby laser, a rod of ruby is irradiated with the intense flash of light from xenon-filled flashtubes. Light in the green and blue regions of the spectrum is absorbed by chromium ions, raising the energy of electrons of the ions from the ground state level to the broad F bands of levels. Electrons in the F bands rapidly undergo non-radiative transitions to the two metastable E levels. A non-radiative transition does not result in the emission of light; the energy released in the transition is dissipated as heat in the ruby crystal. The metastable levels are unusual in that they have a relatively long lifetime of about 4 milliseconds (4 x 10-3 s), the major decay process being a transition from the lower level to the ground state. This long lifetime allows a high proportion μmore than a half) of the chromium ions to build up in the metastable levels so that a population inversion is set up between these levels and the ground state level. This population inversion is the condition required for stimulated emission to overcome absorption and so give rise to the amplification of light. In an assembly of chromium ions in which a population inversion has been set up, some will decay spontaneously to the ground state level emitting red light of wavelength 694.3 nm in the process. This light can then interact with other chromium ions that are in the metastable levels causing them to emit light of the same wavelength by stimulated emission. As each stimulating photon leads to the emission of two photons, the intensity of the light emitted will build up quickly through this cascading process.

Energy level diagram of chromium ions in Ruby
Energy level diagram of chromium ions in Ruby

The ruby laser is often referred to as an example of a three-level system. More than three energy levels are actually involved but they can be put into three categories. These are; the lower level form which pumping takes place, the F levels into which the chromium ions are pumped, and the metastable levels from which stimulated emission occurs. It is a three level laser and as such threshold for laser action is nearly 300 to 400 times when compared with Nd:YAG laser (four level laser) of similar dimensions. Working of this laser has already been discussed earlier. Some important properties of Ruby are listed below:

Important Properties of Ruby

Property Value
Density 3.98 g/cc
Melting Point 2040°C
Young's Modulus 345 Gpa
Compressive Strength 2.0 Gpa
Hardness 9 Mhos, 2000 Knoop
Thermal Expansion 5.8 x 10-6 / °C {20° to 50°C} ; 7.7 x 10-6 / °C {20° to 200°C }
Thermal Conductivity 46.02 W / μm∙K) { 0°C } ; 25.10 W / μm∙K) {100°C} ; 12.55 W / μm∙K) {400°C}
Refractive index at 700 nm 1.7638 Ordinary Ray ; 1.7556 Extraordinary Ray
Birefringence 0.008
Refractive Index vs. Chromium Concentration 3 x 10-3 (Δn / % Cr2O3)
Crystallographic orientation, optical (c - axis) to rod axis 60° within 5°
Fluorescent Lifetime at 0.05% Cr2O3 3 ms at 300 K
Fluorescent Linewidth 5.0 Å at 300K
Output Wavelength 6.94.3 nm
Major Pump Bands 404 nm and 554 nm

Concept of Maiman's Ruby laser is shown below.

Concept and components of first Ruby laser Maiman
Concept and components of first Ruby laser Maiman

Other types of laser operate on a four level system and, in general, the mechanism of amplification differs for different lasing materials. However, in all cases, it is necessary to set up a population inversion so that stimulated emission occurs more often than absorption.

Neodymium Class of Lasers

Neodymium, with chemical symbol as Nd, is a chemical element belonging to the group of rare earth metals. In laser technology, it is widely used in the form of the trivalent ion Nd3+ as the laser-active dopant of gain media based on various host materials, including both crystals and laser glasses.

The strongest laser transition is that from 4F3/2 to 4I11/2 for 1064 nm, but other transitions are available with longer or shorter wavelengths. In order to achieve lasing on those, lasing at the 1064-nm line needs to be suppressed by inserting an appropriate wavelength filter. Neodymium atoms in the ground state absorb photons and are raised in energy to one of the pump bands. The states in these bands have lifetimes on the order of 10-8 seconds, and the atoms quickly drop to the upper lasing level by radiation less transition. The upper lasing level, 4F3/2, has a fluorescent lifetime of about 0.3 msec. A population inversion develops and lasing occurs, with the atoms dropping to the lower lasing level. his level is very close to the ground state, and excited atoms rapidly return to the ground state by another radiation less transition.

The population in level 4I11/2 quickly reduces to zero as excited species jump to the ground state 4I9/2 via multi-phonon emission.

Since the lifetime of the lower states is much smaller than that of the upper states, there is normally negligible population in all these levels, so that neodymium-doped gain media exhibit pure four - level behaviour.

Prominent lines of Nd in YAG
Prominent lines of Nd in YAG

The greatest consideration in the design of a solid-state laser is spectral matching of the pump source to the absorption spectrum of the laser rod. Xenon flashlamps provide the most efficient operation of ruby lasers. Krypton arc lamps and flashlamps are best with neodymium lasers. The krypton flashlamp produces most of its output light in the infrared region of the absorption bands of Nd:YAG and Nd:glass. Thus, it is the best spectral match for these laser materials. Krypton flashlamps are however, not widely used because of their cost. They are far more expensive than xenon lamps, and the xenon lamps also have sufficient output in the desired spectral region, thus making their lower efficiency acceptable.

The most common neodymium-doped gain media are:

  • Nd:YAG = Nd:Y3Al5O12 (yttrium aluminum garnet): the classical choice for 1064 nm, but also usable at 946 nm and 1320 nm (and a few other lines); isotropic; still very common particularly for high power lasers and Q-switched lasers.

The most studied of all the solid state lasers, Nd:YAG was lased in 1964. In Nd:YAG laser, YAG is the host and triply ionized neodymium (Nd3+) is the active ion responsible for the laser output at 1064 nm wavelength. It is a four level laser with high fluorescence efficiency. An Nd:YAG rod of 75mm length and 6mm diameter lases at a very low threshold of less than a Joule with a matching pulse forming network and a xenon-krypton gas mixture flash lamp. It has a high thermal conductivity and can be cooled with fluid coolant efficiently and produce high output at repetition rate of 400 pulses per second (pps) or better. But its efficiency is around 1% due to its very narrow absorption bands; consequently most of the visible output of the flash lamp is unutilized. Typical neodymium doping concentrations are of the order of 1% (atm.). High doping concentrations can be advantageous e.g. because they reduce the pump absorption length, but too high concentrations lead to quenching of the upper state lifetime via up conversion processes. The YAG absorption lines form sharp spikes within closely packed bands. The two important pumping bands in Nd:YAG lasers are in the regions of 730-760 nm and 790-820 nm. Since both of these bands are in the near infrared, these wavelengths are the most desirable for optical pumping of YAG lasers.

Water cooling of the rod combined with the high thermal conductivity of YAG provides a cooling effect sufficient that small-diameter Nd:YAG laser rods may be operated in the CW mode. YAG is the only widely used solid-state laser material capable of CW operation, although other CW solid-state lasers are under development.

  • Nd:Cr:GSGG laser: Nd3+ is the active ion in this case also, the host being gadolinium scandium gallium garnet (GSGG), sensitized with Cr3+. GSGG is a material with higher fracture limit and sensitization with triply ionized chromium (Cr3+) gives it a far better efficiency, compared to Nd:YAG, because Cr3+ absorbs the unutilized part of the emission spectra of the flash lamp and emits in the band corresponding to the absorption band of Nd3+ ion. Its lasing wavelength is 1064 nm.
  • Nd: YVO4 (yttrium vanadate,) for 1064 nm, 914 nm and 1342 nm: very high pump and laser cross sections and larger gain bandwidth, compared with Nd:YAG, thus particularly attractive for low - threshold lasers; also good properties for high power operation with good beam quality (low dn/dT); birefringent
  • Nd:YLF = Nd:YLiF4 (yttrium lithium fluoride) for 1047 nm and 1053 nm: birefringent, long upper state life time, weak thermal lensing: useful for high power Q switched lasers.
  • Nd:GdVO4 (gadolinium vanadate) for 1064 nm and 1341 nm: similar to Nd:YVO4, but having a larger gain bandwidth.
  • Nd:GGG (gadolinium gallium garnet): often used for high power heat capacity lasers
  • Nd:YAP ( yttrium aluminum phosphate) : high thermal conductivity, birefringent
  • Nd:glass : Neodymium atoms are also used as the active elements in Nd:glass lasers. The doping level is usually 1% or less. The absorption spectrum and energy-level diagrams of Nd:glass are similar to those of Nd:YAG, but the glass absorption peaks are much broader and less distinct. The reason for this is that glass is not a crystalline structure as is YAG. Glass is a supercooled fluid and has a random amorphous structure. Neodymium ions in a YAG crystal all have the same spacing from neighboring atoms and very similar environments. In glass the atomic distances and distribution are random, and each ion has a different environment. This causes the energy levels of different ions to shift differently, broadening all the absorption and emission lines considerably. This also results in a somewhat longer lifetime for the upper lasing level. This means that Nd:glass has a higher efficiency than Nd:YAG in the pulsed mode and a broader output linewidth.

As glass has a much lower thermal conductivity than YAG implying that the waste heat is retained in the lasing material longer, resulting in a greater temperature rise. For this reason the transition from the lower lasing level to the ground state in Nd:glass occurs much more slowly as the laser temperature rises during operation. This quickly quenches lasing and requires that Nd:glass lasers operate in the pulsed mode only.

Though this laser also is a four level laser with Nd3+ as the active laser ion, it cannot generate output at high repetition rate due to its very low thermal conductivity, but it can produce much higher energy output as compared to Nd:YAG laser. As both silicate glasses and phosphate glasses are used as hosts; depending on the hosts lasing is at 1061 nm or 1054 nm respectively. Nd:glass lasers typically can be pulsed only once every few seconds, but the larger rods can deliver pulse energies of several hundred joules for relatively small systems and kilojoules for larger ones. They are the most efficient solid-state lasers, and the least expensive. This makes glass popular where high-energy pulses are required. Finally, these neodymium-doped glasses (mostly silicate and phosphate glasses) can be used for laser applications. However, silicate glasses are often more attractive for neodymium-doped optical fibers, which are suitable for fiber lasers and amplifiers.

In all these media (except for glasses), the neodymium dopant ions replace other ions (often yttrium) of the host medium, which have about the same size.

Some Important properties of Nd:YAG crystals are given below

Property Value
Chemical formula Nd3+:Y3Al5O12
Crystal structure Cubic
Density 4.56 g/cm3
Moh hardness 8 to 8.5
Young's modulus 280 GPa
Tensile strength 200 MPa
Melting point 1970 °C
Thermal conductivity 10 to 14 W / μm K)
Thermal expansion coefficient 7 to 8·10-6/K
Birefringence None (only thermally induced)
Refractive index at 1064 nm 1.82
Temperature dependence of refractive index 7 - 10 x 10-6/K
Nd density for 1% atm. doping 1.36∙1020 cm-3
Fluorescence lifetime 230 μs
Absorption cross section at 808 nm 7.7 ∙10-20 cm2
Emission cross section at 1064 nm 28∙10-20 cm2
Gain bandwidth 0.6 nm

Ytterbium doped class of solid-state lasers

Ytterbium is a chemical element belonging to the group of rare earth metals having chemical symbol as Yb. Presently, it has acquired a prominent role in the form of the trivalent ion Yb3+, which is used as a laser-active dopant in a variety of host materials, including both crystals and glasses. It is often being used for high power lasers and for wavelength- tunable solid-state lasers. Energy levels of Yb3+ ions in Yb:YAG, and the usual pump and laser transitions are shown in the adjoining figure.

Energy levels of Yb<SUP>3+</SUP> ions in Yb:YAG
Energy levels of Yb3+ ions in Yb:YAG

Ytterbium-doped laser crystals and glasses have a number of interesting properties, which differ from those e.g. of Nd: doped host materials.

  • They have a very simple electronic level structure, with only one excited state manifold (2F5/2) within reach from the ground state manifold (2F7/2) with near-infrared photons.
  • Pumping and amplification involve transitions between different sublevels of the ground state and excited state manifolds.
  • The quantum defect is always rather small, making them suitable for high power lasers.
  • The gain bandwidth of the laser transitions is typically quite large, compared to Nd : doped crystals thus making them suitable for applications involving wide wavelength tuning, generation of ultra short pulses in mode - locked lasers.
  • The upper state life times are relatively long; typically of the order of 1-2 milliseconds, which is beneficial for Q - switching.

In addition to neodymium and ytterbium, there are other dopants have also been attempted in YAG crystals and laser glasses. For example, Erbium doped laser materials can emit at various wavelengths e.g. Er: Glass laser emits radiation at 1540 nm 1.54 μm, suitable for eye-safe laser applications; Er:YAG emitting at 2.94 μm and is used in dentistry and for skin resurfacing: Er:YAG can also emit at 1645 nm and 1617 nm, as well as at 550 nm and 561 nm; Er:YLF emits at 1730nm and Er:GSGG emits at 2.8 μm.

Active elements from Erbium doped Yttrium Scandium Gallium Garnet crystals (Er:Y3Sc2Ga3012 or Er:YSGG) single crystals are designed for diode pumped solid-state lasers radiating in the 3 μm range. Er:YSGG crystals show the potential of their application alongside with the widely used Er:YAG, Er:GGG and Er:YLF crystals. Flash lamp pumped solid-state lasers based on Cr,Nd and Cr,Er doped Yttrium Scandium Gallium Garnet crystals (Cr,Nd:Y3Sc2Ga3012 or Cr,Nd:YSGG and Cr,Er:Y3Sc2Ga3012 or Cr,Er:YSGG) have a higher efficiency than those based on Nd:YAG and Er:YAG. Active elements prepared from YSGG crystals are optimum for medium power pulse lasers with the repetition rates up to several tens of cycles.

Comparative generation characteristics

Crystal type Er:YSGG Er:YAG
Er concerntation, at. % 38 33
Pumping wavelength, nm 966 964
Stimulated radiation wavelength, μm 2.797; 2.823 2.830
Generation threshold, mW 72 418
Max. Power output at pumping power 720 mW, 966 nm 201 51
Slope efficiency, % 31.1 16.9

The advantages of YSGG crystals compared with YAG crystals, however, are lost when large size elements are used because of the inferior thermal characteristics of YSGG crystals.

Eye Safe Lasers

For a given power level, Lasers emitting in a wavelength region with relatively low hazards for the human eye are known as eye-safe lasers. Lasers with emission wavelengths longer than 1.4 μm usually fall in this category of "eye-safe", because light in that wavelength range is strongly absorbed in the eye's lens and thus cannot reach the significantly more sensitive retina. Wavelengths between 400 nm and 1400 nm are focused by the curved cornea and lens on to the retina; the optical gain is about 100,000-200,000 times. Viewing a laser beam or Point Source will focus all the light on a very small area of the retina, resulting in a greatly increased power density and an increased chance of damage.

Transmission spectra of human eye
Transmission spectra of human eye

Obviously, the quality "eye-safe" depends not only on the emission wavelength, but also on the power level and the optical intensity, which can reach the eye. With sufficient power, as e.g. reached with a fiber amplifier or with a Q- switched laser, the eye can of course still be damaged, e.g. by overheating the eye's lens. Threshold energy density for retinal damage at 1064 nm (Nd:YAG) is 10-6 J/cm2 and at 694.3 nm (Ruby) it is 10-7 J/cm2. For laser wavelengths above 1400 nm, damage to the retina occurs at very high energy density, since the transmission of the eye is negligible as shown in the adjoining figure. For example, Er:Glass laser emits radiation at 1540 nm and threshold density of retinal damage is 1J/cm2. Such high energy density is not normally encountered at work place and these types of lasers are eye safe.

Solid-state lasers have been designed to operate at various wavelengths, but the band of wavelengths from about 1.4 mm to 1.6 mm is of great interest because of eye safety reasons. Maximum permissible exposure levels for eye in this band are several orders of magnitudes greater than invisible and one micron band [ANSI, Z136.1-1993]. Er: Glass class of lasers is gaining much attention due to its radiation wavelength of 1540 nm which is not only safe for eye, but can also be used for rage finder applications as there is an atmospheric window for this wavelength.

On the other hand, eye-safe lasers in the range of 2 - 3 μm wavelength are being used in the fields of coherent Doppler velocimetry, gas detection, space applications and medical operations since water exhibits a strong absorption spectrum in this wavelength region. Er: YAG laser, which emits at 2.94 μm, also falls in this category. Unlike in Nd:YAG lasers, the frequency of Er:YAG lasers is strongly absorbed by water due to atomic resonances. This restricts its use in range finder applications and many other laser applications e.g. surgery, which have water present. Because of this limitation Er:YAG lasers are far less common than relatives such as Nd:YAG and Er: glass.

In addition to the Er: glass, which is the main workhorse in the area of eye-safe lasers, some of the other flash pumped eye safe solid state lasers are Thulium doped YAG, Tm3+:YAG ( around 2 μm ), Holmium doped YAG, Ho:YAG ( 2.1 μm ), Chromium doped YAG , Cr4+:YAG ( 1.35 - 1.55 μm ) , Er:YLF (1730nm), Er:YAG (2940nm), Er:Cr:GSGG ( 2.8μm) and Ho:YLF (2060nm).

Since wavelength around 1.5 micron is of interest, apart from direct generation of this eye safe laser radiation, it has also been generated by shifting the 1064nm output of Nd:YAG to 1540nm by Stimulated Raman Scattering (SRS) technique. Other commonly used technique is based on optical parametric oscillators (OPOs), which involves KTP and periodically poled KTP (PPKTP) crystals.

The energy level diagram of Erbium in glass matrix is shown in the adjoining figures along with the typical absorption and emission cross-sections for the most prominent 4I13/2 → 4I15/2 transition in Er+3spectra.

Energy level diagram of Erbium in glass matrix
Energy level diagram of Erbium in glass matrix
Absorption and emission cross-sections
Absorption and emission cross-sections for 4I13/2 → 4I15/2 transition in Er+3spectra

Table lists the important properties of these erbium-doped glasses. Recently Ytterbium and chromium have also been co-doped with erbium for 1.54-micron applications. Erbium laser glass with Yb ion doping has been found suitable for microlaser system using moderate power diode pumping laser system (DPSS). Using these glasses, the cw 1540 nm has also been obtained with good beam quality and stable output. Chromium co-doping with erbium, on the other hand, in laser glasses are especially suitable for the high power xenon lamp-pumping laser.

Properties of Er: doped Laser Glasses

Property Value
Center lasing wavelength 1.535 μm
Stimulated emission cross section 8.0 x 10-21 cm2
Fluorescence life time 7.9 msec
Refractive index 1.533 at 0.6 μm: 1.521 at 1.535 μm
Temp coeff. Of refractive index, dn/dt -10 x 10 -7 /°C between 20 - 40°C
Transformation temp 450°C
Softening temp 485°C
Thermal coeff of expansion 8.2 x 10-6 /°C between 20 - 40°C
Density 2.90 g / cc

Tunable Lasers

Normally, stimulated emission in solid-state laser is in the form of photons. But it is possible to couple stimulated emission of photons with the phonons, the vibrational quanta of lattice, where the fixed total energy of laser transition can be partitioned between photons and phonons in a continuous way. This has resulted in a new class of solid-state laser called vibronic laser, where a phonon is emitted or absorbed with each electronic transition. Historically, the first tunable vibronic laser, Nickel doped in Magnesium Fluoride (Ni:MgF2) was lased in 1963 at Bell Labs. This followed by a series of vibronic lasers, using nickel, vanadium, cobalt etc. as the dopant material and MnF2, MgO, MgF2, ZnF2 etc. as the host crystal. All these flash pumped lasers worked at cryogenic temperature. Optically pumped Ho:BaY2F6 was the first tunable vibronic laser to operate at room temperature. It can be seen from literature that, chromium (Cr3+) plays a very important role, as dopant, in many of the tunable solid-state lasers. In tunable lasers, output is tunable from visible to infra-red. Some of the important tunable lasers are Alexandrite (BeAl2O4), Emerald (Be3Al2Si6O18) and Titanium:sapphire (Ti:Al2O3) lasers.

Alexandrite Laser

Alexandrite laser was invented in 1974. The laser material is Cr3+ doped chrysoberyl (Cr+3:BeAl2O4). It is tunable from 700 to 820nm, is mechanically strong, chemically stable, has high average power capability, high thermal coefficient, performs better at higher temperature, can be Q-switched and can also be made to lase in the CW mode. As a 3-level system its function is very much akin to that of the ruby laser and lases at a fixed wavelength of 680nm, has high threshold for laser action with low efficiency. As a 4-level laser its function is that of a vibronic laser: that is, phonons, as well as photons, are emitted during lasing. The wavelength tuning is accomplished by controlling the branching of energy between phonons and photons during lasing. Alexandrite lasers have been tuned across most of the spectrum between 701 and 860 nm. The central part of the tuning range is from 700 - 820 nm. Using non-linear wavelength conversion processes such as harmonic generation and Raman shifting, light has been generated at wavelengths from the deep IR (20 μm) to the VUV. In addition to its broad absorption bands throughout the visible spectrum, alexandrite exhibits narrow R line absorption features at wavelengths near 680 nm. These properties together with its long fluorescence lifetime make it an excellent material for both flashlamp and diode pumping. Alexandrite's thermo-mechanical properties make it an excellent performer in high power laser applications.

Material Properties of Alexandrite (Cr+3: BeAl2O4)

Property Value
Operating wavelength 700 - 820 nm
Crystal structure Rhombic
Lattice parameters a = 5.47 Å : b = 9.39 Å : c = 4.42 Å
Hardness 8.5 mohs
Density 3.79 g/cm3
Refractive index 1.74 - 1.75
Axial characteristic Biaxial
Thermal conductivity 0.23 W/cm °K
Stimulated emission cross-section at 300°K 3.0 x 10-19 cm2
Lifetime 260 x 10-6 sec
Absorption loss at 750 nm 0.001 - 0.003 cm-1
Cr dopant concentration 0.03 - 0.50 at. %

Simplified energy level diagram of alexandrite, as a 4-level laser, shown here.

Simplified energy level diagram of Alexandrite Laser
Simplified energy level diagram of Alexandrite Laser

It may be noticed that the upper laser level μmeta-stable level in the figure) is above the energy storage level and consequently the upper lasing level gets more populated from the transitions from the storage level with the rise in temperature of the system. The resulting transitions to ground level are vibronic in nature. i.e. photon emission is accompanied by lattice phonon creation giving rise to 4-level operation.

Emerald laser

Room temperature operation of alexandrite laser induced the search for other materials with similar properties, which resulted in the development of emerald laser in 1980. Chromium doped in beryllium aluminium silicate (Cr3+ in Be3Al2Si6O18) is the common name for emerald. It is a vibronic 4-level laser. The gain and emission cross-section of emerald is almost twice that of alexandrite. It has lower lasing threshold compared to alexandrite and ruby has many similarities with alexandrite like working better at higher temperature and excitation by flash lamps. Further, it is tunable from 730nm to 840nm and can be Q-switched and mode-locked. With its wide spectral bandwidth, it is capable of generating ultra-short pulses. Emerald, like alexandrite operates in a vibronic four level, phonon terminated mode and exhibits gain over a 695-835 nm wavelength range. Its broad fluorescence bandwidth, together with a high gain cross section and 65 μs room temperature fluorescence lifetime, make emerald an excellent laser material for high power, Q-switched, or mode-locked operation. Highly efficient quasi-cw (continuous-wave) laser operation has been achieved in emerald over the 720-842 nm tuning range.

Titanium:sapphire laser (Ti3+:Al2O3)

Titanium:sapphire laser is an important member of the family of vibronic lasers. In this case trivalent titanium is doped in the sapphire host material. Presently, it is the most widely used crystal for wavelengths tunable lasers. It combines the excellent thermal, physical and optical properties of Sapphire with the broadest tunable range of any known material. It can be lased over the entire band from 660 to 1100 nm. Frequency doubling provides tunability over the blue-green region of the visible spectrum. Ti:Sapphire crystals are active media for highly efficient tunable solid-state lasers. They demonstrate good operation in the pulsed-periodic, quasi-CW and CW modes of operation. Ti:Sapphire is a 4-level, Vibronic laser with fluorescence lifetime of 3.2 - 3.6 μm. The peak of the absorption band is 490 - 500 nm which makes it an excellent material for pumping with a variety of sources operating in the green-argon ion, copper vapour, frequency-doubled Nd:YAG or Nd: YLF, and dye lasers are routinely used. Excitation by flash lamp is very difficult due to its short fluorescence lifetime at room temperature. Nevertheless, flash lamp pumping was carried out in 1984 employing a coaxial flash lamp, operating with a pulse width of 5 μs. These flash lamps were specially designed to allow short fluorescence lifetime. These factors and broad tunability make it an excellent replacement for several common dye lasing materials. Peter Moulton was the first scientist, who demonstrated this laser in 1982.

Titanium-doped sapphire (Ti3+:sapphire) is also a widely used transition metal doped gain medium for femtosecond solid state lasers. Immediately after its demonstration, Ti:sapphire lasers quickly replaced most of the dye lasers, which had previously dominated the fields for ultrashort pulse generation and widely wavelength tunable lasers. These ultra short pulses from Ti:sapphire lasers can be generated using passive mode locking, where a pulse duration around 100 fs is easily achieved. However, using advanced precision dispersion compensation techniques, pulses of the order of 5 - 10 fs have also been obtained. Ti:sapphire lasers are also very convenient for pumping test setups of new solid state lasers such as based on neodymium or ytterbium doped gain media, since they can easily be tuned to the required pump wavelength and allow to work with very high pump brightness due to their excellent beam quality and high output power of typically several watts.

Properties of Er: doped Laser Glasses

Property Value
Crystal structure Hexagonal
Lattice parameters a = 4.748 Å; c = 12.957 Å
Axial characteristic Uniaxial
Tuning range 660 - 1100 nm
Pumping range 450 - 532 nm
Ti dopant concentration 0.02 - 0.35 at. %
Refractive index 1.76
Birefringence 0.0082
Density 3.98 g/cm3
Hardness 9 Mohs
Thermal conductivity at 25°C 0.33 - 0.35 W / cm °K
Specific heat at 18°C 761 J / kg °K
Thermal expansion coefficient (20 - 100°C) (4.78 - 5.31) x 10-6 / °K
Absorption coefficient at 510 nm 0.5 - 2.5 cm-1

The adjoining figure shows the energy diagram of the absorption and emission bands of the 3d1 Ti3+ ion. In the diagram, the 2T2 level is the ground state, while the 2E level is the excited state. The closely spaced vibrational sublevels broaden the electronic energy levels. The Ti:sapphire laser is called a vibronic laser because of the close blending of the electronic and vibrational frequencies.

Energy diagram of Ti:Sapphire Laser
Energy diagram of Ti:Sapphire Laser

The absorption band of Ti3+ is in the blue green spectral region, whereas the emission spectrum is slightly red shifted as shown in the figure given here:

Absorption and emission spectra of Ti:Sapphire Laser
Absorption and emission spectra of Ti:Sapphire Laser

Special properties of the Ti:sapphire gain medium can be summarized as follows



  • Sapphire μmonocrystalline Al2O3) has an excellent thermal conductivity, alleviating thermal effects even for high laser powers and intensities.
  • The Ti3+ ion has a very large gain bandwidth μmuch larger than that of rare earth doped gain media), allowing the generation of very short pulses as well as wide wavelength tenability.
  • The maximum gain and laser efficiency is obtained around 800 nm. The possible tuning range is ~650 nm to 1100 nm, although different mirror sets are normally required for covering this huge range, and exchanging mirror sets is a somewhat tedious task. However, the number of required using ultra broadband mirrors could reduce mirror sets.
  • There is also a wide range of possible pump wavelengths, which however are located in the green spectral region, where powerful laser diodes are not available. In most cases, several watts of pump power are used, sometimes even up to 20 W. Originally, Ti:sapphire lasers were in most cases pumped with 514-nm argon ion lasers, which are powerful, but very inefficient, expensive to operate, and bulky. Other kinds of green lasers, which are now being widely used, are frequency doubled solid-state lasers based on neodymium doped gain media such as Nd: YAG, Nd: YLF.
  • The upper state lifetime is rather short (3.2 - 3.6 μs), and the saturation power is very high. This means that the pump intensity needs to be rather high, so that a strongly focused pump beam and thus a pump source with high beam quality is required.
  • Despite the huge emission bandwidth, Ti:sapphire has relatively high laser cross sections, which reduces the tendency of Ti:sapphire lasers for Q - switching instabilities.

If the requirements in terms of pulse duration and output power are less stringent, Ti: sapphire lasers may be replaced with Cr:LiSAF (LiSrAlF6 ) or Cr:LiCAF (LiCaAlF6 ) lasers, which can be pumped at longer (red) wavelengths, where laser diodes are available. These Cr: doped materials are promising new solid-state laser material with a reasonably good tuning range. In the case of LiCAF, the peak lasing wavelength is at 780 nm with a tuning range from 720 to 840 nm. Whereas LiSAF has an even wider tuning range, covering 780-1010 nm with peak lasing wavelength is at 825 nm.

Nonlinear frequency conversion can be used to further extend the range of emission wavelengths of a Ti: sapphire laser system. The simplest possibility is frequency doubling to access the blue, ultraviolet and green spectral region. Another approach is to pump an optical parametric oscillator (OPO), offering a wide tuning range in the near or mid infrared spectral region.

OPO Based frequency conversion setup
OPO Based frequency conversion setup

The output wavelengths of the OPOs are usually tuned by changing the pump Ti: sapphire wavelength. This technique of tuning the OPO wavelength is mechanically simpler than the more common technique of angle-tuning the OPO crystal (which requires a physical rotation of the OPO crystal). In addition, it avoids the redirection of the output beam due to crystal rotation. Another advantage of pump-wavelength tuning is that it is possible to achieve rapid tuning with no moving parts by using an electronically tunable Ti: sapphire laser. The Ti: sapphire laser can be tuned using a conventional multi-plate birefringent filter. A typical OPO based frequency conversion set up for obtaining wavelengths in the range of 1.5 and 2.5 μ m for LIDAR applications is given in the adjoining figure.

A diode-pumped Nd:YLF or Nd: YAG laser is frequency doubled using to pump Ti: sapphire laser. Tuning of the Ti: sapphire laser can be accomplished by the computer-controlled, stepper-motor rotation of a birefringent filter or electro-optical or acousto-optical elements. This tunable radiation is subsequently used to pump one of two optical parametric oscillators to produce tunable mid-IR radiation. Frequency doubling can be accomplished using non-linear crystals like KTP (Potassium titanyl phosphate: KTiOPO4) or LBO (Lithium Triborate: LiB3O5), whereas OPO materials like RTA (Rubidium titanyl arsenate: RbTiOAsO4 ), CTA (cesium titanyl arsenate : CsTiOAsO4 ), RTP ( Rubidium titanyl phosphate :RbTiOP04), can be finally used for obtaining the wavelengths in the range of 2-5 micron.

It is worth mentioning that materials like Potassium Titanyl Arsenate (KTiOAsO4 or KTA) is an excellent optical non-linear crystal developed recently for non-linear optical and electro-optical device applications. The non-linear optical and electro optical coefficients are higher in these materials as compared to KTP and they have the added benefit of significantly reduced absorption in the 2.0 - 5.0 μm region. The large non-linear coefficients are combined with broad angular and temperature bandwidths. Additional advantages of the Arsenates are low dielectric constants, low loss tangent and ionic conductivities orders of magnitude less than KTP. Single crystals of these Arsenates are chemically and thermally stable, non-hygroscopic and are highly resistant to high intensity laser radiation. Crystals of KTA are important for second harmonic generation (SHG), sum and difference frequency generation (SFG)/(DFG), optical parametric oscillation (OPO), electrooptical Q-switching and modulation and as substrates for optical waveguides. OPO devices based on these crystals are reliable, solid state sources of tunable laser radiation exhibiting energy conversion efficiencies above 50%. KTA has a very high damage threshold. No optical damage has been observed at the levels of 10 - 20 GW/cm2 with the picoseconds dye laser pulses.

Wavelength selection

In tunable lasers, wavelength selection is an essential requirement. Some of the wavelengths tuning techniques for selecting a specified wavelength are the use of prism, grating, intra-cavity etalon, birefringent filter etc. The most commonly used technique is the birefringent filter, which was demonstrated in 1973. It consists of a single thin birefringent material located inside the laser cavity at the Brewster angle, with the birefringent axis lying in the plane of the crystal. If the wavelength of interest corresponds to an integral number of full wave retardation, laser functions as if the filter is absent and the specific wavelength is emitted. The laser polarization is modified for any other wavelength and suffers heavy losses at the Brewster surfaces. The losses for the unwanted wave lengths can be increased by increasing the number of crystal plates, which are similarly aligned. By rotating the birefringent crystal in its own plane, the wavelength tunability is achieved.

Birefringent Filter
Birefringent Filter

The adjoining figure depicts the birefringent tuning element employed in most of the tunable lasers. The birefringent element is usually made of crystal quartz or calcite and is mounted at Brewster's angle. Light traveling through this element is resolved into two components, one polarized along the fast axis and one polarized along the slow axis. These two components travel at different speeds and, thus, become more out-of-phase as they travel through the element. The thickness of this element is adjusted such that it results in a retardation of one full wavelength for the wavelength of interest for the slow ray. When this element is used where wavelength band is present and passes through it, only one of the wavelengths will actually be retarded by exactly one wavelength. Other wavelengths will be retarded slightly more or less. The wavelength that is retarded by exactly one full wavelength will emerge with its polarization unchanged. All other wavelengths will have an elliptical polarization with a horizontal component. These horizontal components will be reflected from Brewster's-angle surfaces in the system, producing losses for all wavelengths except the one passed unchanged by the filter.

Additional filter elements can be added to achieve narrower bandwidths. The second element is twice the thickness of the first, and the third element is four times the thickness of the first. Each additional element further reduces the output line width.

The birefringent filter is tuned by rotation about an axis perpendicular to its optical surfaces. If the filter is positioned so that its "slow" axis is horizontal, the slow component of the light experiences the greatest retardation. The angle between the slow axis and the light transmission direction changes with the rotation of the filter. It becomes minimum when the slow axis lies in a vertical plane. Reduction of this angle also reduces the retardation effect. This allows the slow ray to travel faster as the slow axis becomes more vertical. Under these conditions, a different wavelength will experience exactly one full wave retardation at different angular orientations of the filter.

Ceramic Lasers

Solid-state laser technologies made a significant progress last decade. Major part of this progress is attributed to the laser diode (LD) pumping. In the late nineties, the new solid-state laser material, ceramic YAG, achieved prominence because of its tremendous application potential. Two Japanese groups developed ceramic laser by different techniques. Dr. Ikesue demonstrated the first laser oscillation in 1995. But his method, hot press method, is good for microchip lasers only and has limited scalability. Dr. Yanagitani, Konoshima Chemical Company, published the patent on pure chemical method for ceramic YAG laser component. Nanometer size precursor and nano-YAG-crystal grow to micro-crystals with grain size of 10 micron through the solid phase crystal growth. The technique has become a major milestone for the major applications of solid-state lasers including National Ignition Facility (NIF) Programs and Solid-State Heat Capacity Laser (SSHCL) programme for Directed Energy Weapon systems.

The fast growing interest in the development of these ceramic lasers has led to intensive research in this area. Over the past few years, polycrystalline ceramics have emerged as a viable alternative to the single-crystal hosts, which are based on a rare-earth dopant in a crystalline host material. A ceramic laser is a real revolution in solid-state lasers. It has a nature of crystalline laser like large and homogeneously broadened emission cross-section, thermal conductivity, and mechanical constant. But the fabrication process is really glass-like-fabricated.

Lasers based on these ceramic materials have several advantages:

  • Since no tedious growth of single crystals is required, ceramic lasers can be significantly less expensive than conventional lasers.
  • The ceramic materials can be custom-fabricated with spatially tailored doping concentrations and index profiles.
  • The biggest advantage of ceramic laser to the single crystal laser is the scaling to the large aperture size. Samples of the size of 10 by 10 by 2 centimeters have already been fabricated and are being used in heat capacity solid-state laser applications. Demonstration of a large aperture sample of 1m x 1m in the ceramic forming process has already been reported for its application for the meter-size ceramic lasers for laser fusion driver.
  • The slabs of these materials can be obtained regularly, on time, and without unexpected additional costs. Ceramic materials can be made any size and shape. The time required to produce the slabs from start to finish is much shorter than the time to grow crystal boules-days instead of weeks. In addition, multiple samples can be fired in one furnace at the same time.
  • Ceramic slabs are also tougher than single-seed crystal slabs and much less apt to undergo a catastrophic fracture. When a crystal slab fractures, the fracture can "run," extending some distance from the original crack and often branching or making a random turn into the center of the crystal to relieve stress. Because cracks are impeded by grain boundaries, ceramic fractures don't run as easily or randomly.
  • Ceramics also measure lower residual stress, which is stress that resides in a material after it has been manufactured. Significant residual stress distorts the laser beam and can make the material more susceptible to cracking.
  • Further, Ceramics can accommodate higher concentrations of dopants (rare-earth ions such as neodymium), which could permit pumping at wavelengths that might otherwise be impractical.
  • Dopant concentrations are highly homogeneous in ceramics and can be controlled precisely. In crystals, dopants tend to segregate toward the bottom of the growing boule.
  • Ceramics also offer the possibility of novel composite structures. For example, a single slab could have an "active" layer of YAG doped with neodymium ions and another layer composed of YAG doped with chromium ions. Such a design is called a passive Q-switch, which turns on the laser after saturation. Another possible approach is to embed different powders with the same host before sintering the slab to create a gradation of neodymium ions or incorporate the passive Q-switch.

The ceramic laser materials are being produced by forming a nanopowder of ingredients into the desired shape followed by sintering in vacuum to form an aggregate of micro crystals that exhibit optical and thermal qualities almost identical to those of a single seed crystal. Livermore researchers are experimenting with several methods to make transparent ceramics. Like Japanese scientists, they begin with a solution of yttrium, neodymium, and aluminum salts and add a solution of ammonium hydrogen carbonate. The precipitate is then filtered, washed, and dried. At this point, the co-precipitated amorphous carbonate is made up of agglomerates of particles measuring about 10 nanometers in diameter. The particles are heated to about 1,100°C to decompose the carbonates and obtain particles of neodymium-doped yttrium-aluminum-garnet (Nd:YAG) measuring about 100 nanometers in size. Highly agglomerated, the particles are treated ultrasonically, and then the large particles are removed to obtain a uniform small size. In a process called slip casting, a suspension of the fine powder is poured into a plaster of paris mold and allowed to settle. Excess water is poured off, and the mold is set aside to absorb most of the remaining water and dry. The result is a porous structure called a preform structure, which is removed from the mold. The preform still contains many pores and is only about 40 to 45 percent dense. The preform structure is then fired in a vacuum at high temperature for many hours. This sintering process involves surface atom diffusion, resulting in the particles fusing together and decreasing the total surface energy. Some of the pores are squeezed out, and the structure shrinks but still retains its overall shape. Additionally, many physical and thermal properties undergo dramatic improvements during sintering.

Because the sintering process still leaves a few trapped pores, the ceramic parts are subjected to a 1- to 2-hour treatment in a hot isostatic press. The press drives out the last pores by heating the sample to high temperatures under enormous pressure of the order of several hundred megapascals. Provided that no impurities exist, the remaining trapped pores collapse, and the finished part achieves the greater than 99.99 percent theoretical density required for nearly perfect transparency.

Recently, Japanese scientists have developed techniques to produce ceramic parts that rival the transparency of traditional crystals (grown from a single seed) and exceed a single crystal's fracture resistance and robustness of manufacturability.

In addition to the National Ignition Facility (NIF) Programs Directorate and Solid-State Heat Capacity Laser (SSHCL), Livermore researchers have also been looking at other possible applications of these remarkable materials for use in other Livermore lasers. Potential applications include scalable components and advanced drivers for laser-driven fusion power plants.

In its current configuration, the SSHCL has four transparent ceramic insulators, called amplifier slabs, measuring 10 by 10 by 2 centimeters that are pumped by 16 arrays of battery-powered laser diode bars.

With the transparent ceramic slabs in place, the SSHCL can generate 25,000 watts of light for up to 10 seconds at 10-percent duty cycle. The SSHCL is pulsed, turning on and off 200 times per second to generate a beam that can penetrate a 2.5-centimeter-thick piece of steel in 2 to 7 seconds depending on the beam size at the target. The system recently achieved 67,000 watts of average power with five ceramic slabs for short fire durations. The laser, which is powered by batteries, is being pursued as part of the U.S. Army's program to develop directed-energy technologies to defend against missiles, mortar shells, and artillery. Unlike chemical lasers designed for the same purpose, an SSHCL is small enough to be installed on a transport vehicle or helicopter. An SSHCL can also be used to clear land mines. Its pulses can dig through several centimeters of dirt to expose and neutralize a mine.

Yamamoto and colleagues are designing a megawatt-class, solid-state ceramic laser that builds on the success of the ceramics in the SSHCL. The new design features 16 ceramic laser slabs measuring 20 by 20 by 4 centimeters.

Tests show that the transparent ceramics exceed specifications. The amount of scattered light, for example, is similar to that measured from single crystals of Nd:GGG or Nd:YAG. The ceramic slab contains tens of thousands of boundaries between microcrystallites, or "grain boundaries," in the path of the laser light. However, the laser light passing through doesn't "see" the many grain boundaries that measure less than 1 nanometer wide. "The performance of transparent ceramic slabs in the SSHCL is astounding, easily meeting or surpassing the performance of the crystal Nd:GGG slabs."

Recently, scientists at Nanyang Technological University in Singapore, at the University of Electro-Communications in Tokyo and at Konoshima Chemical Co. Ltd. in Japan reported what they believe is the highest efficiency reported from a diode-pumped ceramic Yb:Y2O3 laser. Pumping the laser with 976-nm radiation, of which 2.8 W was absorbed in the ceramic material, they observed 1.74 W of output at 1078 nm with a slope efficiency of 82.4 %. Their Yb: Y2O3 sample had 8 percent atomic doping, producing three absorption peaks and three emission peaks. Two of the emission peaks viz. at 950 and 1031.2 nm overlaid absorption peaks and were poor candidates for laser action. To maximize the quantum efficiency, the scientists pumped their ceramic laser at 976 nm and obtained laser action at 1074 nm. With an end-pumped, 3 × 3 × 2-mm-long Yb:Y2O3 ceramic sample inside a 5-cm-long, nearly hemispheric resonator, they obtained the 82.4 percent slope efficiency.

Heat Capacity Lasers

High average power output of the order of even few tens of kilowatts is not obtainable from solid-state lasers due to their poor efficiency and thermal constraint. Solid-state lasers have very low efficiency and consequently the unutilized input energy heats up the laser rod.

There are three possible modes of laser operation. The first one is single shot operation like the one being used in laser range finders or at higher energy level in the NIF, which uses laser glass as the lasing media. In this case, the thermal effects in the solid-state medium are not very important and the cooling is provided through ambient atmosphere after the laser shot. Since it is a single shot or the time between the shots is much larger, the laser media gets cooled by the time next shot is fired. Net result is that there is no thermal gradient in the lasing media at the time of firing a shot.

The second mode of operation is steady state operation, which is typically the case for most of the solid-state lasers based on Nd: YAG. The laser medium is continuously cooled while it is also being pumped. Solid-state lasers have very low efficiency and consequently the unutilized input energy heats up the laser rod. As the cooling fluid flows over the surface of the laser medium, either rod or a disc, a temperature gradient is developed between the center of the medium and the surface. This is due to the fact that the cooling of the surface is faster than the of the central region of the medium, since the cooling of the same takes place depending on the thermal conductivity of the material, which is very poor. As the material cooling rate is rather low compared to that of surface, temperature gradient is produced in the material. If the thermal gradient increases beyond a certain value, laser action becomes more and more inefficient and may even cease. Further, thermal birefringence thereby produced also considerably reduces the beam quality of the laser. Figure shows the temperature gradient in a laser medium under steady state operation. This temperature profile induces a tensile stress in the medium. Higher the power levels, more is the waste heat deposited in the material and thus higher are the thermo-mechanical stresses (of tensile nature). Since there is a limit to which a material can be subjected to stresses before the fracture limit, it sets the limit how much power we can extract in lasers in steady state conditions.


Temperature profile in lasing media before start of lasing operation in any mode
Temperature profile in lasing media when laser is running in a steady state operation

The limitation of steady state operation can be overcome by operating the laser in a novel mode i.e. heat capacity mode, which is intermediate between the above two modes viz. single shot and a steady state. In this case, single shots are rapidly fired on a time scales, which are short as compared to thermal diffusion times through the laser medium. Under these conditions, the build up of thermal gradients is avoided and the device basically has the thermo-optic properties of a single shot device. The waste heat generated during lasing is stored in the active medium, whose temperature rises from the initially achieved starting value to a temperature where laser operation ceases. At this point the medium is again cooled to the initial temperature so that new lasing sequence can begin. Lasing times of many seconds typically up to 10 sec can be achieved generating up to megawatts of levels of burst power during this time. This burst operation makes the heat capacity laser concept more suitable for applications, which require large amount of energy, but for a short period of time. In the HCL concept, there is inversion of the temperature profile through out the medium, as compared to the normal steady state lasing approach. In fact, total energy that can be extracted depends on the heat capacity of the active medium and the temperature difference over which it is operated. To generate higher output, one has to choose a material with higher heat capacity and avail a technique to increase the temperature difference, like cooling the system to liquid nitrogen temperature.

Figure shows the temperature gradient in a laser medium under heat capacity mode operation. This temperature profile induces a compressive stress in the medium. Higher the power levels, more is the waste heat deposited in the material and thus higher are the thermo-mechanical stresses (of compressive nature). Since for laser materials, compressive fracture strength is about 5 - 6 times higher than that of the tensile strength, the laser can be pumped much harder thereby yielding higher outputs.


Temperature profile in lasing media just after vigorous cooling and before initiation of new burst
Temperature profile in lasing media immediately after laser burst is over in heat capacity mode

After every burst, the laser material is vigoursly cooled for making it ready for the next run. In the beginning of the new run, the temperature profile in the material will be such that the surfaces will be cooler as compared to the center of the material. As soon as we start firing new burst, the surface temperature starts rising faster as compared to the center of the laser material, thus first making the profile even and then inverted later on. On the other hand, in steady state operation, the surface temperature is always less than the center of the laser medium to start with, and this difference continuously increases till a steady state is reached. The temperature difference decides the maximum output power, one can extract from the material. In terms of thermal gradient, the thermal gradient goes on increasing with time in case of steady state operation, whereas in case of heat capacity mode, these gradients rather decreases first for some time and then increases later on. Time within which the gradients develop for the laser to cease really decides the duration of the burst.

The heat capacity concept was demonstrated in 2001 at Lawrence Livermore National Laboratory by constructing a 10 kW average power that operates in this mode. Initial demonstration used laser glass as a lasing medium and pumped was carried out using flash lamps. Although the prototype uses Nd: glass for its laser amplifier disks, the upgraded versions use Nd:GGG. Compared with Nd: glass, Nd: GGG boasts a higher mechanical strength and higher thermal conductivity, which, in combination, allows to rapidly cool the disks between runs and reduce the turnaround time between laser firings. To pump these Nd: GGG amplifier disks, the SSHCL uses arrays of laser diodes instead of flash lamps because diode arrays are more compact and efficient than flash lamps and, more importantly, diode radiation generates less heat in the Nd: GGG laser crystals. The Nd: GGG is also twice as efficient in converting pump energy to output beam energy. However, there was a big challenge to grow the crystals large enough to manufacture the nine 13-square-centimeter slabs needed for the upgraded 100-kilowatt laser. Northrop /Grumman Poly-Scientific, the commercial partner responsible for growing the crystals, have attempted to produce high-optical-quality Nd: GGG crystals up to 15 centimeters in diameter.

In 2001 itself, Professor Ueda of Univ. of Electro-Communications, Tokyo Japan demonstrated the potential of ceramic lasers. Ceramic rods of Nd: YAG of 100 mm length and 3 mm diameter pumped by diode lasers were reported to yield powers of the order of 2 kW with a potential to deliver up to 10 kW. High efficiency operation was demonstrated in the end-pumping scheme and the optical-optical efficiency was measured to be about 60% in 1% and 2% doping. This was almost the best data for a single crystal and was a clear evidence to show the high quality of ceramic YAG material. Since ceramic laser media has many advantages over their crystal counterpart (see section on ceramic lasers), scientist working at Lawrence Livermore National Laboratory immediately thought of using ceramic media for their heat capacity laser programme.

During the last several years, Lawrence Livermore National Laboratory has been developing high-power solid-state lasers for tactical battlefield applications. These lasers are based on a compact, flexible, single-aperture, mobile architecture that can be readily scaled to engagement-level powers (~ 100 kW). Looking at the potential of ceramic media for heat capacity lasers, the lasing medium is a series of diode-pumped solid-state ceramic slabs, producing a beam at a wavelength of approximately 1 micron. During lasing operations, the waste heat is stored in the slabs. In a field device, the slabs would be rapidly interchanged with cool slabs, after several accumulated seconds of lasing.

It has been reported in 2005 that the laboratory laser has four ceramic YAG slabs pumped by diodes at a pulse repetition rate of 200 Hz. The aperture size is 10x10 cm2. With this laser, routine operation has been achieved at a time-averaged power of about 25 kW (125 J, 200 Hz) for several seconds. Since the laser has a pulsed format, this is the power averaged over an interval longer than several pulses. The pulse length is about 0.5 ms, giving a duty factor of 10%. The time-averaged power is the same as the equivalent CW power. With the transparent ceramic slabs in place, the SSHCL can generate 25,000 watts of light for up to 10 seconds at 10-percent duty cycle. This pulsed SSHCL can generate a beam that can penetrate a 2.5-centimeter thick piece of steel in 2 to 7 seconds depending on the beam size at the target. The system recently achieved 67,000 watts of average power with five ceramic slabs for short fire durations. The laser, which is powered by batteries, was conceived as part of the U.S. army's program to develop directed-energy technologies to defend against missiles, mortar shells, and artillery. SSHCL is small enough to be installed on a transport vehicle or helicopter. This SSHCL can also be used to clear land mines. Its pulses can dig through several centimeters of dirt to expose and neutralize a mine.

Based on the success of ceramic slabs, efforts are on to upgrade the system up to 100 kW (500J, 200 Hz) by designing a megawatt-class, solid-state heat capacity ceramic laser based on 16 ceramic laser slabs measuring 20 by 20 by 4 centimeters.

Dye Lasers

Dye Lasers use an organic dye as the gain medium. The wide gain spectrum of available dyes allows these lasers to have high degree of tunability with high resolution and high power. The dye laser initially developed by Schmidt and Schafer and Sorokin and Lankard in 1967 was based on flash lamp pumped. Since the dyes used in tunable dye lasers are fluorescent, another light source is always required to pump the dye in order to achieve the population inversion. The pump beam used to excite the large dye molecules and produce the population inversion is a strong light source either a flash lamp or another laser focused on the dye stream. Typical absorption and emission spectra of a common laser dye molecule, which are large organic molecules, is shown in the adjoining figure. One can see that both emission band and the absorption bands are quite broad but the emission band is lower in frequency as compared to absorption band. The dye will absorb only those wavelengths of light, which are shorter than those, which it emits, since some input energy will always be absorbed in the form of vibrations or heat.

Emission Spectra of Dye Lasers
Emission Spectra of Dye Lasers

The characteristics of the light used in the excitation determine the characteristics of the laser. If a pulsed source like flash lamp is used to pump the dye laser, the beam will also be pulsed, on the other hand, if a continuous-wave laser like argon laser pumps the laser, the dye laser's beam will also be continuous.

The energy absorbed by the dye creates a population inversion, moving the electrons into an excited state. Typically, the dye molecule de-excites spontaneously into a metastable state having relatively longer lifetime. It stays there till stimulated emission occurs from the de-excitation of other molecules in the dye. Since the dyes commonly employed are large and have many lower-energy states available for the excited electron to decay into, a large range of de-excitation energies and thus, a large range of output wavelengths are available to the dye. Typically, in the case of Rhodamine 6G, this spectrum of usable wavelengths is quite large, about 130 nm. The dyes most commonly used in dye lasers are Rhodamine 6G combined and Coumarin which are dissolved in a liquid and pumped through the optical cavity in order to prevent any one part of the dye from becoming exhausted during the excitation process. Now- a- days solid dye cells are also being employed in order to increase convenience and make the system portable. In spite of the large gain bandwidth of organic dyes, no one dye can cover the entire visible spectrum; typical tunable range for various dyes is given the following table:

Some of the Important Dyes and their Wavelength Tuning Range

Name of Dye Tuning Range (nm)
Rhodamine 6G 573-618
Rhodamine B 600-646
Coumarin 47 436-486
Coumarin 102 454-506
Coumarin 307 478-547
Coumarin 153 517-590
Disodium Fluorescein 535-565
Bromo Fluorescein 530-690
Oxazine 170 672-727
Pyridine 2 710-790
Styryl 9 803-875

The most important attribute of the dye laser is its tunability, which gives the user access to essentially any wavelength in the visible and near-visible spectrum. The spectral range of ion-laser-pumped CW dye lasers is essentially complete coverage from 400 to 1000 nm. It is even possible to extend their CW tuning range by using nonlinear optical methods to generate wavelengths further into the ultraviolet and infrared region. Energy band diagram of dye lasers is shown in the figure.

Energy band diagram of Dye Lasers
Energy band diagram of Dye Lasers

Simplified picture of singlet states is shown in the figure below.

Simple Energy band diagram of Dye Lasers
Simple Energy band diagram of Dye Lasers

Typically the dye molecules are large organic molecules and have many internal degrees of freedom both vibration and rotation resulting in broad overlapping of energy levels. For laser oscillation the dye the intense pump laser using either flash lamps or other lasers such as Argon or Krypton excites molecule's absorption band. The process is as follows:

  • Molecule absorbs light and populates fist excited singlet state S1 with electrons
  • Electrons in upper vibrational levels of S1 undergo vibrational relaxation and the electrons move to the lowest vibrational level of S1. This excitation energy is then rapidly redistributed within the S1 state within a time period of few picoseconds. Thus the molecules very quickly dissipate this very high energy by internal conversion - the electron density moves to the lowest excited state, S1. Internal conversion occurs by the electron density transferring from the vibrational levels of the upper excited state to vibrational levels of a lower excited state, which are overlapping. This is a radiationless transition i.e. it does not emit a photon of energy.
  • The molecule decays from the S1 state to the S0 state and the energy reappears as fluorescence photons. This process may take few nanoseconds. Internal conversion may occur in S0 as well.
  • There is a possibility of transfer of energy from S1 singlet state to T0. This happens in case the lifetime of S1 is more than of the order of 100 nanoseconds. If it happens, then the energy is lost and efficiency of dye lasers is reduced.

For an efficient dye Laser the energy transfer from singlet state S1 to Triplet state T0 should be avoided, as it is the main loss mechanism within the dye molecule. This can be avoided if the dye molecules have low fluorescence lifetime for S1 to S0 transition as compared to time required for transfer of population from S1 singlet state to T0, the triplet state. All successful dye lasers use dyes with typical fluorescence lifetime for the S1-S0 transition of the order of few nanoseconds.

Triplet absorption in excited dye systems is a major factor that limits the proper laser action. That is, the laser pulse terminates before the pump pulse ends. In fact, the laser pulse usually terminates before the intensity of the pump pulse has fallen below the threshold excitation value. Efforts have been made to overcome this situation. Most common method often used is to add a second molecule to the dye solution to act as a triplet-quenching agent. Collisions between quencher and dye molecules are responsible for this de-excitation process. Usually triplet quenchers are laser dye specific: for example, Cycloheptatriene and cyclooctatetraene (COT) are good triplet quencher for rhodamine 6G. Adamantane is also sometimes added to some dyes to prolong their life

Since most organic dyes have a large range of wavelengths over which amplification can occur (called the gain bandwidth), lasers built around them can be composed of light waves spanning a range of wavelengths in the spectrum. This makes possible the ability to select the wavelength of the laser light through the adjustment of a prism or grating. This tunability feature allows certain specific applications to be performed at minimal cost as compared to having large number of different monochromatic lasers.

Some of the salient features of dye lasers are listed below:

  • A Negative aspect of dye lasers is that the dyes have limited productive lifetimes.
    The factors that limit the lifetime of laser dyes are mainly the chemical and photochemical degradation of the dye in solution. Representative lifetimes of the typical CW dyes range from 300 to 4000 hours depending on the dye.
  • With broad tuning ranges and narrow line widths, single-mode CW dye lasers can provide an impressively large number of resolution elements.
  • The threshold pump source intensity is given as
    Where h~ is, Planck's constant, νp is the pump frequency, τ is the spontaneous emission lifetime and Φ is fluorescence efficiency. The quantities σp and σc are the singlet pumping, and emission cross sections, respectively. N is the density of dye molecules, l is the dye-jet interaction length, and 'I' is the intensity of the pump laser. T is the total resonator loss, which is the sum of the output coupling transmission and the actual scattering and absorption losses. This gives a rough value for the pump threshold intensity 'I' as (2 - 5) x 105 W/cm2. Such high intensities can be obtained continuously only at the focus of a strong pump laser. Typical geometry is shown here.
  • The optical cavity of the dye laser consists of mirrors. The output coupler is usually a long-radius mirror with transmission typically between 10 and 20 percent. The high-reflectance and the output mirrors are curved and mounted at the proper separation and alignment to produce a focal point in the beam within the dye. Pump light in the form of the laser beam is focused into the dye cell at the same point as the dye laser beam. The actual power density is much higher than the threshold power density typically by a factor of 5. For this the beam from a pump laser must be focused down to a very small diameter, ~10-20 μm. However, the focused beam can create a "hot spot" in the dye. Thermal gradients in the dye solution then result in optical inhomogeneities and subsequent distortion of the output beam.
    Simple cavity system for Dye Lasers
    Simple cavity system for Dye Lasers
  • For dye laser applications, dye-absorption band must overlap one of laser lines. Since the absorption in dye is in UV and blue and emission band is to the red of the absorption band; this means that for visible dye lasers we need strong blue and UV pump lasers. In practice, we are presently limited to the argon and krypton-ion lasers. The main reason is that they are the only CW lasers that can produce high enough power of the order of several Watts in a good single-spatial mode and in the visible or UV region of the spectrum. Other pump sources include high-power green light (532 nm) obtained from frequency doubling the output of CW YAG lasers.
  • The spectral range of ion-laser-pumped CW dye lasers is essentially complete coverage from 400 - 1000 nm. Use of nonlinear optical methods makes it possible to extend their CW tuning range further into the ultraviolet and infrared.
  • The power levels available from CW dye lasers are generally sufficient for spectroscopic applications. Though the output power of CW dye lasers varies with the type of dye, but typical CW systems produce between 100 mW to few Watts of output power.
  • The dyes used in these lasers are highly toxic and so must be handled with proper care. The solvents used for dyes include methanol, ethylene glycol, dioxane, dimethylsulphoxide.
  • The most useful feature of dye lasers is their tunability implying that the lasing wavelength for a given dye may be varied over a wide range. The fluorescent line widths are very broad in the range of 50-150 nm in organic dyes; one can use a wavelength-dispersive optical element such as a diffraction grating or prism in the laser cavity to perform selective tuning. Such tuning can result in extremely narrow linewidths. Initially in order to restrict the laser to operate in a single-transverse mode (TEM00) mode, suitable intra-cavity aperture is used. Further, for obtaining a single longitudinal mode (SLM), suitable dispersive elements, such as prisms, gratings, or a combination of these are introduced. A well-designed multiple-prism grating assembly can restrict oscillation to a SLM. Wavelength tuning is generally achieved by rotation of either a grating or a mirror.
  • Wavelength tuning can also be accomplished using either a tuning wedge which is a thin wedged etalon placed in the optical cavity or a birefringent tuning element, which consists of three birefringent elements mounted together at Brewster's angle. These birefringent elements include full wave plate and additional filter elements. When white light travels through full wave plate, one of the wavelengths will actually be retarded by exactly one wavelength. Other wavelengths will be retarded slightly more or less. The wavelength that is retarded by exactly one full wavelength will emerge with its polarization unchanged. All other wavelengths will have an elliptical polarization with a horizontal component. These horizontal components will be reflected from Brewster's-angle surfaces in the system, producing losses for all wavelengths except the one passed unchanged by the filter. Additional filter elements are added for narrower bandwidths. Each additional element further reduces the output line-width. The three-element filter has a typical output bandwidth of 0.025 nm. However, a CW dye laser with either a birefringent filter or a tuning wedge will have several longitudinal cavity modes present in the laser output. For single-mode operation, a thicker etalon for mode selection must be added to the cavity. In some cases, two etalons are added for greater control of the exact laser frequency. Or alternatively, a well-designed multiple-prism grating assembly can restrict oscillation to a single longitudinal mode SLM
  • The methods mentioned for wavelength selection and tuning in CW dye lasers may also be used with pulsed dye lasers. However, diffraction gratings also may be used as with nitrogen-pumped dye lasers. A light beam incident upon a plane reflection grating will be reflected back along the axis of incidence if the following grating equation is satisfied:
  • Dye lasers can be operated in both pulsed and CW modes. In the pulsed mode, these are usually pumped either by flash lamps or by other lasers such as pulsed-nitrogen lasers or copper vapour laser or excimer lasers or frequency-doubled Nd:YAG . On the other hand, in the continuous mode, the output of a CW argon ion laser generally is preferred as the pumping source.
  • CW dye lasers use a dye jet as the active medium. Their excitation mechanisms are the focused beams of argon ion lasers. Wavelength tuning is accomplished with birefringent filters or tuning wedges. On the other hand, pulsed dye lasers employ a dye cell and are pumped either by the short-duration pulse of a nitrogen laser or copper vapour laser or excimer lasers or flash lamps. However, nitrogen, copper vapour and excimer laser pumped dye lasers can produce several hundred pulses per second with pulse durations of a few nanoseconds, whereas flash lamp pumped systems are typically capable of producing a few pulses per second, with pulse duration of several microseconds.
  • Although liquid dye lasers have been very successful, there has been continuous effort to find new gain media in the solid-state that would simplify the engineering of this class of lasers. The gain media used in these lasers are organic dye doped polymers. There is some work on dye lasers based on solid media, e.g. with the dye in a polymer matrix. Obviously, the solid-state form has many advantages, particularly concerning handling. There are a number of materials, which have been used as solid hosts for laser dyes such as polymers, porous glasses, organically modified silicates or silicate nano-composites, polycom glass (combination of polymer and sol-gel) or inorganic compounds such as tetraethoxysilane (TEOS) and tetramethoxysilane (TMOS). However the most promising solid-state lasers so far has been based on polymeric and sol-gel glass. In general, the polymers used include polyacrylics, polyurethanes, polycarbonates, polymethylmethacrylate (PMMA), modified PMMA known as MPMMA, copolymer of heterogeneous mixtures of monomers such as methylmethacrylate (MMA) and 2-hydroxyethyl methacrylate (HEMA). These new solid materials show excellent lasing efficiencies of 45 - 50 % with good photostability. . Further the laser damage threshold also increases significantly. Typically, Rhodamine 6G dye doped in polymer and porous glass composition can withstand pump pulses of 50 ns with energies of 7-10 J/cm2. In the hybrid matrices based on TEOS, both the lasing efficiency and stability are encouraging. Laser efficiencies of up to 26% and laser emission with no sign of degradation, even after 100 000 pump in operations at the same position of the sample at 10 Hz repetition rate have been demonstrated.
  • Though solid-state lasers based on Ti: sapphire are now being used for a particular wavelength range as these can produce high enough powers without any lifetime restrictions, still dye lasers are dominating the fields of tunable lasers and ultrashort pulse generation for areas such as spectroscopy where large number of wavelengths are required. So as such the future of laser dyes particularly solid-state dye lasers continues to be good.
  • Important properties of dye lasers can be summarized below.

Important Properties of Dye Lasers

Property Value
Wavelength Range 400 - 1000 nm
Average output Power Few miliwatts to few Watts
Slope efficiency Upto 50 %
Threshold intensity (1 - 5) W / cm2
Gain (1 - 2.5) cm-1
Saturation Intensity 3.4 x 109W/cm2
Divergence 1 - 2 mrad
Beam Diameter 0.4 - 0.6 mm
Line width attainable after tuning 0.001 - 0.025 nm

Applications

  • Organic dye lasers, because they are both tunable and coherent light sources, are becoming increasingly important in spectroscopy, holography, and in biomedical applications.
  • Nanosecond pulses, picoseconds and femtosecond pulses can be generated for spectroscopy applications using pulsed laser pumping and mode locking.
  • The concept of selective photothermolysis with the 577-/585-nm pulsed dye laser (PDL) has revolutionized treatment of relatively common port wine stain (PWS) birthmarks. A port-wine stain is a birthmark in which swollen blood vessels create a reddish-purplish discoloration of the skin. Although PWS can appear in any part of the body, they occur more often on the face and persist throughout life. The majority of PWS can be significantly lightened with the PDL.
  • Lithotripsy is a medical procedure that uses shock waves to break up stones that form in the kidney, bladder, and gallbladder. When a laser is used, a train of laser pulses is guided by a fiber to the application site, which ignites plasma at the surface of the stone. The breakdown of the plasma creates a shock wave, which detaches some fragments. After many repetitions, the stone is fragmented into smaller pieces, which then can pass spontaneously. Typical operational parameters of dye lasers used in lithotripsy treatment are: emission wavelength of 504 nm and 595 nm, depending on stone composition; pulse energy in the range 50-120mJ/pulse; pulse duration from 1-42.5 ns; and repetition rate of 1-10 Hz.
  • Arteries can become narrowed or blocked by deposits called plaque. Plaque is made up of fat and cholesterol that builds up on the inside of the artery walls. Angioplasty is a medical procedure to open arteries that are obstructed by plaque. It involves different forms of minimally invasive vascular interventions, which can be exemplified by balloon angioplasty, a procedure in which a balloon is used to open a blockage in a narrowed artery. Laser angioplasty is a promising alternative method to open arteries obstructed by plaque, with potential advantages over surgery, balloon angioplasty, and other forms of vascular interventions. Dye Laser radiation can be introduced into arteries via small optical fibers, thus avoiding major surgery. The radiation can remove plaque rather than displacing it,
  • A recent important application of dye lasers involves isotope separation. Here, the laser is used to selectively excite one of several isotopes, thereby inducing the desired isotope to undergo a chemical reaction more readily

Semiconductor Lasers

Semiconductor lasers or diode lasers or laser diodes as they are generally referred to, were invented almost half a century ago by Robert N Hall and Marshall Nathan in 1962. Diode lasers of the sixties required threshold current densities of 1000 A/cm2 at 77 K temperatures and two orders of magnitude greater, or 100,000 A/cm2 at about 300 K. Moreover these lasers were pulsed. The main challenge was to operate these lasers at room temperatures continuously with low threshold current densities. The first laser diode to achieve continuous wave operation was a double heterostructure operation demonstrated in 1970 simultaneously by Zhores Alferov at Iaffe Physico-Technical Institute, St. Petersburg Russia and Morton panish and Izuo Hayashi at Bell Labs. Continuous developments have resulted in laser diodes with shorter and shorter wavelengths, increasing output power and an improved beam quality. Today reliable laser diodes stacks with powers in the range of kilowatts are available in the market for applications like Diode pumped solid-state lasers. In addition, compared to other types of lasers, laser diodes use very little power. Most laser diodes can operate with voltage as low as 2 V with power requirements determined by their current setting. Electrical to optical efficiencies in excess of 50% are typical in the case of laser diodes. In this way, Laser diodes have thus grown to a key component in modern photonics technology. Advances in crystal growth technologies, such as metallorganic chemical vapor deposition (MOCVD) and molecular beam epitaxy (MBE), the development of double heterostructure lasers and subsequently quantum well lasers, materials passivation technologies, thermal management technologies; all have further contributed to one of the most enabling technological industries today, that of high-power semiconductor lasers.

As compared to other lasers, semiconductor lasers are:

  • Compact and rugged. This ruggedness and small size allow laser diodes to be used in environments and spaces in which other types of lasers cannot operate
  • High efficiency in the range of > 50%
  • Direct excitation with small electric currents,
  • Possibility of direct modulation with applied current
  • Small beam waist
  • Low costs due to mass production and high reliability

However, there are few drawbacks in semiconductor laser diodes as compared to other solid state and gas lasers. These include, their sensitivity to temperature and large beam divergence and lower spectral purity.

The optical gain in a semiconductor lasers are achieved through the recombination of injected holes and electrons resulting in emission of photons in a forward-biased semiconductor p-n junction. This represents the direct conversion of electricity to light, which is a very efficient process, and practical diode laser devices can achieve more than 50% percent electrical-to-optical power conversion rate, that is almost an order of magnitude larger than most other lasers. Over the past few years, efforts are on for a gradual replacement of other laser types by diode laser based-solutions, as the considerable challenges to engineering with diode lasers are being sorted out. At the same time the compactness and the low power consumption of diode lasers have enabled important new applications such as storing information in compact discs and DVDs, and the practical high-speed, broadband transmission of information over optical fibers.

Semiconductor lasers also have basic three components

  • A gain medium that amplifies light (p-n junction)
  • An energy source to create population inversion (electrical current through the junction)
  • A resonant cavity for confining the light (cleaving the semiconductor creates a reflective facet which can be used to create a laser cavity)

Under normal conditions, absorption of photon results in the generation of an electron hole pair: the condition, which is used for detection of light. The recombination of an electron hole pair results in spontaneous emission of photon: the principle of operation of light emitting diodes (LED). Electron hole combination can be stimulated by a photon thus inducing emission of identical photons: the principle of operation of semiconductor lasers.

Solid-state and gas lasers work on narrow optical transitions connecting discrete energy levels between which population inversion is achieved by optical or electrical pumping

Semiconductor lasers, on the other hand, work on transitions between energy bands in which conduction electrons and valence holes radiative recombination across the band gap that determines the emission wavelength.

In a semiconductor laser, the transitions are associated with the electron states in the conduction band and valence band. Since the upper and lower energy states are continuous and hence the semiconductor has a broad gain spectrum implying that the output is not sharp. Thus coherence and mono chromaticity of these lasers are poor.

Indirect-gap semiconductors are inefficient light emitters because in case of indirect band gap materials like silicon, transitions between conduction bands to valence band involve phonon for conservation of momentum. Moreover phonon-assisted photon emission involves three "particles" simultaneously (electron, photon and phonon), its probability is low.

Before we go into the basics of semiconductor lasers, we will briefly outline some of the fundamental points related to semiconductor physics. However, for details reader can consult any textbook on semiconductor physics.

Semiconductor materials are crystalline or amorphous solids whose electrical conductivity is somewhere between that of an insulator and a conductor. Examples of Semiconductors materials are silicon, germanium and gallium arsenide (GaAs), They are neither good conductors nor good insulators: that is why the name semi-conductors. They have a small number of free electrons because the atoms are closely grouped together in a crystalline pattern called known as crystal lattice. However, their ability to conduct electricity can be greatly enhanced by adding certain impurities to this crystalline structure thereby, producing more free electrons than holes or vice versa.

Semiconductors contain two types of mobile charge carriers, holes and electrons.

The holes are positively charged while the electrons are negatively charged.

A semiconductor may be doped with donor impurities such as antimony in silicon so that it contains mobile charges, which are primarily electrons. Semiconductor material with electrons as majority carriers is known as n-type semiconductor

Similarly, a semiconductor may also be doped with acceptor impurities such as boron in silicon, so that it contains mobile charges, which are mainly holes. Semiconductor material with holes as majority carriers is known as p-type semiconductor

Electrical and optical properties of semiconductors are determined by the energy distribution of electrons in these materials.

The energy of electrons in solids, just like the energy of electrons in atoms, is limited to certain discrete values. In crystalline solids, these energy levels are grouped into bands, known as allowed energy bands.

In semiconductors, the last completely filled band is called the valence band. This first empty band above the valence band is called the conduction band. Energy Gap between the Valence band and Conduction band is known as Band Energy Gap The energy gap between these two bands is called forbidden gap or band-gap, Eg.

Insulators have large band-gap energies. The material is an insulator if the energy band-gap is larger than about 3.5 eV. For example, band-gap energy of diamond, which is a good insulator, is Eg = 6eV. The band-gap energy of semiconductors is typically between 0.2 eV and 3.5 eV. Materials with Eg < 0.2eV are generally considered as metals.

The semiconductor contains no electrons at a temperature of absolute zero, T = 0K and its conduction band is empty, and thus behaves like an insulator. As the temperature increases, some electrons are thermally excited into the first empty band, i.e. the conduction band.

Fermi level indicates the occupation conditions of electrons or holes in the semiconductor; it is the energy level to which carriers occupy. Fermi level (EFP) for p-type is near the valence band and EFN for the n-type is near the conduction band.

An interface between two regions of a semiconductor or an interface between two different semiconductor materials is called a junction. Junctions between differently doped regions of the same semiconductor material are called a homojunction, while a junction between two different types of materials is called a heterojunction. A junction between a p-type and an n-type semiconductor is called a p-n junction.

Once the contact is made between the 'n' and the 'p' doped material, electrons diffuse from the n region into the p region where they recombine with the abundant holes. Similarly holes diffuse from p region to n region and combine. Electrons leave behind the positively charged donor ions, so some part of n region will be positively charged. Similarly some part of the p side will be negatively charged.

Simplified p-n junction diagram under no bias
Simplified p-n junction diagram under no bias

Due to the diffusion of both types of carriers away from the junction region, a narrow zone around the junction is totally depleted of mobile charge carriers. This region is called the depletion region. The process happens till dynamic equilibrium takes place: the diffusion of electrons/holes and the drift currents cancel, so in the absence of an external field no net current flows across the junction. In terms of band structure, p-n junction can be represented as shown in the figure:

Note that the system of a p-n junction without bias is in equilibrium and hence the Fermi level EFN for n-type and EFP for p-type must be equal implying that there will be band bending. Thus in the absence of a bias, the bottom of the conduction band on the n-side lies lower than that on the p-side. This prevents net diffusion, as the electrons have to overcome a potential barrier qφ.

However, this equilibrium can be disrupted, by applying an external electric field usually known as biasing the p-n junction.

Positive voltage to the p region and negative voltage to the n region is known as forward bias. This allows the current to flow through the junction.

On the other hand, the junction is reverse biased if a negative voltage is applied to the p region and positive voltage is applied to n region. Under reverse biased condition, very little current small current flows.

The width of depletion layer increases with an increase of a reverse voltage and decreases with an increase in the application of a forward voltage

The p-n junction is the basis of optoelectronics devices, such as the light emitting diode (LED), laser diodes,

Under forward bias conditions, if the external voltage becomes greater than the value of the potential barrier, the current will start flowing through the junction. This is because the negative voltage pushes electrons towards the junction giving them the energy to cross over and combine with the holes, which are being pushed in the opposite direction towards the junction by the positive voltage. Thus forward bias creates extra charge carriers in the junction, lowers the potential barrier, and causes injection of charge carriers, through the junction, to the other side.

The laser operation occurs at a p-n junction, that is the boundary region between p-type and n-type materials. When p-n junction diode is forward biased, then there will be injection of electrons into the conduction band along n-side and production of more holes in valence band along p-side of the junction. At the junction, electrons and holes meet and are attracted to each other because of opposite charges. When they meet, they recombine and emit radiation. When a forward-bias voltage is applied to the junction, the barrier height is reduced and some of the electrons in the conduction band will overlap some of the holes in the valence band. It is worth pointing out that pumping the semiconductor raises some electrons to the conduction band where they rapidly distribute themselves into the lowest available energy levels within the conduction band. On the other hand, the electrons in the valence band occupy the lowest energy levels there, pushing the holes to the top of the valence band.

Simplified p-n junction diagram under forward bias
Simplified p-n junction diagram under forward bias

Thus under forward bias conditions, there are more number of electrons than the number of holes in the junction region because of higher mobility of electrons as compared to that of holes. In other words a population inversion, the necessary condition for laser operation. In this situation, radiative recombination of the holes and electrons can occur. Electrons fall across the energy gap and recombine with holes. At very low currents, a population inversion does not occur even though recombination radiation is emitted. Under these conditions p-n junction behaves as a light-emitting diode (LED). In comparison, to produce a population inversion, comparatively high current is required within the junction region. This situation is indicated in the adjoining figure where p-n junction is shown under forward bias conditions.

One can see that EFN is not equal to EFP in case of biased p-n junction. Fermi levels in such situations are known as Quasi Fermi Level.

Under forward biased condition, the recombination process has to be such that the carriers recombine radiatively. For this the probability of band-to-band recombination, which is the most desirable process for generating high-energy photons, is to be exploited.

In an efficient semiconductor laser or even LED, most of the carriers that recombine must result in production of photons, in other words, the recombination should be a result of band-to-band radiative transition. All other recombination processes in which electron energy is lost in producing phonons, as is the case in indirect band gap materials, or in recombination with ionized impurities are undesirable. The efficiency for an efficient device can be characterized by the term "internal quantum efficiency" ηINT, defined as

ηINT   =   Number of band to band radiative recombinations

Number of carriers crossing junction

and is given as

Internal quantum efficiency

where A is an Einstein coefficient describing radiative recombination and X is a coefficient for nonradiative recombination. Thus it is required that A >> X for efficient photon generation. This explains why indirect band gap materials like silicon are not efficient semiconductor laser materials.

The energy of the photon resulting from this recombination is equal to that associated with the energy band gap. In light-emitting diodes (LED) this light energy is transmitted out through the sides of the junction region. In semiconductor lasers the junction forms the active medium, and the reflective ends of the laser material provide feedback. By imposing the appropriate feedback conditions required by all lasers, stimulated emission dominates and laser action can occur. The structure of the laser diode creates an optical cavity in which the light photons have multiple reflections. Ensuring the light is properly reflected is very important for the operation of the device. Usually, the cleaved ends of the laser diode, with no further coating, form the mirrors for output coupling. The typical reflectivity at the interface between gallium arsenide and air is approximately 36%. However, mirror of higher reflectivity is desired at one end to ensure that the laser output comes from one end. Further to reduce the threshold for laser operation, the reflectivity of output coupler and the total reflector can be realized using dielectric coatings. More sophisticated coatings can then be applied to these facets to tailor their reflectivity. Alternatively, Bragg gratings can be inserted at the ends of the cavity to provide reliable single-frequency operation capable of high-speed modulation.

The thickness of the junction region is small, typically around one micron. Thus, light traveling in the plane of the junction is amplified more than light perpendicular to it and the laser emission is parallel to the plane of the junction.

Though effective photon generation is essential for Semiconductor lasers, however, it is equally important to ensure that these photons come out of the active region of the device; otherwise the device will not be efficient. The generated photons may fail to escape from the active region because of the reabsorption of these photons in bulk material between the active layer and the surface. Thus for a bright, efficient, photon emitter we need to ensure that as many carriers as possible recombine soon after crossing the junction, and do not escape to travel into the bulk material far from the junction. In an ideal device electrons and holes should be "trapped" in the region where recombination is desired. Though initial work on these devices has been on homojunctions (e.g. p-type and n-type GaAs), but these devices somehow suffer from poor electron - hole confinement, poor optical confinement, and inefficient injection of carriers thus requiring very high degree of doping. This implies that the threshold current densities are very high and the device has to work as pulsed and also at low temperature typically 77 K. This trapping can be affected by the use of heterostructures semiconductor structures that use heterojunctions. A heterojunction is a junction between layers of different properties (e.g.: different band gap energies) but having almost the same lattice structure. The Double Heterojunction laser (DH-laser) uses four different layers, for example an n-GaAs -layer followed by a N-GaAlAs-layer, then a p-GaAs-layer and eventually a P-GaAlAs one where N and P indicate larger band gaps. The aluminium containing layers have a lower refractive index. Hence this structure with the p-layer as the active region in between them provides good wave guidance. These structures confine the injected electrons and holes to a narrow region about the junction. This requires less current to establish the required concentration of electrons for population inversion. Further these structures also help in photon confinement. A dielectric waveguide around the optical gain region helps to increase the photon concentration and elevate the probability of stimulated emission. This reduces the number of electrons lost traveling off the cavity axis.

For example heterostructure laser diode can be fabricated using two different band gap materials namely GaAs and AlGaAs. The double-heterostructure (DH) laser diode, which consists of a thin layer of low bandgap material such as GaAs sandwiched between two high bandgap layers such as AlGaAs, is one of the most commonly studied geometry. The bandgap discontinuity confines the free electrons and holes to the active region, meaning that more electron-hole pairs can contribute to the amplification. Further, the semiconductor with a wider band gap (AlGaAs) will also have a lower refractive index than GaAs. This difference in refractive index is what establishes an optical dielectric wave-guide that ultimately confines photons to the active region. Use of such structures help in confining both the injected electrons and holes and also the emitted photons to a narrow region about the junction. This as such requires less current to establish the required concentration of electrons for population inversion. Typically, the DH laser has a room temperature threshold current density two orders of magnitude smaller then the homojunction device.

As mentioned earlier, that the homojunctions are no longer being used because of poor confinement of both carriers as well as emitted photons thus paving the way for heterojunctions for the present day practical devices which dominate most applications. These devices have basically a stripe geometry, in which the gain region is confined to a narrow stripe region. This confinement of the laser operation within a stripe region is usually accomplished by either gain guiding or by index guiding. Both these methods confine the light in such a way that the losses due to beam spreading are minimized thereby reducing the current requirements for laser operation. Further, since in the stripe geometry, aperture is limited and the dimensions in the directions parallel and perpendicular to the junction are comparable thus reducing astigmatism. Index-guided lasers employ steps in the index of refraction both parallel and perpendicular to the junction to confine the light. On the other hand, gain guiding structure makes use of composition changes for confinement in the plane of the junction. This is done, by adjusting the charge carrier density in the region, which results in the required refractive index. However, refractive index changes in the direction perpendicular to the junction to confine the light, just as like index-guided devices.

Gain-guided lasers are comparatively easier to fabricate as compared to index-guided lasers, but they have weaker confinement, which leads to somewhat poorer beam quality and stability. Further Gain-guided devices also have somewhat larger astigmatism than index-guided devices. These factors restrict the use of gain-guided devices for some applications where beam quality is important, but for most of the applications gain-guided diode lasers are suitable.

Another rapidly upcoming area of semiconductor laser technology is the development of high-power linear arrays. These devices are high-radiance diode sources suitable for applications like pumping solid-state lasers. The array is fabricated as a bar with a number of stripe laser sources. These devices are capable of emission of kilowatts of optical power. Such lasers are usually available as bar structures and may be stacked to form two-dimensional arrays.

Most of the laser diodes are edge emitters that are also called Fabry-Perot (FP) diode lasers since the cavity is essentially similar to that of a conventional gas or solid-state laser but formed inside the semiconductor laser diode chip itself. The mirrors are either formed by the cleaved edges of the chip or one or both of these are anti-reflection (AR) coated and external mirrors are added for high performance.

However, recently, surface emitting laser diodes (VCSEL: Vertical Cavity Surface Emitting Laser) have also become of interest for special applications. VCSEL have the optical cavity axis along the direction of current flow rather than perpendicular to the current flow as in conventional laser diodes. The active region length is very short compared with the lateral dimensions so that the radiation emerges from the surface of the cavity rather than from its edge. The reflectors at the ends of the cavity are dielectric mirrors made from alternating high and low refractive index quarter-wave thick multilayer.VCSELs emit their beam from their top surface. The cavity is formed of a hundred or more layers consisting of mirrors and active laser semiconductor all formed epitaxially on a bulk substrate. This approach provides several very significant technical advantages in terms of Beam characteristics and lasing threshold. The beam from a typical VCSEL exits from a circular region 5 to 25 um in diameter. Since this is much larger than for the FP laser diode, the divergence of the resulting beam is much lower. And, because it is also circular, no corrections for asymmetry and astigmatism are required - a simple lens should be able to provide excellent collimation. Lasing threshold drive current is an order less than the edge emitting laser diodes. Further, the packing density of such devices can be an order of magnitude higher than for FP laser diodes. As of now the output power is less and VCSEL technology is in its infancy its potential is just beginning to be exploited.

Distributed Feedback Lasers (DFB) is a special category of lasers under edge emitters, which incorporates a distributed grating that acts as a distributed reflector. This results in single mode lasers of high stability, which is the requirement of telecom industry.

Recently quantum well devices for semiconductor technology are being pursued seriously. A quantum well is a very thin layer of semiconductor material between two layers with larger values of band-gap. If the layer is thin enough, 20 nm or less, comparable to the deBroglie wavelength, (λ ≈ h/p), quantum mechanical properties of electrons become important. This changes the energy-level structure of the material. Quantum well devices may incorporate a single quantum well or multiple quantum wells, with a number of alternating thin layers of high-band-gap and low band-gap material. The use of quantum wells in laser devices allows optimizing the properties of the material for the specific application. Quantum well devices offer lower threshold current and higher output power than devices without quantum wells.

Conventional semiconductor lasers can emit wavelengths upto few microns depending upon the band gap of the material. However, in order to generate longer wavelengths, new class of semiconductor lasers, known as quantum cascade lasers can be considered. In conventional semiconductor lasers, the lasers action is due to interband transitions through the recombination of electron hole pairs across the band gap, On the other hand, in Quantum Cascade lasers laser emission is achieved through the use of intersubband transitions in a repeated stack of semiconductor multiple quantum heterostructures. Wavelength range from 2.75-250 μm has been generated using these structures.

Most commonly used material for semiconductor lasers are the III-V compounds such as GaAs, AlGaAs, InGaAs and InGaAsP depending upon the desired lasing wavelength. Recently, GaN/AlGaN and InGaN/AlGaN are also being used to achieve emission in the blue and ultraviolet regions.

The important characteristics of semiconductor lasers are listed below:

Diode output power vs drive current
Diode output power vs drive current

Adjoining figure shows the output power of a semiconductor lasers as a function of current. Above a threshold current, at which the laser diode starts lasing, the laser diode shows almost a linear dependence between optical output power and laser current. Below the threshold the spontaneous emission is predominant and the optical amplification is not sufficient the device behaves like a LED. The current through the junction must exceed a minimum threshold value. It means that it must provide enough holes and electrons so that the radiation generated by their recombination exceeds the losses. Losses may arise from several reasons such as spreading of light out of the active region, transmission of light through the mirrors, and absorption of light by free carriers in the junction.

This steeply rising light output curve can be extrapolated backward to the zero light output intercept, which defines the threshold current. Further, in the linear region, the slope of the output vs current curve yields the electrical-to-optical power conversion efficiency, also known as slope or quantum efficiency. The values of slope efficiency vary from 30% - 80%.

Quantum efficiency

The output characteristics of these devices are slightly different from those of other type of lasers. Because of their small size these have beam divergence angles of as much as high as 20o - 30o. The high value of divergence of semiconductor lasers is because of diffraction of the light waves when couple out of the laser structure. Inside the laser, the light waves are limited to the active zone. Since the active light-emitting area is rectangle-shaped with different length and breadth, the parallel and vertical divergence are also different. This results in the appearance of an elliptical spot at some distance from the emitting area. The ratio of vertical to parallel divergence, measured in the far field, is called the ratio of axes. If we focus such a beam, it will be observed that the focus of the vertical and the focus of the parallel divergence are not congruent but are shifted against each other: the effect known as astigmatism. In order to make a diverging beam parallel, a positive (convex) lens can be used to produce a collimated or focused beam. However, without further correction, the beam profile will be elliptical and the focal distances in both the axes will not be the same due to the astigmatism. Usually a pair of wedge-shaped prisms can be used to circularize the elliptical spot shape. By adjusting the relative orientations of the two prisms, it is fairly easy to effectively correct for this beam characteristic. The astigmatism may, however, be corrected, by the use of cylindrical lenses to form a circular profile. Still the beam divergence is substantially larger than what one is accustomed to with lasers. However there are some like Vertical cavity surface emitting lasers (VCSEL), which have square or round emitting areas and, therefore, can produce relatively symmetrical beam.

The characteristic curve (output power vs. current) of a semiconductor laser strongly depends on the temperature. Higher the temperature, higher is the threshold current and smaller is the slope of the curve in the laser region. The threshold current density for laser operation increases rapidly with increasing temperature. Typically, at the cryogenic temperature of 77 K, the threshold current in a gallium arsenide laser is about one tenth that of the room temperature value. This means that cooling to cryogenic temperatures changes the operating characteristics of the laser. The shift in the threshold current is due to the temperature dependent nature of the carrier concentration in the active layer, whereas the decrease in slope is due to an increasing probability for non-emitting recombination processes. Further, increase in temperature also affects the spectral distribution. With the increase of temperature, the crystal expands and thus increasing the resonator length. Further the refractive index increases whereas the bang gap decreases with increase in temperature. Net result is the output spectral lines drift to longer wavelengths. Typically, the wavelength shift for 808 nm diodes is generally around 3 nm per 10oC. Thus, it is necessary to stabilize the laser temperature. Most of the semiconductor diodes are operated at 77 - 200 K.

Another effect of temperature is on the life of semiconductor lasers. When the temperature is reduced by about 10 degrees, the lifetime is almost increased twice. This is why these lasers are mounted onto a heat sink to avoid an overheating by power dissipation. Life time up to 100.000 hours have been reported

Since charge carriers electrons and holes are injected into the device from the n- and p- side, these lasers are also sometimes called injection lasers.

The coherence length of semiconductor laser diodes is low. Typical values for an index guided Fabry-Perot laser, emitting a single spectral line at 825 nm is 7cm, whereas for a gain guided Fabry-Perot laser, the coherence length is 300μm only.

Typical gas lasers, with a diffraction-limited beam emerging from an aperture around 1.5 millimeters in diameter usually have a circular beam with divergence of a few tenths of a degree. Thus semiconductor diode lasers typically have much lower radiance than other types of lasers. Also, their radiation cannot be focused so well as the light from better-collimated lasers.

Typical gain and Saturation intensity of semiconductor lasers is 103 cm-1 and 2.5 x 109 W/cm2 respectively.

Laser diodes offer many advantages, including small size, lightweight, low power consumption, and high efficiency. They have become widely used as light sources for a wide variety of applications, including compact disk players, printers, magneto-optic data storage, and optical-fiber telecommunications.

Difference between Diode Laser Laser can be summarized as follows


Semiconductor Laser LED
Generation through stimulated emission Generation by spontaneous emission
Monochromatic and coherent light beam Divergent and incoherent light beam
Power output kilowatts Power output in miliwatts
Require feedback mechanism like optical resonator Does not require feedback mechanism
Expensive Cheap
Requires temperature and current stability Easy to handle. No such controls are required
Generally spectral width less than 5 nm Spectral width upto 100 nm

Most Common semiconductor materials and their wavelengths


Material Wavelength
GaAs/AlGaAs 720 - 850 nm
GaAs/InGaAs 900 - 1100 nm
GaAs/AlGaInP 635, 650, 670 nm
GaN/InGaN 380, 405, 450, 470 nm
InP/InGaAsP 1000 - 1650 nm

Applications

  • Telecommunication
  • Optical storage
  • Solid state Laser pumping
  • Material processing such as Welding, drilling and cutting
  • Medical applications in dermatology, dentistry, ophthalmology, in surgery of tumors, kidney stone
  • Barcode scanning
  • Inspection, measurement and control
  • Laser printers
  • DVD drives
  • Laser pointers
  • Environment monitoring

References

Fiber Lasers

We have discussed solid state lasers in earlier section. High power solid state lasers have found wide spread applications in defence, industry, science and technology due to their characteristics such as peak as well as average power and energy, high beam quality and robust and compact. But solid state lasers suffer from thermally induced stress, reducing the quality of the laser output especially when power scaling is carried out. Classic solid state lasers employ laser media in the cylindrical rod geometry of various dimensions. In this case the temperature profile of the rod is with the highest temperature in the centre and slowly coming to lower levels at the periphery in a parabolic shape, causing thermal lensing and consequent generation of thermal stress producing low quality laser output. Due to this, power scaling with high beam quality is compromised in solid state lasers. Other solid state geometries with thin discs and plates also suffer from this problem. The advent of high power fiber optic lasers in this century is a welcome alternative to conventional solid state lasers in generating high power laser output, capable of

The combination of two innovative technologies i.e. active optical fiber as the laser medium and high power semiconductor laser diodes as pump source has resulted in the development of high power fiber optic lasers. It consists of an optical fiber cable with an active core doped with rare earth ions, resonator mirrors and pump light source. Laser diode light, pumped in to the fiber core is absorbed by the active ions throughout the fiber length and is converted to laser radiation. They are guided in the waveguide structure of the fiber cable. Fiber integrated mirrors like Bragg gratings are used as the resonator mirrors. The fiber optic laser is compact and it has long term stability, without any thermal lensing problems. Long length of the optical fibers offer large surface area due to their long length as well as having large ratio of surface to volume, ensures fast heat dissipation producing diffraction limited output. Thermal load also is distributed over the entire length of the fiber cable, thus reducing the thermal lensing build up problems as seen in the case of solid state lasers. It may be mentioned that typical laser rod (active media) dimension in the solid state laser is about few centimeters in length and few millimeters in diameter, whereas the typical length of optical fiber is several meters.

Optical fiber cable is the active component of fiber optic laser, with the core doped with rare earth ions. The choice of the rare earth ion is decided by the wavelength requirement. The fiber cable is a cylindrical dielectric wave guide structure made of silica glass embedded in an outer cladding material of slightly lower refractive index. For environmental protection they are normally incorporated into cables. When light rays are incident at the core at angles lower than the critical angle, they undergo total internal reflection and are guided through the core. Rays with any other inclination lose their energy in the cladding and are not guided. In fact the fiber geometry determines the fiber modes that can propagate in the cable. Propagation characteristics of laser photons in the optical fiber laser are determined by the differences in the refractive indices of the core and the cladding, unlike in the case of solid state laser where it is decided by the external resonator mirrors. The light from the diode laser is pumped directly into the fiber and this fiber is coupled to the laser fiber using fiber couplers. Any misalignment will cause losses. The alignment must be accurate to few micrometers and loss should be less than 0.2% for best results. Fiber couplers are basically one way devices in the sense that the light from the pump source is not sent back to the source.

Fiber optic laser is also a solid state laser, but an unconventional one. Here the photons move in a zig-zag path guided by total internal reflection phenomenon unlike in the case of solid state lasers, where the photons move to and fro in a straight line between resonator mirrors. Pump radiation (high power semiconductor diode laser) launched in to the fiber, is absorbed by the active core and is converted to laser radiation. Both the pump radiation and the laser are guided in the active waveguide structure. The laser cavity is constructed with dielectric mirrors at the end of the fiber.

A variety of materials have been used as the basic fiber material in which the rare earth ions like ytterbium, erbium, neodymium, thulium or praseodymium are doped. These rare earth ions generate their own typical laser output wavelengths. The fiber materials should be transparent at the wavelength of operation. Silica glass is the most widely used material for operation in the visible and near infra-red operation. The zirconium fluoride materials are used for operation at longer wavelengths. The resonator mirrors that form the cavity are of various types. They may be multilayer dielectric mirrors, distributed Bragg reflectors or semiconductor saturable absorber mirrors. Normally, semiconductor diode lasers are used as the pump sources depending on their output wavelengths. Since rare earth ions have different absorption characteristics, the pump source are chosen accordingly.

Fiber optic lasers have the capability of generating high power laser output with superior beam quality compared to solid state lasers, due to their long interaction length in the active medium, thus generating much higher laser gain than in the case of solid state lasers. They also have higher electrical efficiency, lower laser threshold, lower maintenance problems and are highly reliable, compact and sturdy.

Fiber lasers involve optical fibers as the gain media. In general, the gain medium is a fiber doped with rare earth ions such as erbium, (Er3+), neodymium (Nd3+), ytterbium (Yb3+), thulium (Tm3+), or praseodymium (Pr3+). These lasers fall under the category of solid-state lasers because the gain medium of fiber lasers is similar to those of solid-state lasers. After the invention of laser in 1960, the application of a glass fiber with a cladding was proposed in 1966, and by 1970, fibers with losses of approximately 20 decibels per kilometer were demonstrated. Since then, progress in the fiber technology - both its production as well as applications, has been tremendous. Fibers with losses of less than 0.2 dB/km have been demonstrated long back in 1979. However, the wave guiding effect and the small effective mode area usually lead to substantially different properties of the lasers based on fibers. These lasers have the advantages of being cost effective, compact and portable and are easy to operate with a stable output.

Before we discuss fiber lasers, we define some of the terms widely used for this class of lasers.


  • There are six transmission bands for fiber optic communication. These are O Band (1.26 to 1.31 micron), E band (1.36 to 1.46 micron), S band (1.46 to 1.53 micron), C Band (1.53 to 1.565 micron), L band (1.565 to 1.625micron) and U Band (1.625 to 1.675 micron). Seventh band which is also being used by private networks is around 850 nm.
    Single Mode and Multimode fibers
    Single Mode and Multimode fibers
  • Single Mode and Multimode: In multimode fibers the core diameter is greater than the core diameter of single-mode fibers, making the light to have several propagation modes, i.e. the light goes through the fiber core using several paths and not using a single path, like in single-mode fibers. Multimode fibers have a core diameter 50 to 100 microns (typical commercial values are 50, 62.5 and 100 microns) and a cladding diameter of 125 microns. Multimode fibers can be classified into graded-index and step-index, depending on the refractive index between the core and the cladding; in graded-index there is a gradual change between the core and the cladding, while in step-index this change is abrupt, hence the name. Step-index fibers can transmit data up to 50 Mbps, while graded index fibers can transmit data up to 1 Gbps. Multi-Mode fibers are also known as MMF and they are used in short-distance cables.
  • Step index fibers result in limited distances because of modal dispersion. Since the modes extend through the fiber at different angles, their lengths are slightly different. The result is that light takes less time to travel down some modes (the shorter ones) than others dispersing or spreading down the light pulse. With short fiber spans, there is a little spreading out of the pulse but on longer distances typically more than a kilometer, the spread is so much that it is of not much use.
  • A graded index fiber on the other hand virtually eliminates modal dispersion by gradually decreasing the refractive index out towards the cladding, where the modes are longest. Then waves on longer modes travel faster than on the shorter modes, so the entire pulse reaches at the receiver at almost the same time. GI fibers, one can extract a bandwidth of 200MHz over a distance of 2km. Typically a step index fiber would have almost one tenth of its performance. Beyond a distance of 2km, even GI fibers not only requires high laser power source, but also modal problems creep in making the noise to signal ratio very poor.
  • Single-mode: single-mode fibers are used in long-distance cables, but they require connectors with better precision and as such are expensive devices. In this kind of fiber the light has only one way of traveling inside the fiber core, hence its name. The core diameter is between 7 and 10 microns and its cladding diameter is around 125 microns, so both multi-mode and mono-mode cables have the same diameter, what makes the difference is the diameter of the core. There are three types of single-mode fibers: non dispersion-shifted fiber (NDSF), dispersion-shifted fiber (DSF) and non-zero-dispersion-shifted fibers (NZ-DSF).
  • In case of single mode fibers, many multimode problems such as modal noise, and modal dispersion are no longer an issue. However the tolerances are very tight and thus the cost is higher. Non Dispersion Shifted Fibers (NDSF) carries signal in the O band transmission window at 1.310 micron. Although the range is greatly increased but there is other problem namely chromatic dispersion. Any light pulse, howsoever the precise laser may be, contains a large number of frequencies. Since the refractive index is frequency dependent, the waves end up traveling at different velocities resulting in pulse dispersion. At the same time wave-guide dispersion also affects the wave velocity. As a result, part of electric field and magnetic field extend into cladding resulting in making the wave faster. At 1.310 micron, chromatic dispersion and wave-guide dispersion cancel each other. However outside this wavelength, dispersion increases thus restricting the length of fiber.
  • S band (1.46 to 1.53 micron) and C Band (1.53 to 1.565 micron) transmission windows are better suited for longer distances and they have less attenuation and works well with optical amplifiers such as Erbium doped fiber amplifiers. Dispersion shifted fibers (DSF) move optimal dispersion points to higher frequencies by altering the core cladding interface. Zero dispersion shifted fibers (ZDSF) moves the zero dispersion frequency from 1.310 micron by increasing the wave-guide dispersion until it cancels out the chromatic aberration at 1.55 micron. However, devices like Wavelength Division Multiplexing (WDM), Dense Wavelength Division Multiplexing (DWDM) and Erbium doped fiber amplifiers operate in this range. Signals traveling over ZDSF can combine to create additional signals that may be amplified by Erbium doped fiber amplifiers and superimposed on DWDM channels resulting in noise like four wave mixing.
  • Non-zero dispersion fibers avoids four wave mixing by moving the zero dispersion point above the range of Erbium doped fiber amplifiers. The signal still transmits in S and C band with only a moderate amount of dispersion. The operation in this manner is rather useful as it provides a minimal level of interference needed to separate DWDM channels from one another. Therefore non-zero dispersion fibers are being widely used. The idea is to move λ0 to either end of the 1550 nm band, thus ensuring that all of the wavelength channels have slightly different optical speeds in the fiber. Common wavelengths are around 1.530 micron, 1.497 micron, 1.452 micron and 1.560 micron. The advantage that these fibers have over DSF is a compromise solution of a slightly lower degree of integrated dispersion compensation for a higher tolerance to non-linear distortion effects. These are either with positive dispersion Non-zero dispersion fibers and negative non-zero dispersion fibers.
  • Some long-haul fiber paths usually alternate between positive non-zero optical regime and negative non-zero optical regimes to provide self-dispersion compensation with uniformly low dispersion across the minimum-loss window at 1550 nm.

Numerical Aperture (NA)

The Numerical Aperture (NA) of a fiber is defined as the sine of the largest angle an incident ray can have for total internal reflectance in the core. Rays launched outside the angle specified by a fiber's NA will excite radiation modes of the fiber. A higher core index, with respect to the cladding, means larger NA. However, increasing NA causes higher scattering loss from greater concentrations of dopant.

Qualitatively, NA is a measure of the light gathering ability of a fiber. It also indicates how easy it is to couple light into a fiber.

Numerical Aperture of a fiber
where α is the half acceptance angle.

Cutoff Wavelength

The cutoff wavelength, λc, is the minimum wavelength in which a particular fiber still acts as a single mode fiber. Below the cutoff wavelength, higher order modes are able to propagate which means that the fiber becomes a multimode fiber at this wavelength. It is given as

Cutoff wavelength

where a is the fiber core radius, n1 = ncore and n2 = nclad are the refractive indices of core and cladding, Vc is the cutoff number.

V number is a normalized frequency parameter of a fiber and can be used to express many fiber parameters such as number of modes at a given wavelength, mode cut off conditions, propagation constants etc. The V number is a dimensionless parameter, which is often used in the context of step index fiber. It is defined as

V number

The V number can be interpreted as a kind of normalized optical frequency. It is relevant for various essential properties of a fiber:

If V is less than 2.405 then the fiber is a single mode fiber but if V is greater than 2.405 then fiber is multimode.


  • For V values below ≈2.405, a fiber supports only one mode per polarization direction.
  • Multimode fibers can have much higher V numbers. For large values, the number of supported modes of a fiber can be calculated approximately.
  • V number is related with the number of modes is the fiber as:
           M≈ V2/2 for step index fiber and
           M≈ V2/4 is the number of modes for graded index fiber.
  • V number determines the fraction of the optical power in a certain mode, which is confined, to the fiber core. For single-mode fibers, that fraction is low for low V values (e.g. below 1), and reaches ≈90% near the single-mode cut-off at V ≈ 2.405.
  • A low V number makes a fiber sensitive to micro-bend losses and to absorption losses in the cladding. However, a high V number may increase scattering losses in the core or at the core-cladding interface.
Structure of fiber laser
Structure of fiber laser

Fiber laser is similar with other lasers. The typical structure of fiber laser consists of three parts: pumping source, the gain medium and resonant cavity. Pumping source commonly uses high-power semiconductor laser, the gain medium is fiber core doped with rare earth ions. The resonant cavity is mainly constituted by a fiber Bragg grating and other optical feedback elements. The optical fiber is placed between two selective reflector mirrors. The pumping light enters the optical fiber via coupling optical system, and the laser output is taken through the collimation optical system and the filter. Its structure is shown in the adjoining figure. When the pumping light is incident onto the fiber core, which is doped with rare earth ions, it is absorbed by these rare earth ions. As a result of absorption of photons, there occurs energy level excitation and the process of laser emission starts involving the processes of spontaneous and stimulated emissions.

One of the most common types of fiber lasers is the Erbium-Doped Fiber Laser. The reason for using Erbium is because the Erbium atoms have very useful energy levels. There is an energy level that can absorb photons at a wavelength of 980nm, and this then decays to a meta-stable state equivalent to 1550nm. This means that we can use a cheap diode laser 'pump source' at 980nm and we get a very high quality, and potentially very high power beam out at 1550nm.

Double Clad Fiber Geometry
Double Clad Fiber Geometry

Within the doped fiber, we have our 'laser medium', which are the erbium atoms. The photons that are emitted are confined inside the fiber core. To create the laser cavity, Bragg Gratings are added.

A Bragg Grating is a section of glass that has stripes in it where the refractive index has been changed. Each time the light goes across a boundary between one refractive index and another, a part of it is reflected back. If one has enough stripes, the grating acts like a very efficient mirror.

The pump source can be a cheap diode laser. Diode lasers can also be stacked, in order to get higher power from a number of diode lasers, which can be used to pump a single fiber laser.

The problem is that the fiber core is too small for us to focus the low-quality diode laser into it. To get around this, one has to focus the pump laser into the much-larger cladding around the core. To contain the pump laser beam, the fiber is given an outer sheath as a cladding. This way, the pump beam bounces around inside the fiber. Every time it crosses the core, a bit more pump light is absorbed.

Double Clad Fibers
Double Clad Fibers

Fiber lasers based on single mode fiber can generate a high quality laser beam that is almost diffraction limited, but it requires that the pump sources should also have a diffraction limited beam quality thus restricting the lasers to be generally a low power lasers. Multimode fibers on the other hand usually lead to poor beam quality. In order to overcome this problem, recently double clad fibers (see adjoining figure) are being widely used. In these types of fibers, the pumping laser light partly travels in the single mode core but the major portion of pumping light is restricted to the inner cladding. The inner cladding has a relatively larger area as compared to that of the core and also has a much higher numerical aperture, so that it can support a large number of propagation modes thereby allowing the efficient launch of the output of high power laser diodes. It may be mentioned that the ratio of the areas of inner cladding and core is an important parameter. This area ratio should not be too large, as it would reduce the pump intensity in the core thereby compromising the power efficiency. Area ratios usually used are of the order of 100-1000 are common. In addition to inner cladding, which is usually made of silica, there is outer cladding, which is of lower refractive index as compared to that of inner one. Typically inner cladding is of silica and outer cladding is of fluorine-doped silica or a polymer based cladding.

The simplest design of inner cladding is that of a circular pump cladding and a centered core, which is relatively easy to make and use, but in this kind of design, the propagation modes in the inner cladding are related to helical rays, which have poor overlap with the core. As a result pump light absorption is incomplete. Modes with poor core overlap can be avoided by using a modified design with a lower symmetry. Examples of such geometries include elliptical, D-shaped or rectangular inner cladding. Further, such pump claddings also often better matching the properties of pump sources such as beam shaped diode bars.

Pump Coupling techniques in Fiber Lasers

End-pumped coupling technique
End-pumped coupling technique

A fiber laser can be end pumped or side pumped. In end pumping, the light from one or many pump lasers is fired into the end of the fiber. In side pumping, pump light is coupled into the side of the fiber; actually, it is fed into a coupler that couples it into the inner core.

End-pumped coupling technique

End-coupling scheme for fiber lasers
End-coupling scheme for fiber lasers

From the end-pumped coupling schematic diagrams shown in figure above, it can be seen that, in end-pumping it is used the LD laser source to produce the pumping light, generally through the battery of focusing lenses, then coupling the laser beam into the DCF. End pumping technique as shown in the above figure is the simplest, easy to operate and the coupling efficient is over 50%. When the requirement of the coupling efficient is not high, end pumping is a very good choice. However, if one requires large number of pumping laser diodes for high power applications, the overall pumping wastage is very large as the double clad fiber is to be connected with other fibers in that case. In that case side-pumping technique is more useful: it can be either V-groove type or the micro prism type.

V type groove pump coupling is realized by stripping a bit of DCF outer cladding and then making a series of V type grooves in the inner cladding. The laser produced by LD pump source is focused through the "V" groove, and then by changing the direction in the base is carried out to couple in the double clad fiber. It may be pointed out that that the groove depth must be lower than the half of the inner cladding depth. It is a relatively an expensive method.

In case of coupling through the micro-prism, a micro-prism is inserted into the fiber by stripping the outer cladding of DCF. The LD pump source produces the laser beam, which is incident on the micro prism. After reflection through the micro prism, the pump light is coupled to the fiber. Compared with the V-type groove technique, coupling of pump light through the micro-prism helps in the fiber laser producing the smooth radius beam pump source. The process is however complex and it is difficult to realize.


  • The Fiber laser consists of a coil of appropriate double-clad doped fiber, two reflectors and a pump source. The pump source can be a single diode laser or a fiber laser or an array of single-emitter diodes or a stack of diode bars.
  • Multi-Diode pumped fiber laser
    Multi-Diode pumped fiber laser
    Continuous-wave fiber lasers can be either single- or multimode (in terms of transverse modes). A single mode produces a high-quality beam for materials working or sending a beam through the atmosphere, while multimode industrial lasers can generate higher raw power. If an application does not require the extremely high intensities resulting from single mode operation, the higher total power from multimode operation is often an advantage—for example, for some kinds of cutting and welding, and particularly for heat-treating, where a large area is illuminated.
  • The first fiber lasers with mW output power were realized in the early sixties. For the next few decades, the innovative concept of fiber lasers was only a low-power laboratory curiosity. In more recent years, the output power of CW fiber lasers has increased nearly exponentially up to tens of kW-level and with excellent beam quality.
  • Currently, the available pump power, with adequate good beam quality, limits the power scaling rather then the fiber technology itself.
  • High-power, multimode single-emitter diodes or diode bars pump the modern fiber laser, typically through a cladding surrounding a single-mode core. This single-mode core is typically 5 to 12 μm in diameter. The double-clad fiber consists of an inner single-mode core doped with the appropriate rare-earth ions such as neodymium, erbium, ytterbium and thulium. The cladding is made of undoped glass that has a lower index of refraction. The pump light is injected into the cladding and propagates along the structure, passing through the active core and producing a population inversion. Though single emitter laser diodes are costly, better beam quality (half spot width multiplied by the half the divergence angle) and much longer lifetime make single emitter laser diodes a unique choice of future as a pumping laser source. There are advantages for single-emitter pump diodes. The main advantage is that they do not require water-cooling and can be introduced to the active medium via fiber at very high efficiency, with no additional bulk optics or alignment required. Also, the single-emitter diode can produce higher output power and has better beam properties and lifetimes of greater than 200,000 hours of operation, both in CW and modulated regimes.
  • The emission wavelength is a function of choices in the doped fiber and by any type of reflector (a typical example would be Bragg gratings).
  • Single-mode fiber lasers are available in the commercial market from a few watts to few thousand of watts of output. In addition, single-mode fiber lasers have been produced to more than 20-kW level for special projects employing a more expensive fiber technology. These devices are typically continuous in operation; however, the units can be modulated to 50 kHz or more. In the modulated mode, the units have a peak equal to the average CW power. The emission exits via a single-mode fiber with an M2 less than 1.1. Laser transverse mode is a pure Gaussian distribution.
  • As the profile is a function of the single-mode fiber rather than thermal operating point, as is the case with conventional solid-state lasers, a fiber laser produces the same beam profile over the entire operating range. The modulation is accomplished by turning the pump diodes off and on, allowing the device to be modulated at a high frequency or in single-pulse operation. Contrary to the conventional solid-state laser, the fiber laser, with its perfect cross section, does not require a warm-up time and can operate in a wide range of ambient conditions in a stable (beam quality and power) manner. These lasers are available with both randomly and linearly polarized outputs and typically can operate from 10 to 100 percent of specified power without any change in divergence or the final focus spot diameter.
  • Multimode kilowatt fiber lasers: Kilowatt-class and above lasers are produced by fiber combining several single-mode fiber lasers in parallel and then launching them through larger-core-diameter step-index fibers produce lasers. At this point, the laser is no longer single-mode; however, the resultant beam quality is better than that of most commercial industrial kilowatt-class lasers.
  • QCW fiber lasers: The newest types of fiber lasers are the QCW. These devices feature a high peak power and a lower average power and can be manufactured at a substantially lower cost than a CW version. For example, a QCW laser with a peak power of 20 kW and an average power of 2 kW is about five times cheaper than a 20-kW CW laser. They are ideally suited for numerous industrial applications requiring a long pulse duration and high peak power, such as spot welding, seam welding and drilling. Designed to displace existing YAG lasers due to their minimal required maintenance and low up-front costs, QCW lasers can be easily retrofitted into most existing systems. Both single-mode and multimode versions have been developed.
  • Q-switched fiber lasers are typically constructed by firing a low-power, nanosecond-pulsed seed laser with an integral pigtailed modulator through a chain of fiber amplifiers. A fiber amplifier, as with a fiber laser, is constructed using the same techniques; however, the lasers do not contain end reflectors that induce laser action. These lasers are completely monolithic and can produce nanosecond pulses with frequencies from 20 to >200 kHz.
  • Wall-plug efficiency or radiant efficiency is the energy conversion efficiency with which the system converts electrical power into optical power. It is also defined as the ratio of the radiant flux (i.e. the total optical output power) to the input electrical power. Overall fiber-laser efficiency is the result of a two-stage process. First is the efficiency of the pump diode. Semiconductor lasers are very efficient, with an electrical to optical efficiency in the range of 50 - 60 %. If this output can be matched carefully to the fiber laser's absorption line, the result is the pump efficiency, which is of the order of 40 - 50%. The second is the optical-to-optical conversion efficiency. Usually high excitation and extraction efficiency can be achieved, producing optical-to-optical conversion efficiency on the order of 60% to 70%. The result is wall-plug efficiency in the 25% to 35% range.

Applications

Industrial fiber lasers are utilized in materials processing applications in all of the major high-power and low-power markets, including automotive welding and cutting, sintering, marking, scribing, drilling and heat treating. The single-mode lasers, with the ability to attain high fluency levels and to be focused to micron-sized spots, have changed previous beliefs relating to process parameters. On the kilowatt level, the fiber laser has attained higher speeds of cutting and weld penetration than conventional technology operating at the same power level. Fiber lasers have some advantages over other lasers for materials processing. For example, the near-IR wavelengths of fiber lasers are absorbed well by metals. The beam can also be delivered by fiber, which allows a robot to easily move the beam focus around for cutting and drilling. The industrial market is now the largest market for fiber lasers; much of the action right now is at the kilowatt-class power level. In case of high-power cutting and welding—for example, replacing resistance welding for high-speed sheet steel, solving the problem of material distortion caused by resistance welding. Power and other feedback controls allow fiber lasers to cut a very precise curve, especially going around corners. Particularly interesting is their use in automotive work. The automotive industry is moving to high-strength steel to produce cars that meet durability requirements but are relatively light for better fuel economy; the problem is how to cut the high-strength steel. And that's where they turn to fiber lasers. It's very difficult, for example, for conventional machine tools to punch holes in this kind of steel; however, fiber lasers (and other types of lasers as well) can easily cut these holes. In addition, fiber lasers now offering up to 100-kW output power have greatly increased the power available at the 1-μm region, up from the previous 5-kW level.


  • Few kW multimode fiber laser is extremely useful for cutting and drilling concrete. It is very important to use this technique particularly in building the structures earthquake proof. For doing so one requires to reinforce steel bar in the concrete to bolster its strength. Conventional percussion drilling can crack and weaken concrete, but fiber lasers cut it without fracturing.
  • Q-switched fiber lasers are used, for example, in LIDAR. Typically it contains an eye-safe erbium fiber laser with a 4 kW peak power, a 50 kHz repetition rate, and 5-to-15-ns pulse duration. These lasers are also used for pulsed materials processing.
  • There is a lot of interest in smaller fiber lasers for micro-and nanoscale machining. For surface ablation, if the pulse duration is made shorter than few tens of ps, then there is no material splatter, just ablation, eliminating the formation of kerfs and other unwanted artifacts on the metal being cut. Pulses down in the range of femtosecond regime don't heat the surrounding area, allowing material work without damaging or weakening the surrounding area. In addition, holes can be rapidly cut with high aspect ratios. Typically, small holes through 1-mm-thick stainless steel using 800-fs pulses at a 1 MHz repetition rate can be drilled within few milliseconds.
  • Because of their compact size, wavelength choice and single-mode operation, fiber lasers offer the medical community a tool for an array of medical applications that rely on specific wavelengths and fiber delivery. Maintenance-free operation makes it very attractive for doctors and others in the medical profession. One can also do surface machining of transparent materials - for example, the human eye. To cut flaps for LASIK surgery, femtosecond pulses are tightly focused with a high-numerical-aperture lens onto a spot below the eye's surface, causing no damage at the surface, but breakdown of the eye material at a controlled depth. The surface of the cornea, the smooth surface of which is important for vision, escapes unharmed. The flap, separated from underneath, can then be pulled up for ablative excimer-laser lens shaping. Other medical applications for picosecond and femtosecond fiber lasers include shallow-penetration surgery in dermatology, and use in certain kinds of optical coherence tomography (OCT).
  • For the scientific and government communities, fiber lasers' wide wavelength range, availability of narrow linewidths, polarized or unpolarized emissions, short pulse durations, single-mode operation, insensitivity to environmental conditions and compact size are an ideal solution for many sophisticated applications, including some that only fiber lasers can accomplish. Scientific applications of femtosecond fiber lasers include laser-induced breakdown spectroscopy, time-resolved fluorescence spectroscopy, and general materials research.
  • Fiber lasers can satisfy extreme power requirements. The U.S. Navy's Laser Weapon System (LaWS), tested by the Naval Sea System Command, had many fiber lasers, incoherently combined into one beam and fired through a beam director. The 33 kW systems were used to shoot down an unmanned aerial vehicle (UAV). Although the beam was not single-transverse-mode, the system is of interest because it can be constructed of standard, easily available components.
  • Lockheed Martin has been investing internal research and development funding to mature the technologies and integrated system operation. It started by developing a ground-based 10-kilowatt prototype using a commercial laser, a program called Area Defense Anti-Munitions, or ADAM. In 2012 and 2013, ADAM shot down 19 small-caliber Quassam-like rockets and an unmanned aerial system in flight. The system could also disable two Zodiak-type boats.
  • In 2014, scaling up power by integrating Lockheed Martin 30-kilowatt Accelerated Laser Demonstration Initiative (ALADIN) fiber laser with the ADAM (Area Defense Anti-Munitions) architecture and beam control to create the Advanced Test High Energy Asset, or ATHENA, prototype was demonstrated. In the early testing, ATHENA stopped the engine of a truck, and in 2015 it helped in defeating four quad copter drone targets.
  • In 2015, Lockheed Martin began production of at 60-kilowatt system for a U.S. Army vehicle - the first laser built using a modular technique. The US Army has the option to add more modules and increase power from 60 kilowatts to 120 kilowatts or even more as a result of the laser's modularity.
  • The primary difference is achieving higher power in fiber lasers in ATHENA and U.S. Navy's Laser Weapon System (LaWS) is that Lockheed Martin approach uses spectral beam combination to increase the power of the laser to send one beam, whereas discrete laser devices incoherently combine on the target in case of U.S. Navy's Laser Weapon System (LaWS).
  • Laser Beam Combining: There is a requirement to combine a number of separate laser beams into a single beam. Most commonly, the need is to provide a high power levels in industrial lasers and particularly in laser directed energy weapons has led to an interest in scalable systems in which an arbitrary number of otherwise identical laser beams can be added together to realise overall power levels in the 10s to 100s of kW. In the most elemental form of beam combining we might envisage using some form of semi-transparent mirror (beam splitter), which transmits one beam and reflects the other. Unfortunately, this simplistic technique doesn't work efficiently because we always lose half the total power that is available. For example, a 50:50 beam splitter would lose 50% from each beam and any other splitting ratio would lose power from the beams proportionally e.g. a 70:30 beam splitter would lose 70% of one beam and 30% of the other. Perhaps the most commonly encountered beam combining element is the well-known dichroic mirror, which transmits one wavelength or range of wavelengths whilst reflecting others. This has been used effectively for many years and enables beams of different wavelengths to be combined with high efficiency. If the beams are polarized, as with many laser systems, this property can be exploited to provide an efficient combination method using a polarizing beam splitter. Reflective polarizers reflect s-polarized light whilst transmitting the p-polarizations. More sophisticated and recent techniques are discussed below:
  • The simplest approach is incoherent beam combining, which directs many laser beams in the same direction, increasing total power but not increasing the beam brightness. More sophisticated techniques can raise brightness. One is combining beams coherently, so their amplitudes add constructively. An alternative is combining beams of different wavelengths, as in wavelength-division multiplexing, which avoids the complexities of phase matching but produces wider-band laser emission.
  • For two incoherent sources of intensity I1 and I2, the resultant intensity is
    Intensity of incoherent sources
    For I1 = I2 = I0, then Iresultant = 2I0

    On the other hand, for coherent waves, however, there will be redistribution of intensities, and we will have
    Intensity of coherent sources max
    and
    Intensity of coherent sources min
    For I1 = I2 = I0, then
    Intensity of incoherent sources recalculated
    Thus for incoherent sources, the amplitudes are squared first and then added to get the resultant intensity, whereas for coherent sources, the amplitudes are added first and then squared to get the maximum intensity.
  • Incoherent combination of the high-quality beams from fiber lasers can generate much more directional beams that are getting serious consideration for use as laser weapons. "Beam brightness at the source is of limited importance when considering realistic [directed-energy] propagation scenarios in turbulent atmospheres," Fiber laser arrays are living up to that promise. The US Navy assembled an array of six 5.5-kW industrial fiber lasers to produce the 33 kW Laser Weapon System (LaWS), which in 2010 engaged representative targets over the water at a distance of one nautical mile. Encouraged by those tests, the Navy is scaling LaWS to the 100 kW range.
    Incoherent beam combining technique
    Incoherent beam combining technique
  • Laser beam combining techniques allow increasing the power of lasers far beyond what it is possible to obtain from a single conventional laser. One step further, coherent beam combining (CBC) also helps to maintain the very unique properties of the laser emission with respect to its spectral and spatial properties. Such lasers are of major interest for many applications, including industrial, environmental, defense, and scientific applications. Recently, significant progress has been made in coherent beam combining lasers, with a total output power of 100 kW already achieved. Scaling analysis indicates that further increase of output power with excellent beam quality is feasible by using existing state-of-the-art lasers.
  • Combining laser beams can boost power at the target far above that produced by a single laser. Incoherent beam combining achieves propagation efficiencies of greater than 90%, while avoiding the complexities of coherent or spectral beam combining. Incoherent combining of laser beams is achieved by overlapping the individual laser beams on a target with a beam director consisting of independently controlled steering mirrors with optional adaptive-optics capabilities. This approach does not require phase locking or polarization locking of the individual lasers, and can be readily scaled up to a compact and reliable directed-energy system. In an example that illustrates the essence of incoherent beam combining, the beams from a hexagonal array of seven fiber lasers are combined with a beam director of individually controlled steering mirrors (see above figure). The individual fiber lasers have an initial spot size large enough so that diffractive spreading is not significant over the propagation range. For example, a straightforward calculation involving only diffraction shows that a Gaussian beam with a 4 cm spot size that is focused onto a target at a range of 5 km will have a spot size of only 4 cm on the target. Typically, atmospheric turbulence will cause more beam spreading than diffraction.
  • The coherent beam combining techniques can be subdivided into side-by-side combining (tiled aperture) techniques, leading to a larger beam size but reduced divergence and filled-aperture techniques, where several beams are combined into a single beam with the same beam size and divergence.
  • Techniques for side-by-side combining may have been inspired by the earlier Implementation of phased-array antennas in radio frequency and microwave transmitters and receivers. In the optical domain, the realization is more difficult due to the much smaller wavelength, which introduces correspondingly tighter mechanical tolerances. A simple example of side-by-side combining, four beams may be arranged to obtain a single beam as shown in the figure. In actual practice, there may be very small gaps of low intensity between the beams. If the beams are all mutually coherent, and the relative phases are properly adjusted to obtain essentially plane wavefronts over the whole cross-section, the resulting beam has a much lower beam divergence as compared to single beams. As a result, the beam quality is preserved, and the brightness can be four times that of the single beams. Fiber laser beams can be combined coherently by tiling output of an array of phase locked lasers (or amplifiers) across an aperture, or arranging transform optics and a diffractive optical element to fill an aperture with light from a more widely spread array of fiber emitters.
    Side-by-side combining of laser beam
    Side-by-side combining of laser beam
  • To understand the principle of filled-aperture techniques, let us consider a case of a 50% reflectivity beam reflectivity. Overlapping two input beams at this beam splitter will in general lead to two outputs, but a single output can be obtained if the two beams are mutually coherent and adjusted such that there is destructive interference for one of the outputs. In other words, if two mutually coherent laser beams are incident on a 50/50 beam splitter, both beams can be combined into one output port, as long as they are locked in phase. The resultant combined beam has the same quality as each input, but with double the power and brightness. By cascading multiple two-port combiners, or by using beam splitters with more than two ports, this concept can be generalized to more than two lasers.
    Combining mutually coherent laser beams
    Combining mutually coherent laser beams
  • In any case, mutual coherence of the combined beams is essential. Apart from phase coherence, the beams involved must have a stable linear polarization, and the amplitude fluctuations should also not be excessive.
  • Recently, a robust and compact design has been demonstrated for coherent beam combining, which can handle kilowatt class of fiber lasers. Adjoining figure shows such architecture. A master oscillator is split to seed N fiber channels, which each consist of actuators for phase, polarization, and path length, followed by a kilowatt-class fiber amplifier chain. The tight confinement of light within these narrow fiber cores imposes nonlinear limits on the combinable fiber powers. To launch light into free space, the kilowatt fiber tips were mounted into a close-packed, thermally stable array. A spherical mirror simultaneously collimates and images the array onto a reflective diffractive optical element (DOE). This diffraction grating is specially ruled such that laser beams falling at different angles emerges parallel. In this case, the DOE—which has essentially the same power handling as a typical high-energy laser mirror—is a multi-port beam splitter with a lithographically etched, periodic-surface-relief, low-angle structure. When illuminated by N phase-locked input beams at angles matching the DOE diffractive orders, the DOE can function as a beam combiner. To enable servo locking of the fiber phases, polarizations, and path lengths, single beam sample of the combined output is sent to path, polarization and phase controller. The use of coherent detection enables scaling of this controller to any number of fiber-laser channels with high control bandwidths, allowing for the rejection of vibrational disturbances that may be coupled from the environment or platform of operation. The beam combining efficiency for a 3 kW of input laser has been achieved more than 80% with a near diffraction limited beam quality. Beam quality factor M2 was of the order of 1.2.
    Coherent beam combining of high-powered lasers
    Coherent beam combining of high-powered lasers
  • Though the coherent combining of fiber lasers has been demonstrated for 3kW laser power, however the concept can be extended to an efficient scaling path toward 100kW-class lasers with diffraction-limited beam quality.
  • Commercially available High Power CW Fiber Lasers cover output power range from 1 kW to over 100 kW and feature a wide range of operating wavelengths, single-mode and multi-mode options, high stability and extremely long pump diode lifetime. These lasers are water-cooled and can be supplied with a built-in or standalone chiller. The lasers are available with a wide variety of fiber terminations, collimation optics and processing heads. IPG Photonics is one of the manufacturers.
  • IPG manufactures high power CW Ytterbium lasers in 1 to >100 kW range and Erbium, Thulium and Raman fiber lasers in 1 to 5 kW range.
  • After successfully testing Laser Weapon System (LaWS), which was deployed in 2014 on the amphibious transport dock USS Ponce, the US Navy plans to fire a 150-kw weapon system in 2018. And further plans to enhance the laser power to something like 250 kW or 500 kW by 2020.

References