Semiconductor Laser

Background

A laser, which is an acronym for Light Amplification by Stimulated Emission of Radiation, is a device that converts energy into light. Electrical or optical energy is used to excite atoms or molecules, which then emit light. A laser consists of a cavity, with plane or spherical mirrors at the ends, that is filled with lasable material. This material can be excited to a semistable state by light or an electric discharge. The material can be a crystal, glass, liquid, dye, or gas as long as it can be excited in this way.

The simplest cavity has two mirrors, one that totally reflects and one that reflects between 50 and 99%. As the light bounces between these mirrors, the intensity increases. Since the laser light travels as an intense beam, the laser produces very bright light. Laser beams can also be projected over great distances, and can be focused to a very small spot.

The type of mirror determines the type of beam. A very bright, highly monochromatic (one wavelength or one color) and coherent beam is produced when one mirror transmits only 1-2% of the light. If plane mirrors are used, the beam is highly collimated (made parallel). The beam comes out near one end of the cavity when concave mirrors are used. The type of beam in the first case makes lasers very useful in medicine since these properties allow the doctor to target the desired area more accurately, avoiding damage to surrounding tissue.

A semiconductor laser converts electrical energy into light. This is made possible by using a semiconductor material, whose ability to conduct electricity is between that of conductors and insulators. By doping a semiconductor with specific amounts of impurities, the number of negatively charged electrons or positively charged holes can be changed.

Compared to other laser types, semiconductor lasers are compact, reliable and last a long time. Such lasers consist of two basic components, an optical amplifier and a resonator. The amplifier is made from a direct-bandgap semiconductor material based on either gallium arsenide (GaAs) or InP substrates. These are compounds based on the Group III and Group V elements in the periodic table. Alloys of these materials are formed onto the substrates as layered structures containing precise amounts of other materials.

The resonator continuously recirculates light through the amplifier and helps to focus it. This component usually consists of a waveguide and two plane-parallel mirrors. These mirrors are coated with a material to increase or decrease reflectivity and to improve resistance to damage from the high power densities.

The performance and cost of a semiconductor depends on its output power, brightness, and operating lifetime. Power is important because it determines the maximum throughput or feed rate of a process. High brightness, or the ability to focus laser output to a small spot, determines power efficiency. Lifetime is important because the longer a laser lasts, the less it costs to operate, which is especially critical in industrial applications.

The simplest semiconductor lasers consist of a single emitter that produces over one watt of continuous wave power. To increase power, bars and multibar modules or stacks have been developed. A bar is an array of 10 to 50 side-by-side individual semiconductor lasers integrated into a single chip and a stack is a two-dimensional array of multiple bars. Bars can produce 50 watts of output power and last over 5,000 hours. Because such high powers produce a lot of heat, cooling systems must be incorporated into the design.

History

The concept behind lasers was first proposed by Albert Einstein, who showed that light consists of wave energies called photons. Each photon has an energy that corresponds to the frequency of the waves. The higher the frequency, the greater the energy carried by the waves. Einstein and another scientist named S. N. Bose then developed the theory behind the phenomenon of photons' tendency to travel together.

Laser action was first demonstrated in the microwave region in 1954 by Nobel Prize winner Charles Townes and his co-workers. They projected a beam of ammonia molecules through a system of focusing electrodes. When microwave power of appropriate frequency was passed through the cavity, amplification occurred and the term Microwave Amplification by Stimulated Emission of Radiation (M.A.S.E.R.) was born. The term laser was first coined in 1957 by physicist Gordon Gould.

Townes also worked with Arthur Schawlow and the two proposed the laser in 1958, receiving a patent in 1960. The first practical laser was invented that same year by a physicist named Theodore Maiman, while he was employed at Hughes Research Laboratories. This laser used a pink ruby crystal surrounded by a flash tube enclosed within a polished aluminum cylindrical cavity cooled by forced air. Two years later, a continuous lasing ruby was made by replacing the flash lamp with an arc lamp.

In 1962, laser action in a semiconductor material was demonstrated by Robert Hall and researchers at General Electric, with other United States researchers soon following. It took about another decade for the first semiconductor diode laser to be developed that could operate at room temperature, which was first demonstrated by Russian researchers. Bell Labs followed the Russian researchers' success, while also improving laser lifetimes. In 1975, Diode Laser Labs of New Jersey introduced the first commercial room-temperature semiconductor laser.

Despite this progress, these lasers still were inadequate for telecommunications applications. Instead they found wide use (after other performance and lifetime improvements) in audio compact disks after Philips (The Netherlands) and Sony (Japan) developed a CD in 1980 using a diode laser. By the end of the decade, tens of millions of CD players were being sold every year. More recently, digital video disks have become available for optical storage, which are also based on diode lasers.

As power has increased, semiconductor lasers have expanded into other applications. Since 1995, performance of high-power diode lasers has jumped by a factor of 25. With this higher reliability, large groups of diode lasers can now be combined to create "stacks" of up to 25 individual diode lasers.

In 1999, laser-diode revenues represented 64% of all lasers sold, up from 57% in 1996, and were projected to reach 69% in 2000. In terms of units sold, semiconductor lasers have accounted for about 99% of the total (over 400 million units), which means most laser light is now produced directly or indirectly (via diode pumping) by semiconductor lasers. In addition to industrial applications, semiconductor lasers are being used as pump sources for solid state lasers and fiber lasers, in graphics applications such as color proofing and digital direct-to-plate printing, and for various medical and military applications (target illumination and ranging). In 2000, Laser Focus World estimated that about 34% of medical therapy lasers were of the semiconductor type.

Raw Materials

The conventional semiconductor laser consists of a compound semiconductor, gallium arsenide. This material comes in the form of ingots that are then further processed into substrates to which layers of other materials are added. The materials used to form these layers are precisely weighed according to a specific formula. Other materials that are

used to make this type of laser include certain metals (zinc, gold, and copper) as additives (dopants) or electrodes, and silicon dioxide as an insulator.

Design

The basic design of a semiconductor laser consists of a "double heterostructure." This consists of several layers that have different functions. An active or light amplification layer is sandwiched between two cladding layers. These cladding layers provide injection of electrons into the active layer. Because the active layer has a refractive index larger than those of the cladding layers, light is confined in the active layer.

The performance of the laser can be improved by changing the junction design so that diffraction loss in the optical cavity is reduced. This is made possible by modifying the laser material to control the index of refraction of the cavity and the width of the junction. The index of refraction of the material depends upon the type and quantity of impurity. For instance, if part of the gallium in the positively-charged layer is replaced by aluminum, the index of refraction is reduced and the laser light is better confined to the optical cavity.

The width of the junction can also affect the performance. A narrow dimension confines the current to a single line along the length of the laser, increasing the current density. Peak power output must be limited to no more than 400 watts per cm (0.4 in) length of the junction and current density to less than 6,500 amperes per centimeter squared at the junction to extend the life of the laser.

The Manufacturing Process

Making the substrate

• 1 The substrates are made using a crystal pulling technique called the Czochralski method, where a crystal is grown from a melt. The elements are first mixed together and then heated to form a solution. The solution is then cooled, which solidifies the material. A seed crystal is attached to the bottom of a vertical arm so that the seed barely contacts the material at the surface of the melt. The arm is raised slowly, and a crystal grows underneath at the interface between the crystal and the melt. Usually the crystal is rotated slowly in order to avoid producing impurities in the crystal. By measuring the weight of the crystal during the pulling process, computer controls can vary the pulling rate to produce any desired diameter.

Growing the layers

• 2 The most common method for growing the layers onto the substrate is called liquid-phase epitaxy (LPE). Layers that have the same or fixed crystal-growth direction as that of the substrate can be grown on the substrate when the substrate comes into contact with a solution of the desired composition. As the temperature is decreased, the semiconductor compound (such as GaAs) comes out of the solution in crystalline form and is deposited onto the substrate.

  An LPE system consists of a reactor (where the layers are grown), a substrate loading system, a pump and exhaust system (for removing air or impure gases after the materials are put in or taken out), a gas flow system (to move hydrogen gas through the reactor to remove impure gases) and a temperature control system. Pure materials are used for making the reactor so that the layers are not contaminated. The loading box is usually filled with nitrogen gas to purge the air while opening the reactor. The reactor typically consists of a quartz tube, in which a graphite boat and boat holder are placed. The graphite boat consists of an outer frame, a substrate holder, a spacer and a melt box.

• 3 The source materials for the layers are first rinsed and etched in order to clean the surface. After drying the etched materials, they are loaded into each melt box of the graphite boat. To grow each layer, the materials are first melted by heating to a specific temperature and then the substrate holder is pulled along with the substrate from the first melt to the next. The substrate is kept at each melt for a certain time under a fixed cooling rate, usually 33°F (0.5°C) per minute, according to a specific program designed for each composition. The temperature is automatically controlled using thermocouple sensors.

Fabricating the laser device

• 4 After the layered structure is grown, several other processes are completed to form the laser device. First, the substrate is mechanically polished until the thickness decreases to 70-100 microns in preparation for cleaving. Next, a very thin silicon dioxide film is formed on the substrate surface. Stripes are formed by photolithography and chemical etching. Contact electrodes are applied using an evaporation method. Next, a laser resonator is formed by cleaving the wafer along parallel crystal planes. The completed laser devices are then attached to a copper heat sink on one side and a small electrical contact on the other.

Quality Control

The substrate onto which the semiconductor structure is grown must meet certain requirements regarding crystal direction, etch pit density (EPD), impurity concentration, substrate thickness and wafer size. The crystal direction must be within several degrees. Etch pits, which are rectangular hills or holes, are revealed by selectively etching the substrate with some type of acid solution. The etch pit density (number of etch pits per square centimeter) is used for estimating dislocation density, which affects the laser lifetime. An EPD of 10 ³ per centimeter squared or less is required. Impurity concentrations are around 10 ¹⁸ per cubic centimeter. Substrates can range in size up to 3 in (7.6 cm) in diameter and typically are sliced into 350-micron thick pieces.

After the growth process, the surface of the semiconductor wafer is examined by an optical microscope. To examine the layered structure, a ground or cleaved cross section of wafer is stained and etched to increase the contrast of the layers using a scanning electron microscope. X-ray diffraction is used to determine the compositions of the layers and to measure the lattice patterns of the structure. The impurity concentration and refractive index of the layers is also measured using several analytical methods. After the laser device is fabricated, such operating parameters as voltage/current curves, threshold current density and spectral characteristics are measured.

The Future

Industry analysts from Frost & Sullivan predict that the diode laser systems market will reach nearly $4.6 billion by 2005. This growth is partially due to expanding applications in materials processing as high-power diode lasers become less expensive than solid state lasers. The compact size and electrical efficiency also make high-power semiconductor lasers attractive for industrial applications such as heat treating and welding. New material compositions and processing methods are also being developed to expand the applications.