All Optical Tutorials on Thorlabs.com

Laser Diode Tutorial

Semiconductor Lasers

Semiconductor lasers are comprised of a large group of binary, ternary, and quaternary elements from Groups II -VI from the periodic table. These lasers can have emission ranges from the blue (~400 nm) to the IR by combining elements from Groups II and VI or Groups IV andVI, respectively. This large emission range combined with the small device footprint, low operating current, low operating cost, and high efficiency makes semiconductor lasers one of the most important and widely used classes of lasers in use today. Please see our Coherent Sources page for a complete listing of all laser diodes offered by Thorlabs. For applications requiring diodes not listed on our Coherent Sources page, please contact technical support.

Fabry-Perot Laser Diodes

The simplest type of semiconductor laser is the Fabry-Perot (FP) laser diode.  In this device, two parallel ends of the semiconductor are cleaved along the crystal axis, creating reflective mirrors forming a Fabry-Perot laser cavity with the semiconductor as the gain medium. Optical coatings are typically applied to the mirror facets to optimize the output power, with the laser emission taken from the low-reflectivity front facet  and high-reflectivity on the back facet to reduce the overall mirror loss. The gain spectrum of the semiconductor medium is quite broad and supports lasing over many longitudinal modes of the FP cavity.  Consequently, FP laser diodes typically operate with multiple longitudinal modes. 

For FP laser diodes that are fabricated such that the optical beam is confined in an optical waveguide with a single-transverse mode (which is also called “single-spatial mode,” but commonly just shortened to “single mode”), the longitudinal mode spacing is determined by Δv = c/2nL, where  c is the speed of light, L is the laser diode chip length, and n is the group index of refraction of the semiconductor waveguide. It is often more convenient to express the mode spacing in terms of wavelength  (Δλ = λ² /2nL), which is more readily measured directly on an optical spectrum analyzer.  

For example, taking typical values for the group index of refraction n = 3.5 and cavity length L = 1 mm, yields longitudinal mode spacing of Δλ = 0.05 nm @ 635 nm and Δλ = 0.3 nm @ 1550 nm.  The number of lasing longitudinal modes and the ratio of the power in the various modes, which is characterized by the side-mode-suppression ratio (SMSR), is influenced strongly by the type of material used to form the semiconductor gain medium (AlGaAs, InGaAsP, AlGaInP, etc.) as well as the bias current and temperature.  For GaAs-based FP lasers, it is often possible to adjust the bias current/temperature such that single-longitudinal mode operation can be achieved with an SMSR of 5-10 dB over a limited current/temperature operating range.  In order to obtain more stable single-longitudinal mode operation with high SMSR (>30 dB) and narrow linewidth (<10 MHz), other types of laser diodes such Distributed Feedback (DFB), Distributed Bragg Reflector (DBR), and Vertical Cavity Surface Emitting Lasers (VCSEL) should be used.  These type of semiconductor lasers will be discussed below.

In general, the output power and wavelength that a laser diode displays is tunable by altering the temperature and/or current. This is extremely evident with IR laser diodes where small changes in temperature greatly affect the small band gaps. Thus almost all laser diodes are temperature tunable, though this tunability is generally small (~10s of nm). Laser diodes also display some current-based power tunability. Increasing the input current increases stimulated emission up to a specified value; however above that value, spontaneous emission starts to compete with the stimulated emission process. It is thus recommended to keep input currents within the specified range for each laser diode.

The polarization of the stimulated emission is generated parallel to the junction plane while spontaneous emission is unpolarized. For high polarization efficiency (50:1 or greater), it is also recommended to keep the input current within the specified range.

The first FP laser diodes utilized a single semiconductor material, predominantly GaAs, to form a single p -n junction diode. These devices were subsequently labeled homojunction laser diodes.^1-4 While these early FP lasers demonstrated the principles of the laser diode, cryogenic temperatures were necessary for cw operation to prevent destruction due to the high current density threshold, J_th≈ 10⁵ A/cm², inherent to these devices. The advent of the heterostructure laser diode several years later reduced the high threshold current density and allowed the development of a wide range of room temperature FP laser diodes.

Figure 1 shows a FP chip on submount laser diode. The chip has two welded gold contacts to the n- and p- doped semiconductor layers. This chip, FPL2000C, is manufactured to form a FP laser cavity tuned to emit 30 mW of CW light at 2000 nm. The FP laser cavity yields a spectral bandwidth of ~15 nm (nom.). This diode features a quatum well structure, which is described below.

Thorlabs' offers several different types of FP laser diodes. Packaged FP laser diodes can be found for the Visible and IR. We also offer butterfly, chip on submount, and C-Mount packaged versions to complete our LD product line. High power, broad area laser (BAL) diodes are also available. For other FP laser diode options please contact tech support.

Heterostructure Laser Diodes

Heterostructure arrangements allowed the widespread development of room temperature cw operation laser diodes. Single Heterostructure (SH) arrangements were first developed, followed quickly thereafter by the Double Heterostructure (DH) laser diode. The DH laser diode is one of the most commonly used laser diodes today. Briefly, DH laser diodes feature low threshold current densities, room temperature operation, and high efficiency.

Double Heterostructure laser diodes are comprised of a thin, active region (100 - 200 nm), surrounded by two thicker (1 - 2 μm) cladding layers, which form the p - n junction. Figure 2 shows a typical DH laser diode structure. In this example, the GaAs active area is 0.15 μm thick with 1 μm cladding layers of p-Al_0.3Ga_0.7As andn-Al_0.3Ga_0.7As. This structure is mounted to a thick GaAs substrate. The configuration reducesJ_th ≈ 1 - 3 kA/cm² and increases efficiency compared to the single junction laser diodes due to

1. Photon confinement in the GaAs active region due to the larger index of refraction of GaAs (n = 3.6) compared to the p- and n- cladding layers (n = 3.4).
2. Carrier confinement in the GaAs active region due to the smaller band gap (E_g ≈ 1.5 eV)of the GaAs compared to the p- and n- cladding layers (E_g ≈ 1.8 eV).
3. Reduction in photon absorption arising from the differences in band gap of the active and cladding layers. Only photons created with energy equal to or greater than the larger band gap cladding layer are absorbed. This results in only minor absorption at the blue tail of the emission profile.

There are several limitations to the DH laser diode that affect wavelength ranges and device performance. The largest drawback to the DH laser diode is the strict lattice matching condition. Lattice mismatches greater than 0.1% can result in interfacial strain between the active and cladding layers, producing non-radiative electron hole recombination. The lattice matching restriction reduces the number of elements that can be used in the active and cladding regions, resulting in a decrease in possible wavelength ranges and increase in J_th .

Thorlabs offers a variety of DH laser diodes in the visible and NIR.

Quantum Well Laser Diodes

Quantum Well (QW) laser diodes are a special class of DH laser diodes where the active area thickness, D, approaches the deBroglie wavelength.

There are many benefits of using QW structures over the DH, or bulk, structures. Only the most important results are discussed here. The reader is directed to Reference 5 and the references therein for further information on QW vs. bulk optical properties. QW structures benefit from a large increase in differential gain compared to standard bulk DH structures. In addition, this gain is less affected by changes in temperature compared to the analogous bulk structure.

A sample QW structure is shown on the left in Figure 3. The QW is comprised of a 10 nm thick GaAs active area surrounded by two confinement layers of Al_0.2Ga_0.8As, each with a thickness 100 nm. Surrounding the confinement layers are two thick, 1 μm layers of a high-band-gap, low-refractive-index material, Al_0.6Ga_0.4As.

This configuration features threshold current densities that are 4 - 5 time smaller than comparable DH heterostructures (J_th ≈ 100 - 300 A/cm²). A reduction in thickness is acheived and photon confinement is increased due to the particular outer cladding/confinement structure utilized. In addition, QWs show an improvement in device performance due to the increase in gain compared to bulk structures.

Quantum Well structures also reduce the strict lattice matching parameters found in DH structures. For very thin QWs, the mismatch in the lattice structure may be 1 - 3%. QW laser diodes with large interfacial lattice differences can operate without the boundry mismatch problems associated with DH laser diodes (non-radiative electron-hole recombination). These types of QWs are typically deemed Strained QWs (SQWs) and have opened up new wavelength ranges previously unavailable to DH laser diodes. These SQWs also improve the absorption characteristics, efficiency, and threshold current density compared to unstrained QWs. Further information on SQWs is available in Reference 5.  

QW structures may be layed out singly between two high band gap materials or in a series arrangement with alternating QW (narrow band gap) / barrier (high band gap) materials. This latter arrangement is termed the Multiple Quantum Well (MQW) arrangement. A MQW structure is shown to the right in Figure 4. In this structure, a repeating unit of a 5 nm layer of low band gap and a 4 nm layer of high band gap is deposited between a p -n junction.

Varying the material composition of the QW or barrier, the layer or barrier thickness, or the number of QWs (e.g. Multi, ML725B8F) can change the emission characteristics of the MQW.⁶ The only restriction in a MQW arrangement is the high band gap material (thickness and band gap) must be sufficient to eliminate electron tunneling, which would greatly reduce device efficiency.

Thorlabs offers a variety of MQW laser diodes in the visible and NIR.

Distributed Feedback Laser Diode

The Distributed Feedback (DFB) laser diode incorporates a grating into one of the cladding layers surrounding the active layer of a DH laser diode. The diffraction grating etched (or deposited) provides much narrower laser line widths and high temperature stability compared to DH laser diodes. Reflections at the end facets are not necessary in DFB laser diodes since the diffraction grating selects the wavelengths that are present in the gain medium.

A DFB laser diode is shown to the right in Figure 5. The InGaAsP active layer, which has a band gap corresponding to emission at λ = 1550 nm, is surrounded by a cladding layer of InGaAsP, which has a slightly larger band gap corresponding to emission at λ = 1300 nm. One of the cladding layers has a varying thickness of period Λ. Each cladding layers is bordered on one side by a high band gap, low-refractive-index material (either p-InP orn-InP). 

The variation in the cladding layer produces a refractive index along the z-direction, n_eff , that is dependent on the z location, 

where the brackets refer to an average over the x-direction, orthogonal to the longitudinal axis. The transverse beam profile along the x-coordinate has a narrow width almost completely contained within the active area and cladding layers. We can impose a periodicity along the z-axis to the refractive index:

where n₀ and n₁ are the refractive indicies of the cladding and substrate layers,Λ is the pitch of the periodic change in refractive index along the interface, and φ is the phase factor. Utilizing Bragg's Law from a grating or other periodic elements, forwards and backwards propogating beams are coupled if 

where <n_eff > is the average refractive index along the z-axis. Under this simplified analysis, it is observed that only one wavelength can exist for a given pitch, Λ. A detailed analytical treatment beyond the simplified description here is available in References 7 and 8 and references therein. Thorlabs offers two DFB laser diodes operating at 1310 nm and 1550 nm. For other DFB laser diode needs please contact technical support.

Vertical-Cavity Surface Emitting Laser Diodes

Vertical-Cavity Surface Emission Laser (VCSEL) diodes are a unique class of laser diodes where the emission occurs perpendicular to the active layer/junction plane. This is in contrast to the previously described laser diodes, where the light propogation/amplification is parallel to the junction plane. VCSEL lasers are typically employed where low threshold currents,J_th ≈ 3 - 5 kA/cm², and high emitter densities are required. The actual threshold currents (approximately a few milliamps) are much lower than FP laser diodes because of the small active area.

The short active area/gain medium limits scattering and absorption, improving efficiency. The short active area also allows VCSELs to typically operate in a TEM₀₀ mode, even when high currents well above the threshold level are applied. VCSELs are not, however, recommended for high-power applications due to the vertical emission and short active area, which limits the gain length.

A VCSEL cavity is shown in Figure 6 to the right. Because the emission from these laser diodes is perpendicular to the junction plane, a high density of emitters can be produced over a small area. In addition, these devices can be configured for very high packaging density applications, as emitters can be very closely spaced compared to typical FP laser diodes.

The active area of a VCSEL is comprised of several strained QW layers, approximately 5 - 10 nm thick, utilizing a high band gap material of 4 - 6 nm thickness between the wells. Figure 7a shows the strained QW structures forming the active area of a VCSEL.

The strained quantum wells, which are contained within two cladding layers, are shown in Figure 7b. These cladding layers make up the laser cavity of the VCSEL. The thickness of the laser cavity is approximately 1λ. VCSELs operating in the NIR (λ = 1-3 um) with similarly sized laser cavities have a mode spacing of Δλ ≈ 100 - 300 nm. This mode spacing allows single longitudinal mode emission at varying input currents.

The laser cavity is contained within a repeating structure of λ/4 thick high-refractive-index / low-refractive-index layers. This layering of many (on the order of 15 - 25) Bragg reflectors produces one mode with peak reflectivity at each of the quarter-wave stacks and is therefore amplified by the gain medium. Other modes present in the cavity are reduced by destructive interference from the λ/4 stacks. This vertical structure is shown in Figure 7c. The structure is anchored on a thick substrate and metal contact. The emission surface features a λ/2 thick layer (for phase matching) and metal contact with a circular aperture of approximately 5 - 10 μm.

Thorlabs' offers several packaged VCSELs in the NIR. For VCSEL solutions at other wavelengths please contact technical support.

¹ Hall, R.N., Fenner, G.E., Kinhsley, J.D., Dills, F.H., Lasher, G., Coherent Light Emission from GaAs Junctions,Phys. Rev. Lett. 9, 366 (1962).
² Nathan, M.I., Dumke, W.P., Burns, G., Dills, F.H., Lasher, G., Simulated Emission of Radiation from GaAsp-n Junctions, Appl. Phys. Lett.1, 62 (1962).
³ Holonyak, Jr., N. and Bevacqua, S.F., Coherent (Visible) Light Emission from Ga(As_1-xP_x) Junctions, Appl. Phys. Lett. 1, 82 (1962).
⁴ Quist, T.M., Keyes, R.J., Krag, W.E., Lax, B., McWhorter, A.L., Rediker, R.H., Zeiger, H.J., Semiconductor Maser of GaAs, Appl. Phys. Lett. 1, 91 (1962).
⁵ Svelto, O. and Hanna, D.C., Principles of Lasers, 4th ed., Plenum Press, New York (1998).
⁶ Kittel, C., Introduction to Solid-State Physics, 6th ed., Wiley, New York (1986).
⁷ Kogelnik, H. and Shank, C.V., Stimulated Emission in a Periodic Structure, Appl. Phys. Lett. 18, 152 (1971).
⁸ Otsuka, K., Winner-Takes-All and Antiphase States in Multimode Lasers, Phys. Rev. Lett. 67, 1090 (1991).