Coherent light generators – Particular active media – Semiconductor
Reexamination Certificate
2001-03-30
2004-10-26
Wong, Don (Department: 2828)
Coherent light generators
Particular active media
Semiconductor
C372S046012
Reexamination Certificate
active
06810054
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to diode laser devices.
BACKGROUND OF THE INVENTION
A laser is a device that produces monochromatic, coherent light through stimulated emission of photons from the atoms or molecules of a lasing medium, which have been excited from the ground state to a higher energy level. For example, the lasing material may be a rare-earth element (e.g., a lanthanide such as Nd or Er) dispersed as a dopant within a carrier. The laser medium is part of an optical cavity or resonator defined by highly reflective surfaces; these reflect light back and forth through the medium to stimulate emission. More specifically, if an energy-level population inversion is created by excitation of the laser medium, the spontaneous emission of a photon from an excited atom or molecule returning to its ground state can stimulate the emission of photons of identical energy from other excited atoms or molecules. The initial photon thereby generates a cascade of photons that have identical energy and are exactly in phase, and this cascade effect is multiplied as the photons travel repeatedly between the mirrors. The mirrors do not have perfect reflectivities, however, allowing a portion of this cascade to exit the cavity and thereby provide a laser output.
Excitation of the lasing medium can be accomplished in any of a variety of ways, including optical pumping, current injection, or the use of an electrical discharge. A diode semiconductor laser utilizes current injection, and such a laser is typically fabricated initially from an intrinsic semiconductor heterostructure grown on a substrate. The final p-i-n structure for a laser diode, which is obtained by adding n and p doping stripes to the top and bottom portions of the structure, differs from an ordinary p-n junction in containing an intervening intrinsic layer consisting of either a single quantum well (SQW) or multiple quantum wells (MQWs). A widely studied quantum well system consists of alternating layers of GaAs and Al
1-x
Ga
x
As. In this system, the energy difference between the unfilled conduction band edge and the filled valence band edge of GaAs, referred to as its “bulk energy gap,” is smaller than the corresponding gap in Al
1-x
Ga
x
As. Since 60% of the difference between the two bulk energy gaps at the heterojunctions of this system is accomodated by the conduction band, with the remaining 40% accommodated by the valence band, the conduction and valence band wells occur in the same layer of the GaAs structure.
At T=0° K, no electrons populate the conduction band of a laser diode heterostructure. As the temperature is raised, a fixed number of electrons in thermal equilibrium with the lattice are excited across the T=0° K energy gap. The conduction band then becomes partially filled through a wide range of energies according to a Fermi distribution. As more electrons continue to be promoted across the gap with increasing temperature, states near the conduction and valence band edges gradually become filled and unavailable for photon absorption, and the majority of the absorption is shifted to higher energies. Absorption at the T=0° K gap, on the other hand, is totally suppressed when a non-equilibrium concentration of charge carriers is injected into the wells so as to completely fill the band edges up to some quasi-Fermi level. The new effective energy gap that results, &Dgr;F, is the difference between the quasi-Fermi level in the conduction band (F
c
) minus the quasi-Fermi level in the valence band (F
&ngr;
).
The T=10° K energy gap of a quantum well can itself be increased relative to the energy gap of the bulk material by reducing the thickness of the quantum well. The reason for this is simply that the quantum-mechanical effect of the discontinuity in the band edge is such as to admit only discrete sub-band states within the well, and each sub-band has associated with it a different confinement energy relative to the zero-point energy at the bottom of the well. Other factors affect the energy position of band edge recombination (and the absorption spectrum of the material); these include band-gap renormalization (shrinkage of the energy gap) due to carrier-carrier interactions at high carrier concentration, and temperature.
As explained above, to achieve lasing in any structure, optical feedback must be provided (via mirrors) through a gain medium. In a semiconductor laser diode, the expression for the gain coefficient, g(&ngr;), for a laser of this type at a given frequency &ngr; is given by &agr;(e
(&Dgr;F−h&ngr;)/kT
−1). From the condition that &Dgr;F>h&ngr;>E
gap
(where E
gap
is the T=10° K energy gap at thermal equilibrium), lasing cannot occur in a region of the crystal without the injection of a non-equilibrium concentration of carriers into that region. Typically, the cleaved output facets of a semiconductor crystal serve as the mirrors.
Laser diodes generally fall into two categories: “gain-guided” structures and “index-guided” structures. The present invention is concerned with the latter. For both types of structures, output light generated from electron-hole recombination in the quantum well is optically confined in the plane of the well layer, since this layer generally has a higher refractive index than the adjacent barrier layer. A crude rule of thumb illustrating this point is the so-called Moss rule, which states that n
4
E
gap
=77. In index-guided structures, refractive index differences are also introduced within the plane of the quantum well, further confining the output light laterally within the active region near the region of the doping stripes. Thus, refractive-index differences confine the light both transversely (keeping it within the well layer) and laterally (affecting the distribution of light within the well layer).
In gain-guided structures, although the light is not strictly confined laterally under the doping stripe by differences in refractive index, any light escaping the active region is not amplified simply because the non-equilibrium carrier concentrations injected into the well by the doping stripe do not diffuse far from the active region before recombining.
One of the advantages a properly designed index-guided structure provides over gain-guided structures—an advantage utilized in this application—is the ability to confine the output light to a single lateral mode of the cavity. This results in output light that is diffraction-limited and can be focused to a desired spot size with a longer image distance. Consequently, a greater depth of focus near the image plane is achieved. On the other hand, index-guided structures that confine light to a single lateral output mode suffer from a serious drawback. Their cavity size is generally smaller than 10 &mgr;m, which restricts output power to levels that do not damage the output facets. (Output damage to the facets is caused by heating due to enhanced absorption at the facets, which is itself a result of band bending induced from high surface recombination at the facet.) Several strategies have been employed in the past to overcome this difficulty, including the use of larger cavities having curved output facets, non-absorbing mirror facets, and implementation of arrays of single-mode cavities that are weakly optically coupled and are designed to lase only in the fundamental mode of the coupled system. Each of these strategies is discussed below.
When curved output facets are used with a larger cavity, some of the less-collimated light originating from higher-order lateral modes of the cavity is reflected away from the optical axis and rejected before it can be fed back through the gain medium of the cavity. Light reflected back and forth n times by the curved mirror facets of an unstable resonator cavity acts as if it has passed straight through a system of n repeating diverging lenses. The optical-transfer matrix of any system of repeating lenses such as this is obtained by applying Sylvester's Theorem to the transfer matrix of one
Sousa John Gary
Stoltz Steven
Jackson Cornelius H.
Presstek Inc.
Testa, Hurwitz & Thibeault. LLP
LandOfFree
Mode-limiting diode laser structure does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Mode-limiting diode laser structure, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Mode-limiting diode laser structure will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3284202