The present disclosure relates to lasers and, in particular, to quantum cascade laser and related methods.
Quantum cascade lasers (QCLs) differ from traditional semiconductor diode lasers in that they do not use a p-n junction for light emission. Multiple active regions are “cascaded” so that each injected electron can emit multiple photons and therefore enhance laser gain. Each active region includes a multi-layered semiconductor material structure. This multi-layered semiconductor material structure is designed to have an electronic band structure that gives the desired emission wavelength, and is manufactured with nanometer-level thickness control.
Most commercially available QCLs are of the “edge-emitting” variety. In these approaches, linear ridges are created by etching vertically through the layered structure, which has been grown on a wafer substrate. These are cleaved so that the active region comprises a rectangular ridge several millimeters in length, which forms a waveguide. The laser radiation is amplified by passing back and forth along the axis of the ridge. Laser emission occurs on the end facets of these ridges.
In
Generally, a QCL may comprise a substrate, and a sequence of semiconductor epitaxial layers adjacent the substrate and defining an active region, an injector region adjacent the active region, and a waveguide optically coupled to the active region. The active region may include a plurality of stages. Each stage may have an upper laser level and a lower laser level defining respective first and second wave functions. The upper laser level may have an upper laser level average coordinate, and the lower laser level may have a lower laser level average coordinate. The upper laser level average coordinate and the lower laser level average coordinate may have spacing therebetween of less than 10 nm. Wave functions for all active region energy levels located below the lower laser level may have greater than 10% overlap with the injector region. Energy spacing between uppermost laser levels may exceed 60 meV. A depth of the active region may be greater than 100 meV.
Additionally, the upper laser level average coordinate and the lower laser level average coordinate may be derived based upon ∫x ψ2dx. The active region may have a width exceeding 15 μm.
In some embodiments, the waveguide and the active region may define a double-channel shallow-ridge configuration with channel width exceeding 50 μm and a ridge width exceeding 15 μm. In other embodiments, the waveguide and the active region may define a two-level ridge configuration, and a first level ridge may extend into the active region. Also, a second level ridge may extend to a depth less than that of the first level ridge. Moreover, some embodiments may have the waveguide and the active region defining a multi-channel configuration with greater than 2 channels.
Moreover, the QCL may include an insulator layer over the sequence of semiconductor epitaxial layers, and an electrically conductive layer over the insulator layer. The waveguide may define first and second opposing facets, each facet configured to emit a laser output.
Another aspect is directed to a QCL system. The QCL system may include a QCL comprising a substrate, and a sequence of semiconductor epitaxial layers adjacent the substrate and defining an active region, an injector region adjacent the active region, and a waveguide optically coupled to the active region. The waveguide may define first and second opposing facets configured to respectively emit first and second laser outputs. The active region may include a plurality of stages. Each stage may have an upper laser level and a lower laser level defining respective first and second wave functions. The upper laser level may have an upper laser level average coordinate, and the lower laser level may have a lower laser level average coordinate. The upper laser level average coordinate and the lower laser level average coordinate may have spacing therebetween of less than 10 nm. Wave functions for all active region energy levels located below the lower laser level may have greater than 10% overlap with the injector region. Energy spacing between uppermost laser levels may exceed 60 meV. A depth of the active region may be greater than 100 meV. The QCL system may further comprise a plurality of mirrors, and a half-wave plate configured to combine the first and second laser outputs.
Yet another aspect is directed to a method of making a QCL. The method may include forming a sequence of semiconductor epitaxial layers adjacent a substrate and defining an active region, an injector region adjacent the active region, and a waveguide optically coupled to the active region. The active region may include a plurality of stages, each stage having an upper laser level and a lower laser level defining respective first and second wave functions. The upper laser level may have an upper laser level average coordinate, and the lower laser level may have a lower laser level average coordinate. The upper laser level average coordinate and the lower laser level average coordinate may have spacing therebetween of less than 10 nm. Wave functions for all active region energy levels located below the lower laser level may have greater than 10% overlap with the injector region. Energy spacing between uppermost laser levels may exceed 60 meV. A depth of the active region may be greater than 100 meV.
The present disclosure will now be described more fully hereinafter with reference to the accompanying drawings, in which several embodiments of the invention are shown. This present disclosure may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the present disclosure to those skilled in the art. Like numbers refer to like elements throughout, and base 100 reference numerals are used to indicate similar elements in alternative embodiments.
The 10 W continuous wave (CW) optical power level is presently considered to be potentially difficult with single QCL emitters, which undermines QCLs positioning as a technology of the future in some applications (where high power is needed in a small package). The present disclosure describes a design for a compact laser module projected to deliver over 30 W of optical power into a high brightness Midwave Infrared (MWIR) beam. The unprecedented combination of size, weight, and power for the module may be helpful in these applications and other high power MWIR laser applications.
Broad-Area Configuration
Optical power from traditional narrow-ridge QCLs will likely stay close to 5 W [1, 2] (which results into 3.5 W to 4.0 W on system level) due to two main reasons: limited wall plug efficiency and optical damage of the front facet. CW MWIR QCL efficiency has been clamped at approximately 20% for the last five years and device modeling results indicate that there is not much room for further improvement. At the same time, even if higher efficiency devices were available, it would be very difficult to substantially increase power level as QCLs tend to fail at optical power densities of 10 MW/cm2, which roughly corresponds to total power level of 5 W for narrow (10 μm-wide) devices.
Increase in power level from single QCL emitters, therefore, reduces to (a) increase in laser core dimensions (power scaling) and (b) increase in the emission aperture size (lower optical power density at the output facet).
The broad-area configuration QCL 23 (substrate 24,
Laser core thermal resistance can be drastically reduced and laser self-heating can be suppressed by reducing laser core thickness. Device modeling results demonstrate that this will lead to a lower temperature for the same total electrical power dissipated in the thinner core, or, which is of primary interest here, to the same average temperature for much wider devices [3,4]. This makes power scaling with lateral device dimensions possible. In the following, a laser design and laser packaging tailored to the broad-area configuration is disclosed. The design ideas presented here can also be used for optimizing performance of traditional (narrow ridge) QCLs (See
Laser Core Design
CW broad-area QCL performance depends on laser thermal resistance as well as on its pulsed laser characteristics, such as injection efficiency for the upper laser level, ηi, differential modal gain, Γg, waveguide losses, αw, and transparency current, Jtr. The present disclosure provides analysis of several QCL designs in the broad-area configuration and observed a general trend: structures that have a low transparency current tend to outperform structures with a higher transparency current even when pulsed laser efficiency is the same in both cases. This is illustrated below.
Measured ηi, Γg, αw, and Jtr for the two structures were used for CW optical power projections in the broad-area configuration (ridge width>10 μm) using the model developed and validated in Ref. 4. It was determined that the structure with Jtr=0.4 kA/cm2 had a much more efficient power scaling with increase in ridge width, i.e. a higher projected total CW optical power. Qualitatively, the root cause of the different projected behavior is the following. In addition to increase in ridge width, power scaling assumes employment of long cavity HR-coated devices that have low mirror losses, αm. A combination of a low Jtr and low mirror losses results into a low threshold current density Jth as the latter is defined by (Γg and αw were approximately the same for the two designs):
The low Jth, in turn, leads to a lower laser core temperature at threshold and, therefore, a higher slope efficiency in CW mode. As a consequence, one of the goals in broad-area QCL design should be a reduction in transparency current density.
Transparency current density is a function of carrier backscattering to the lower laser level 3 and injection efficiency into the upper laser level 4, both very difficult to predict from first principles.
However, the active region/injector coupling cannot be made arbitrarily large for the following reasons:
The new QCL design principles that were developed in this disclosure and described below may provide an approach to all the potential problems listed above. Specifically, laser core layers thicknesses and compositions are carefully adjusted to achieve one or more the following design characteristics in a single structure:
The disclosed design provides a larger active region depth and/or active region/injector coupling than either single-phonon, double-phonon, or non-resonant extraction designs and, in contrast to the bound-to-continuum design, it is based on a vertical laser transition. Therefore, it is substantially different from all the main QCL designs reported in literature.
All these design characteristics are combined in an exemplary laser design presented in
The design is projected to have a high performance in the broad-area configuration. The design will also likely outperform other active region designs such single-phonon, double-phonon, bound-to-continuum, and non-resonant-extraction designs in the traditional narrow ridge configuration.
Beam Quality Control
Discrimination against high-order modes is one of the technical challenges in broad-area QCL design. Possible approaches include tilted-cavity waveguide geometry [5], employment of FIB-defined high order mode discriminating features into the waveguide [6], and other methods. The simplest and most practical solution addressed here is the shallow-ridge waveguide (SRW) configuration [7, 8], where etching of the ridge is stopped before reaching the laser core (
The SRW configurations was explored in Refs. 7 and 8 only in pulsed mode and no SRW design principles for CW operation were outlined in those reports. In addition, only gain-guided configuration was explored in Ref. 8 that is unlikely to be sufficient to ensure an efficient high-order mode suppression in CW mode: laser core self-heating in CW mode reaches its highest value at the center of the laser core, which favors higher order modes. Therefore, a large mode discrimination in terms of losses/gain and mode overlap factor has to be applied to achieve a high quality beam.
The present disclosure provides a numerical analysis for the optimal design of the SRW described below. The shallow ridge configuration uses etching features that leaves the slab laser core intact, as shown in
To simulate the effects of current spreading on gain, and thus gain guided modes, assumptions were made about the temperature behavior of the device. Threshold is treated with a 140 K characteristic temperature [9]:
Transparency current is taken to have the same behavior with characteristic temperature T0, forcing differential gain having an inverse relationship:
This gives form for gain (and imaginary index) in the laser core:
Taking this as the temperature and (growth direction) current dependent form of gain in the laser core, the simulation is run in two steps. The first finds steady state current and temperature distribution assuming that the epi-side is kept at 293 K and 0 V, with a positive voltage applied to the substrate side.
Current spreads beyond the width of the shallow ridge as it reaches the laser core. Temperature buildup occurs nearly entirely in the laser core. Temperature spreads through the substrate around the area of high current density (high heat generation) and reaches the heat sink.
The current distribution on the epitaxial side of the laser core features spikes in density just below the walls of the central ridge, creating “ears”. Overall gain begins to degrade before the “ears” exceed rollover due to device heating, so local spikes in current density should not be a concern. The substrate side shows a more Gaussian behavior, with current spreading causing a wider and lower distribution. Gain is then stronger in the epi-side direction. Full width at half maximum (FWHM) of the current spreading for the 30 μm ridge is >50 μm.
Temperature build up in the device (diagrams 83, 84 of
The depth of the channel provides vertical confinement. In general, deeper channels lead to greater mode confinement in the laser core, and thus, more modal gain. There is a “squeezing” effect for higher order modes, allowing them to spread laterally with deep enough channels, pushing them toward the high gain central region. Channel depth should be chosen (after central ridge width) to allow just enough gain to not exceed mirror losses for higher order modes to allow for maximum fundamental mode net gain.
The overall width of the channels was constructed to allow for an overall width of 150 μm for central ridge and channels combined. Wide channels are necessary to provide as much loss as possible for higher order modes. This width left 100 μm of un-etched material on each side for structural support during epi-down mounting. The optimal design for maximum gain with no net gain for higher order modes was a 30 μm ridge with 60 μm wide by 2.7 μm deep etched channels in a 10 mm long device.
The structure reported in Ref. 7 was the only shallow-ridge RWG QCL structure explored so-far. It had 50 μm-wide channels. Channel width<50 μm is typically used for QCLs. The present disclosure provides that it is advantageous to use >50 μm wide channels to ensure higher order mode suppression in high duty cycle operation and only narrow ridges without metal contact openings at the edges of the chip are sufficient for structural support in the epi-down configuration required for CW operation.
Current spreading and mode losses are controlled with choosing proper etching depth in the proposed configuration. Mode losses can be independently controlled in the modified “two-level etch” configuration presented in
The traditional double-channel SRW can also be extended to a multiple-channel (>2 channels) design shown in
Facet Damage Control and Packaging
As mentioned above, high power CW operation leads to QCL output facet damage. The problem is exacerbated (damage threshold reduces) when HR-coating is applied to the back facet and AR-coating is applied to the front facet of the laser due to a residual absorption in the coating materials. To maximize optical power from a single emitter (especially in the broad-area configuration with a very high output power level), the present disclosure provides a concept of a new packaging for a single QCL emitter that can be used both for traditional QCLs and for the broad-area devices.
A QCL chip in this configuration is used without any coating material applied to either facet or with coating material that has a high damage threshold. The basic idea of the new design is to independently collect optical power emitted from the two laser facets and then recombine the two beams using one of many possible beam combining approaches, such as the spatial beam combining (placing the two beam close to each other), polarization beam combining, etc.
Specifically, beam from several uncoated chips lasing at different wavelengths can be combined into a single high brightness beam using polarization beam combining.
Referring now additionally to
Moreover, the QCL 123 illustratively includes an insulator layer 128 over the sequence of semiconductor epitaxial layers, and an electrically conductive layer 130 over the insulator layer. Additionally, the upper laser level average coordinate and the lower laser level average coordinate are derived based upon ∫xψ2dx.
In illustrated embodiment, the waveguide 126 and the active region 125 define a double-channel shallow-ridge configuration with channel width exceeding 50 μm and a ridge width exceeding 20 μm. The active region 125 has a width exceeding 20 μm. Also, the QCL 123 defines two channels 131, 132 abutting opposing longitudinal sides of the ridge.
Referring now additionally to
Referring now additionally to
Referring now additionally to
The QCL 423 system further comprise a plurality of optical elements 443-448b configured to combine the first and second laser outputs into a combine laser output 450. The plurality of optical elements 443-448b comprises first and second collimating lenses 448-448b configured to collimate laser outputs respectively from the first and second opposing facets 441, 442. The plurality of optical elements 443-448b comprises first, second, and third mirrors 443-445 configured to direct the first and second laser outputs, a half-wave plate 446 receiving the second laser output, and a polarizer 447 configured to reflect the first laser output and pass the second laser output.
Yet another aspect is directed to a method of making a QCL 123. The method includes forming a sequence of semiconductor epitaxial layers adjacent a substrate 124 and defining an active region 125, an injector region 127 adjacent the active region, and a waveguide 126 optically coupled to the active region. The active region 125 includes a plurality of stages, each stage having an upper laser level and a lower laser level defining respective first and second wave functions 63-66, 68, 70-72, 74-77 (diagrams 62, 67 & 73). The upper laser level has an upper laser level average coordinate, and the lower laser level has a lower laser level average coordinate. The upper laser level average coordinate and the lower laser level average coordinate have spacing therebetween of less than 10 nm. Wave functions for all active region energy levels located below the lower laser level have greater than 10% overlap with the injector region. Energy spacing between uppermost laser levels exceeds 60 meV. A depth of the active region is greater than 100 meV.
Referring now additionally to
Many modifications and other embodiments of the present disclosure will come to the mind of one skilled in the art having the benefit of the teachings presented in the foregoing descriptions and the associated drawings. Therefore, it is understood that the present disclosure is not to be limited to the specific embodiments disclosed, and that modifications and embodiments are intended to be included within the scope of the appended claims.
This application is a divisional of U.S. application Ser. No. 17/002,129 filed Aug. 25, 2020, which is a divisional of U.S. application Ser. No. 16/136,713 filed Sep. 20, 2018, now U.S. Pat. No. 10,811,847, which is based upon prior filed Application No. 62/608,761 filed Dec. 21, 2017, which are hereby incorporated herein in their entireties by reference.
This invention was made with government support under award No. N00014-16-1-2424, which was awarded by the Office of Naval Research's funding. The government has certain rights in the invention.
Number | Name | Date | Kind |
---|---|---|---|
5263037 | Trutna, Jr. | Nov 1993 | A |
8014430 | Patel et al. | Sep 2011 | B2 |
8068524 | Patel et al. | Nov 2011 | B1 |
8121164 | Lyakh | Feb 2012 | B1 |
9077153 | Maulini et al. | Jul 2015 | B2 |
9608408 | Lyakh et al. | Mar 2017 | B2 |
9917418 | Razeghi | Mar 2018 | B2 |
10020635 | Lyakh | Jul 2018 | B1 |
10177535 | Lyakh | Jan 2019 | B1 |
20030042476 | Nakahara | Mar 2003 | A1 |
20090213890 | Patel et al. | Aug 2009 | A1 |
20130010823 | Maulini et al. | Jan 2013 | A1 |
20130121359 | Mansour et al. | May 2013 | A1 |
20140233596 | Fang | Aug 2014 | A1 |
20150103857 | Maulini et al. | Apr 2015 | A1 |
20150184994 | Flanders | Jul 2015 | A1 |
20150333472 | Demir | Nov 2015 | A1 |
20160322788 | Lyakh et al. | Nov 2016 | A1 |
20180048118 | Lyakh | Feb 2018 | A1 |
20180254603 | Patel et al. | Sep 2018 | A1 |
Entry |
---|
A. Lyakh et al., “High-Performance continuous-wave room temperature 4.0-um quantum cascade lasers with single-facet optical emission exceeding 2 W”, Proc Natl Acad Sci USA. Nov. 2, 2010; 107(44): 18799-18802 (Year: 2010). |
Lyakh et al. “High-Performance continuous-wave room temperature 4.0-1.Jm quantum cascade lasers with single-facet optical emission exceeding 2 W”, Proc Natl Acad Sci US A. Nov. 2, 2010; 107(44): 18799-18802 (Year: 2010) See U.S. Appl. No. 17/002,129. |
Dhirhe et al. “Quantum cascade lasers with an integrated polarization mode converter” Optics Express; Nov. 5, 2012 / vol. 20, No. 23 / 25711; pp. 7 See U.S. Appl. No. 17/002,129. |
Number | Date | Country | |
---|---|---|---|
20230163571 A1 | May 2023 | US |
Number | Date | Country | |
---|---|---|---|
62608761 | Dec 2017 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 17002129 | Aug 2020 | US |
Child | 17938817 | US | |
Parent | 16136713 | Sep 2018 | US |
Child | 17002129 | US |