Semiconductor Lasers

The History of Lasers

The Physics of Quantum Mechanics

The basic ideas of quantum theory were first introduced by Max Planck when investigating patterns of electromagnetic radiation emitted by a hot body. Einstein’s theory of special relativity helped to resolve some the problems that classical physics had in describing particles moving at speeds comparable to that of light in his explanation of the photoelectric effect. As physicists sought new ways to solve other phenomena not explained within the framework of classical physics, Erwin Schrödinger and Werner Heisenberg formulated the new and consistent physical theory known as quantum mechanics. This new theory forced scientists to reexamine the relationship between particles and waves.

The New Atom

As scientists slowly uncovered more of the properties of the atom, it became apparent that electrons could only contain distinct amounts of energy. The basic model was modified to give electrons specific orbits in which to exist that extend from the nucleus outward. Electrons can only exist in these orbits and not between them. Each level holds a fixed number of electrons with an associated and specific energy. Once a level is full, or if an electron gains enough energy, it will jump to the next energy level. Any level that is not full is called an incomplete level. These incomplete levels largely determine how atoms bond and interact.

The most successful model of atomic structure is the quantum mechanical nuclear model. This model does not give us a view of the atom that is intuitive, yet it explains many phenomena that the simple model cannot. Since the electron acts like a wave of energy and a particle, the Bohr model proved to be unreliable in making predictions. The quantum model helps to explain this wave-particle duality of the electron as well as the electron's discrete energy behavior. The distinct levels of energy available to an electron are described as being quantized.

The Crystal Lattice

When individual atoms are brought close together, remarkable things begin to happen. First, the quantized levels of the atom overlap and seem to blur into near-continuous levels of energy. Second, the atoms align themselves into crystals, a repeated atomic structure in three dimensions. These crystals are referred to as lattices. There are several forms of crystal lattices. The first and most basic structure is called the simple cubic (see figure 2).

The unit cell, consisting of the smallest piece that retains the properties of that crystal, is a cube with an atom on each corner. The next crystal lattice is the body-centered cubic. The shape is that of a cube with an atom in the center (see figure 3).

The third form is the face-centered cubic (see figure 4). This cube has an atom on each corner and an atom on each face of the cube.

These lattices are held together by atomic forces called bonds. Two or more atoms combine to form a molecule due to a net attractive force that exists between them. Bonds are classified into Ionic, Metallic, Van der Walls, Mixed, or Covalent bonds. The kind of resultant bond is dictated by the electrons in the outermost shells of bonded atoms. Covalently bonded solids, such as silicon, have a tetrahedral configuration called a diamond crystal structure (see figure 5).

Although there are atoms at each corner and at the center of each face, there are also four atoms within the unit cell. In some semiconductors, such as GaAs, each type of atom in the lattice structure is arranged in alternating positions. Such geometry is called the zinc blend structure (see figure 6).

Basics of Semiconductor Theory

Materials can be categorized into conductors, semiconductors, or insulators by their ability to conduct electricity. Electrons, at thermal equilibrium with its environment (such as the solid materials in which the electrons exist), are governed by the Fermi statistics for their energy distribution. The so-called Fermi function, f(E), gives the probability with which a quantum state at energy E is occupied by an electron. The most important property is the Fermi energy, EF, which enters f(E) as a key parameter. According to the Fermi statistics, a quantum state can have a maximum of one electron. The Fermi function is expressed as

whose physical meaning is the probability of electron occupancy for an energy state at energy E. These allowed energy levels tend to form bands. The highest filled level is known as the valence band. Electrons in the valence band do not participate in the conduction process. The first unfilled level above the valence band is known as the conduction band. In metals, there is no energy gap; the conduction band and the valence band overlap, allowing free electrons to participate in the conduction process. Insulators have an energy gap that is far greater than the thermal energy of the electron, while semiconductor materials, the energy gap is typically around 1eV. The diagram below summarizes the energy band model of solid materials (see figure 7).

Types of Semiconductors

Elemental semiconductors are semiconductors where each atom is of the same type such as germanium (Ge) and silicon (Si). These atoms are bound together by covalent bonds, so that each atom shares an electron with its nearest neighbor, forming strong bonds. Compound semiconductors are made of two or more elements. Common examples are GaAs or InP. These compound semiconductors belong to the III-V semiconductors, so called because the first and second elements can be found in group-III and group-V of the periodic table respectively.

Intrinsic semiconductors are essentially pure semiconductor material. An extrinsic semiconductor is formed from an intrinsic semiconductor by adding impurity atoms to the crystal in a process known as doping. To take the simplest example, consider Silicon. Since Silicon belongs to group IV of the periodic table, it has for valence electrons. In the crystal form, each atom shares an electron with a neighboring atom. In this state, it is an intrinsic semiconductor. B, Al, In, Ga all have three electrons in their valence band. When a small proportion of these atoms are incorporated into the crystal, the dopant atom has an insufficient number of bonds to share with the surrounding Silicon atoms. One of the Silicon atoms has a vacancy for an electron. It creates a hole that contributes to the conduction process at all temperatures. Dopents that create holes in this manner are known as acceptors. This type of extrinsic semiconductor is known as p-type because it creates positive-charged carriers. Elements that belong to group V of the periodic table such as As, P, Sb have five electrons in their valence band. When added as a dopant to intrinsic Silicon, the dopant atom contributes an additional electron to the crystal. Dopants that add electrons to the crystal are known as donors and the semiconductor material is said to be n-type. Figure 8 is a graphical representation of such concepts.

Doping of compound semiconductors is slightly more complicated. The effect of the dopant atom depends upon the site occupied by the atom. In III-V semiconductors, atoms from group II act as acceptors when occupying site of the group III atom, while atoms in VI act as donors when they replace atoms from group V. Dopant atoms from group IV have the unique property that allows them to act as acceptors or donors depending on whether they occupy the site of group III or group V atoms respectively. Such impurities are known as amphoteric donors.

Laser Theory

Laser technology deals with the concentration of light into very small, powerful beams. All lasers must have three basic elements: a lasing medium, an energy source, and an optical resonator. The concept of atomic energy levels and the Bohr’s model of the atom explain spontaneous spectral absorption and emissions lines in terms of quantum jumps between energy levels. Nonetheless, the basic operation of a laser depends on a principle formulated by Albert Einstein in 1917 regarding the transitions between energy levels in an atom, or molecule.
Einstein found that if a photon with an energy equal to an allowed transition strikes an atom in an excited state, it might stimulate an electron to make a transition to a lower energy level, with the emission of another photon. It is significant that when stimulated to de-excite, the emitted photon will be inphase and in the same direction as the stimulating photon. In addition, the photon released by the atom as it is de-excited will have a total energy exactly equal to the difference between the excited and lower energy levels as shown in figure 9.

The difference between energy levels is calculated using Planck’s constant times the speed of light, divided by the wavelength.

The wavelength of the photon energy can be found by rearranging equation 2 to solve for wavelength (l).

A point to consider is that stimulated emission is an amplification process—one photon in, two out—as illustrated by figure 10. The many photons produced by stimulated emissions are the source of the intense, coherent light in a laser.

Only certain materials serve well as lasing media. Stimulated emissions alone will not produce significant amplification of light unless a condition called population inversion occurs. Most electrons reside in the lowest energy levels, and the higher energy levels are depopulated. This distribution can be changed by the addition of excess energy in precise and discrete amounts.
The addition of energy to a laser that causes population inversion is called pumping. A common method of pumping used by laser designers is optical pumping. In this method, a light source illuminates the lasing medium. Electrons in the lasing medium absorb incident photons at the correct energy causing them to jump to a higher energy level. These excited electrons are now ready to give off a photon by stimulated emission. To obtain light amplification, however, the emitted photons must be confined in the system long enough to allow them to stimulate further emissions. This is accomplished by using a resonant cavity. This is essentially a rectangular cavity with highly polished surfaces at the ends as figure 11 exemplifies.

Types of Lasers

The light produced by lasers is in general far more monochromatic, directional, powerful, and coherent than that from any other light sources. Nevertheless, the individual kinds of lasers differ greatly in these properties as well as in wavelength, size, and efficiency. There is no single laser suitable for all purposes, but some combinations of laser properties are capable of accomplishing things that were impracticable before its invention. The laser has found numerous practical uses, ranging from delicate surgery to measuring the distance between the Earth and the Moon.
Of the several different types of lasers produced today, the following are the most significant.

Gas-discharge lasers

Gas lasers use a mixture of helium and neon enclosed in a glass tube. A flow of coherent (one frequency) light waves is emitted through the output coupler when an electric current is discharged into the gas. If mirrors are positioned at the ends of the discharge tube, laser action results. Though the conditions are unusual and occur for only a few of the many wavelengths at which the discharge emits, most gases can be made to exhibit laser action at some wavelength under certain discharge conditions. The continuous light-wave output is monochromatic.

Gas-dynamic lasers

If a hot gas is allowed to cool rapidly, the number of molecules in a low-energy state may decrease more rapidly and fall below the number in a higher energy state, thus permitting laser action. With this type of lasers, high power outputs of more than 30,000 watts can be obtained.

Chemical lasers

Certain chemical reactions produce enough high-energy atoms to permit laser action to take place. For example, laser action can occur in carbon dioxide if it is present when the element hydrogen and fluorine are reacting to produce hydrogen fluoride.

Free-electron lasers

Lasers of this type are more efficient than any other variety in producing beams of very high power radiation. Such devices are tunable, so that they can be made to operate at microwave to ultraviolet wavelengths. Theoretically, they have the potential of generating laser radiation of X-ray wavelength, though present technology is still incapable of such short wavelengths.

Semiconductor lasers

Semiconductor lasers are made from semiconductor p-n junctions and are commonly called injection laser diodes (ILDs). The excitation mechanism is a direct-current power supply, which controls the amount of current to the active medium. The output light from an ILD is easily modulated, making it very useful in many electronic communications applications.

The Semiconductor Laser

The main reasons behind the role played by semiconductor lasers are their continued performance. Improvements, especially in low-threshold current, high-speed direct current modulation, ultra-short optical pulse generation, narrow spectral linewidth, broad linewidth range, high optical output power, low cost, low electrical power consumption, and high wall plug efficiency are some of the reasons behind their popularity.
In the course of semiconductor laser development, the simple p-n junction has been discarded and replaced by a heterostructure containing several semiconductor layers of different compositions. Heterostructures were first introduced in 1969 when they had an immediate impact on device performance. Many of these achievements were based on joint progress in material growth technologies and theoretical understanding of a new generation of semiconductor lasers – the quantum well (QW) lasers.

Basic Principle of Operation

A semiconductor laser is a p-i-n diode. When it is forward-biased, electrons in the conduction band and holes in the valence band are injected into the intrinsic region (also called active region) from the n-type doped and the p-type doped regions, respectively. The electrons and the holes accumulate in the active region and are induced to recombine by the lasing optical field present in the same region. The energy released by this process (a photon for each electron-hole recombination) is added coherently to the optical field (laser action).
The first semiconductor lasers where made from heavily doped p-n junctions. Under conditions of forward bias the electrons and holes would recombine at the barrier junction producing some laser emission at high currents. These devices were inefficient and had high threshold currents as the majority carriers tended to drift away from the junction interface. It was soon discovered that more efficient lasers could be produced by the implementation of a heterostructure design.
In conventional bulk semiconductor lasers, as shown in figure 12, a double heterostructure (DH) is usually used to confine the injected carriers and the optical field to the same spatial region, thus enhancing the interaction of the charge carriers with the optical field.

For optical radiation at frequency n to experience gain (amplification) rather than loss in a semiconductor medium, the separation between the Fermi energies of electrons and holes in the medium must exceed the photon energy hn. To reach this state, a certain minimum value of injected carrier density Ntr (transparency carrier density) is required. This transparency carrier density is maintained by a (transparency) current in a semiconductor laser, which is usually the major component of the threshold current and can be written as

where w is the laser diode width and L is the laser cavity length. Jtr is the transparency current density, which can be written as

where e is the fundamental electron charge, d is the active layer thickness, and tc is the carrier lifetime related to spontaneous electron-hole recombination and other carrier loss mechanism at injection carrier density Ntr.
The equation 4 shows that a reduction in the active layer thickness d will conduct to a reduction in the transparency current density, which is usually the major component of the threshold current density.

Practical Applications

Semiconductor lasers have become the technology of choice for many important applications because of their small size, low cost, high reliability, and excellent spectral and modulation characteristics. Lasers play a key role in a wide variety of applications and in such diverse areas as spectroscopy, communications, metrology, and micro-sensors and micro-instruments.

Spectroscopic

Spectroscopic applications typically require single-wavelength operation with some degree of tenability. Highly stable single-longitudinal-mode operation may be achieved by a distributed feedback (DFB) or distributed Bragg reflector (DBR) laser cavity. Both types of devices require the incorporation of a sub-micron lithographically defined grating buried within the laser structure, which is accomplished using state-of-the-art fabrication and epitaxial growth techniques. DFB and DBR lasers have been commercially developed in indium phosphide (InP) material systems at the fiber optic wavelengths of 1.3 and 1.55 microns by a host of companies. However, the availability of semiconductor lasers at the specific wavelengths suitable for spectroscopy applications is extremely limited. State-of-the art single-mode DFB lasers at 1.37 and 1.43 microns have been fabricated at JPL for incorporation into spectroscopy instruments and are currently being used for atmospheric trace-gas measurements in JPL’s aircraft programs.
Single-mode semiconductor lasers that can produce continuous wave (CW) output at wavelengths between 2 and 5 microns at temperatures above 200 Kelvin providing further opportunities to develop much smaller instruments for spectroscopy. JPL, in collaboration with several laboratories and universities, is developing gallium antimonide (GaSb)-based single-mode semiconductor lasers in the 2- to 5-micron-wavelength range that could operate at these temperatures. To obtain single-mode operation in this material system, the laterally coupled DFB design recently developed at the JPL Microdevices Laboratory will be used. The laterally coupled design is best suited to fabricate DFB lasers in material systems for which epitaxial regrowth embedding the grating structure is extremely difficult.

Communications

Communication is a second application of semiconductor lasers. The semiconductor laser’s relatively high output power, high frequency of operation, and capability of carrying extremely wide bandwidth signal makes it ideally suited for high-capacity communication systems. Laser communication systems employ either fiber optic or free-space transmission channels. Fiber optic communications require laser wavelengths matched to the minimum loss or dispersion wavelength of a transmitting fiber. The evolution of such technologies based on low-loss fiber transmission at 1.3 and 1.55 microns has benefited from intensive commercial development. While free-space communications have no inherent wavelength restrictions based on transmission properties, the development of solid-state laser sources has been dominated by the commercial availability of high-quality yttrium aluminum garnet (YAG) lasers at the 1.064 microns range.

Metrology

Another application of semiconductor lasers is metrology. NASA's stellar interferometry missions require laser sources suitable for high-precision interferometric metrology. Interferometric metrology systems are used to form optical trusses for structure definition and stabilization and for measurement of the stellar interferometer baseline. For these applications, the lasers need to be high power, extremely frequency-stable, and have a very narrow linewidth - i.e., long coherence length - so that an interferometric measurement can be performed with 1-10-kilometer path-length differences between the spacecraft comprising the stellar interferometer. In addition to performance, the choice of technology used to implement the metrology laser sources must also take into account the ruggedness, reliability, size, weight, and power consumption requirements associated with space qualification and deployment on future small and inexpensive spacecraft. Though semiconductor lasers are by far the simplest and most reliable type of lasers available, high-power (greater than 100 milliwatts) narrow-linewidth (less than 20 kilohertz) semiconductor lasers have not been available to date for these applications.

Micro-sensors and Micro-instruments

The semiconductor diode laser, as a discrete device or when integrated with optics, electronics, or microsensors, offers unique opportunities in the miniaturization of scientific instruments and sensors. The light from many lasers is relatively powerful and can be focused by a conventional lens system to a small beam of great intensity. Thus, even a moderately small-pulsed laser can vaporize a small amount of any substance and drill narrow holes in the hardest of materials. Ruby lasers, for example, are used to drill holes in diamonds for wire drawing dies and in sapphires for watch bearings. For biological research, a finely focused laser can vaporize parts of a single cell, thus permitting microsurgery of chromosomes.
Strong heating can be produced by a laser at a place where no mechanical contact is possible. Thus, one of the earliest applications of lasers was for surgery on the retina of the eye. Lasers are also used for small-scale cutting and welding. They can trim resistors to exact values by removing material and can alter connections within integrated arrays of microcircuit elements. A pulse of light from a laser can vaporize a sample of a substance for analysis by suitable instruments. By this method, an extremely small sample can be analyzed without introducing contaminants.

Advantages/Disadvantages of Semiconductor Lasers

The Advantages

Semiconductor lasers offer many advantages including small size, lightweight, low power consumption, and high efficiency. Recent research developing surface emitting lasers has shown the ability to produce lasers in 2-dimensional arrays, which can be manufactured on a single silicon wafer. The small size of these lasers has allowed major advances in optical communications. Because ILDs have a more direct radiation patter, it is easier to couple their light into an optical fiber. This reduces the coupling losses and allows smaller fibers to be used.

The Disadvantages

One of the disadvantages of semiconductor lasers is lower power output. Gas and chemical lasers can produce outputs up to 1015 watts. Another disadvantage is high sensitivity to static electricity. Special antistatic environments are necessary when handling these devices. Special power supplies are also needed to protect the laser from transient voltage spikes during turn-off or turn-on. The beam profile of semiconductor lasers has an astigmatism, which must be corrected with cylindrical lenses.
Table 1 summarizes the three most important types of semiconductor laser materials. The variation of energy bandgap and index of refraction can be controlled by changing to composition parameters x and y.

Future Developments

Research into the basic structure and characteristics of semiconductors promise a future filled with new technological products. Although semiconductor lasers already play a key role in information services, optical communication, and in the entertainment industry, the next generation of smaller, more efficient and more dependable devices will revolutionize many electronic devices. By confining electrons in dimensionally limited environments and by using new and exotic semiconducting compounds, researchers near the point of being able to create designer atoms. In addition to this, much effort is being focused on expanding the power and relatively narrow frequency bands presently attainable by semiconducting lasers. There is also research underway that endeavors to expand the operating environments in which semiconductor lasers are able to operate. Through proper design of the laser, environmental conditions such as temperature, vibration and even pressure will cease to be a limiting factor in semiconductor laser applications.

Extended Lifetimes

Many applications that are well suited for semiconductor lasers require that the device be robust and long lasting. To this end, many researchers focused on overcoming the limited life span and fragility of the lasers. A life span of 1,000,000 hours was an important benchmark that justified the economics as well as alleviated installation and repair complications

Advances in material science led to improvements in fabrication techniques. These advances continue to reduce the cost of production and increase the precision of the devices. Current techniques, such as molecular-beam epitaxy (MBE), deposit layers of semiconductor material with a thickness of only a few atoms. Such ultra thin layers promise to produce very efficient laser action while at the same time using less current. The reduced current means less dissipated heat and thus prolongs the useful life span of the laser. Using MBE, continuous operation of 1,000,000 hours was achieved and even surpassed.

Another benefit of the low-current operation of future semiconductor lasers will be near-continuous operation without expensive power supplies and higher speeds. Future lasers will be able to be switched on and off at frequencies over 60GHz.

Smaller Dimensions

Researchers are busy fabricating semiconductor lasers that are near infinitesimal in size. The electron that is confined in such an ultra-minute space acts in a quantized manner; therefore, quantum mechanics must take over as the dominating physical theory. Three basic devices are used to confine electrons: quantum wells, quantum wires, and quantum dots. A quantum well is two extremely thin sheets of the same semiconductor material sandwiching a second type of semiconductor material. Slicing a narrow strip from the quantum well produces a quantum wire and by dicing the wire, a quantum dot is obtained (see figure 13).

By reducing the dimensions and manipulating the exact composition of the semiconductor material, scientists can literally custom design the resulting electrical properties of the device. Extremely efficient quantum well lasers are currently used in many common devices such as compact disk players, but quantum wires and quantum dots are still being studied in the laboratory. Since such quantum devices contain few conduction electrons, the energy bands present in the bulk material reduce to a few levels that can then be utilized for specific applications (see figure 14).

More and Wider Wavelengths

As there are relatively few semiconducting elements, researchers have few building blocks with which to work. Since there are great, commercial applications for semiconductor lasers (especially in the shorter wavelengths) many scientists have begun to explore this frontier. Recently, there has been progress in designing a ZnSe laser that emits light in the blue-green spectra. There has also been speculation that a GaN based laser could emit near-ultraviolet wavelengths (about 400 nm). Another team of researchers has succeeded in building a blue-purple (417 nm) laser utilizing an indium gallium nitride (5).

Future Applications

The ability to construct smaller, more reliable semiconductor lasers in a variety of wavelengths opens the door for numerous future applications. Applications of lasers operating in the lower wavelengths include high-density optical recording, display systems, spectroscopy, micro-instrumentation, printing, and medicine. The ruggedness, size, and power requirements of today's solid-state semiconductor lasers promise to be a stimulus for future technologies.

References

Bie, Alex. “Tunnel Diodes and Other Goodies.” Popular Mechanics. Feb. 1999.
___. “From Vacuum Tubes to Laser Diodes.” Popular Mechanics. Mar. 1999.

Corcoran, Elizabeth, and Glenn Zorpette. “Diminishing Dimensions.” Scientific American. Oct. 1997.

Development of High Speed DFB and DBR Semiconductor Lasers, NSL, NanoStructures Laboratory, MIT. http://nanoweb.mit.edu/annual-report00/25.htm

“Diode Emits Blue-Purple Pulses at Room Temperature.” Laser World News. http://elecomm.electra.inha.ac.kr/Labs/micropro/webzine/news_2.htm

Forouhar, Siamak. “Semiconductor Lasers for NASA Applications.” NASA Office of Space Access and Technology. http://mishkin.jpl.nasa.gov/spacemicro/TDL_PAPER

Gourley, Paul L. “Nanolasers: Semiconductor Lasers Have Shrunk to Dimensions Even Smaller than the Wavelength of the Light They Emit.” Scientific American. March 1998.

Humphries, James T., and Leslie P. Sheets. Industrial Electronics. Albany, New York: Delmar Publishers, 1993.

Kasap, S. O. Principles of Electrical Engineering Materials and Devices. New York: McGraw-Hill, 1997.

Leutwyler, Kristin. “Plastic Power: Polymers Take a Step Forward as Photovoltaic Cells and Lasers.” Scientific American. Dec. 1996.

Makino, T. “The Effect of Carrier Capture and Escape on the Optical Gain of Quantum-well Semiconductor Lasers.” Canadian Journal of Physics. Feb. 1999.

Olegi, D. “Blue-Green Semiconductor Lasers.” Europhysics, News Extra, Aug. 1995.

Practical Systems Take Shape. National Academy of Sciences, Washington, DC http://www4.nas.edu/beyond/beyonddiscovery.nsf/web/laser

“Solid-State Lasers: Upward to a Focus.” Science News. November 6, 1993.

Tomasi, Wayne. Advanced Electronic Communications Systems. Englewood Cliffs, New Jersey: Prentice Hall, 1994.

Wilson, Jerry D. College Physics. Englewood Cliffs, New Jersey: Prentice Hall, 1994.

Wilson, Jim. “Have Atoms, Will Travel.” Popular Mechanics. July 1997.