HIGH-POWER SINGLE-MODE TRIPLE-RIDGE WAVEGUIDE SEMICONDUCTOR LASER

Abstract

To achieve high-power single transverse mode laser, we here propose a supersymmetry (SUSY)-based triple-ridge waveguide semiconductor laser structure, which is composed of an electrically pumped main broad-ridge waveguide located in the middle and a pair of lossy auxiliary partner waveguides. The auxiliary partner waveguides are designed to provide dissipative modes that can phase match and couple with the higher-order modes in the main waveguide. By appropriately manipulating the gain-loss discrimination of the modes in the laser cavity, one can effectively suppress all the undesired higher-order transverse modes while keeping the fundamental one almost unaffected, thereby ensuring stable single-mode operation with a larger emitting aperture and accordingly a higher output power than a conventional single-transverse-mode ridge waveguide diode laser.

Claims

1. A high-power single transverse mode laser, comprising: a triple-ridge waveguide (TRW) structure, having a main waveguide in between a pair of auxiliary waveguides on respective lateral sides of the main waveguide, wherein the main waveguide is relatively wider than either of the auxiliary waveguides on either side thereof, to support high-order modes of the main waveguide, while optical loss is introduced by the two auxiliary waveguides, which two auxiliary waveguides are respectively configured to support guided modes that only couple with the high-order modes of the main waveguide, to filter out the high-order modes in the main waveguide.

2. A high-power single transverse mode laser according to claim 1, wherein the pair of auxiliary waveguides each have propagation constants β which match with those of guided modes associated with the main waveguide other than the main waveguide fundamental mode.

3. A high-power single transverse mode laser according to claim 1, wherein the pair of auxiliary waveguides are respectively separated from the main waveguide therebetween by a trench having the same trench width d.

4. A high-power single transverse mode laser according to claim 3, wherein the trench width d is less than or equal to 2.0 μm.

5. A high-power single transverse mode laser according to claim 4, wherein the trench width d is in a range of from 1.35 μm to 0.9 μm.

6. A high-power single transverse mode laser according to claim 3, wherein the trench has a trench depth which is less than about 1000 nm.

7. A high-power single transverse mode laser according to claim 6, wherein the trench depth is about 900 nm.

8. A high-power single transverse mode laser according to claim 4, wherein the main waveguide has a width w.sub.M≤20.0 μm, and each of the pair of waveguides have widths w.sub.L (waveguide left width) and w.sub.R (waveguide right width) which are each ≤8.0 μm.

9. A high-power single transverse mode laser according to claim 8, wherein the trench width d is about 0.9 μm, the trench depth is about 900 nm, the main waveguide has a width w.sub.M about 12.0 μm, and each of the pair of waveguides have widths w.sub.L about 4.9 μm and w.sub.R about 7.25 μm, respectively.

10. A high-power single transverse mode laser according to claim 8, wherein the trench width d is about 1.1 μm, the trench depth is about 900 nm, the main waveguide has a width w.sub.M about 10.0 μm, and each of the pair of waveguides have widths w.sub.L about 3.95 μm and w.sub.R about 5.95 μm, respectively.

11. A high-power single transverse mode laser according to claim 8, wherein the effective index difference Δn.sub.eff between the ridge and the trench is about 3.5×10.sup.−3.

12. A high-power single transverse mode laser according to claim 6, wherein: the main waveguide has a width w.sub.M≤20.0 μm, and supports at least three transverse electric (TE) modes TEM 0, TEM 1, and TEM 2; and wherein the TE mode 0 for one of the auxiliary waveguides (TEL 0) and the TE mode 0 for the other of the auxiliary waveguides (TER 0) are respectively chosen to couple with the TEM 1 and the TEM 2 modes.

13. A high-power single transverse mode laser according to claim 1, wherein: the TRW structure includes an InGaAs/GaAs epitaxial wafer; and the laser operation wavelength λ is in a range of 0.95 μm to 1.05 μm.

14. A methodology for providing a high-power single transverse mode laser, comprising: providing a triple-ridge waveguide (TRW) structure, having a main waveguide in between a pair of auxiliary waveguides on respective lateral sides of the main waveguide, with the main waveguide relatively wider than either of the auxiliary waveguides on either side thereof, to support high-order modes of the main waveguide, and respectively configuring the two auxiliary waveguides to support guided modes that only couple with the high-order modes of the main waveguide, to introduce optical loss by the two auxiliary waveguides, which filters out the high-order modes in the main waveguide.

15. A methodology according to claim 14, wherein the pair of auxiliary waveguides each have propagation constants β which match with those of guided modes associated with the main waveguide other than the main waveguide fundamental mode.

16. A methodology according to claim 14, wherein the pair of auxiliary waveguides are respectively separated from the main waveguide therebetween by a trench having the same trench width d.

17. A methodology according to claim 16, wherein the trench width d is less than or equal to 2.0 μm.

18. A methodology according to claim 17, wherein the trench width d is in a range of from 1.35 μm to 0.9 μm.

19. A methodology according to claim 16, wherein the trench has a trench depth which is less than about 1000 nm.

20. A methodology according to claim 19, wherein the trench depth is about 900 nm.

21. A methodology according to claim 17, wherein the main waveguide has a width w.sub.M≤20.0 μm, and each of the pair of waveguides have widths w.sub.L (waveguide left width) and w.sub.R (waveguide right width) which are each ≤8.0 μm.

22. A methodology according to claim 21, wherein the trench width d is about 0.9 μm, the trench depth is about 900 nm, the main waveguide has a width w.sub.M about 12.0 μm, and each of the pair of waveguides have widths w.sub.L about 4.9 μm and w.sub.R about 7.25 μm, respectively.

23. A methodology according to claim 21, wherein the trench width d is about 1.1 μm, the trench depth is about 900 nm, the main waveguide has a width w.sub.M about 10.0 μm, and each of the pair of waveguides have widths w.sub.L about 3.95 μm and w.sub.R about 5.95 μm, respectively.

24. A methodology according to claim 21, wherein the effective index difference Δn.sub.eff between the ridge and the trench is about 3.5×10.sup.−3.

25. A methodology according to claim 19, wherein: the main waveguide has a width w.sub.M≤20.0 μm, and supports at least three transverse electric (TE) modes TEM 0, TEM 1, and TEM 2; and wherein the TE mode 0 for one of the auxiliary waveguides (TEL 0) and the TE mode 0 for the other of the auxiliary waveguides (TER 0) are respectively chosen to couple with the TEM 1 and the TEM 2 modes.

26. A methodology according to claim 14, wherein: the TRW structure includes an InGaAs/GaAs epitaxial wafer; and the laser operation wavelength λ is in a range of 0.95 μm to 1.05 μm.

27. A methodology according to claim 14, wherein configuring the two auxiliary waveguides includes phase matching the two auxiliary waveguides with the higher-order modes in the main waveguide, such that the higher-order modes, except for the fundamental mode of the main waveguide, will split into symmetric (in-phase) and anti-symmetric (out-of-phase) supermode pairs, whereby the TRW structure effectively suppresses unwanted higher-order modes and ensures single-mode lasing in a relatively broader main ridge waveguide.

28. A methodology according to claim 14, further including optimizing the respective auxiliary waveguide widths and trench widths by performing parameter sweeping based on two waveguide coupling.

29. A methodology for providing an edge-emitting laser diode capable of high-power single-transverse-mode operation based on the principle of unbroken supersymmetry (SUSY), comprising: providing a triple-ridge waveguide (TRW) structure, having a main ridge waveguide in between a pair of lossy auxiliary ridge waveguides, with the main ridge relatively wider than either of the auxiliary ridges; and respectively configuring the two auxiliary waveguides to support guided modes that only couple with the higher-order modes of the main waveguide, to introduce optical loss by the two auxiliary waveguides, which suppresses unwanted higher-order modes in the main waveguide other than its fundamental mode.

30. A methodology according to claim 29, further including providing additional loss in the auxiliary waveguides using ion implantation during fabrication of the auxiliary waveguides.

31. A methodology according to claim 30, further including: providing secondary cross-sectional trenches respectively within the auxiliary waveguides; and providing absorbing layers associated with the auxiliary waveguides, wherein the absorbing layers comprise at least one of Germanium (Ge), Chromium (Cr), alloys of Ge and Cr, and other material with high absorption for the lasing wavelengths of the edge-emitting laser diode.

32. A high-power single transverse mode laser, comprising: a ridged waveguide structure, having a main waveguide and at least one auxiliary waveguide on a lateral side of the main waveguide, wherein the main waveguide is relatively wider than the at least one auxiliary waveguide, to support high-order modes of the main waveguide, and the at least one auxiliary waveguide is configured to support at least one guided mode which couples with a higher-order mode of the main waveguide other than its fundamental mode, to suppress at least one higher-order mode in the main waveguide.

33. A high-power single transverse mode laser according to claim 32, further comprising: another auxiliary waveguide situated on a lateral side of the main waveguide opposite to that of the at least one auxiliary waveguide, so that the main waveguide is situated in between the pair of the two auxiliary waveguides, wherein the main waveguide is relatively wider than the auxiliary waveguide, and the pair of auxiliary waveguides each have propagation constants β which match with those of guided modes associated with the main waveguide other than the main waveguide fundamental mode.

34. A high-power single transverse mode laser according to claim 33, further including implanted ions associated with the auxiliary waveguides for providing additional loss in the auxiliary waveguides.

35. A methodology according to claim 34, further including: secondary cross-sectional trenches respectively within the auxiliary waveguides; and absorbing layers associated with the auxiliary waveguides, wherein the absorbing layers comprise at least one of Germanium (Ge), Chromium (Cr), alloys of Ge and Cr, and other material with high absorption for the lasing wavelengths of the high-power single transverse mode laser.

Description

BRIEF DESCRIPTION OF THE FIGURES

[0029] A full and enabling disclosure of the presently disclosed subject matter, including the best mode thereof, directed to one of ordinary skill in the art, is set forth in the specification, which makes reference to the appended Figures, in which:

[0030] FIG. 1A depicts an example schematic diagram of an example currently disclosed triple-ridge waveguide (TRW) laser structure, representing fundamental principles of single-mode lasing in the currently proposed triple-ridge waveguide (TRW) laser;

[0031] FIG. 1B graphically depicts propagation indices of guided modes in a single ridge waveguide structure as a function of the ridge width;

[0032] FIG. 1C illustrates a cross-section of the presently disclosed FIG. 1A TRW laser structure, together with illustration of mode coupling mechanism;

[0033] FIG. 2A on the left side illustrates electric field component Ex distributions of the TE modes supported by the isolated RW structures, while the right side thereof illustrates Ex distributions of the TE modes in the TRW structure sorted from the lowest-order (top) to the highest-order (bottom) mode (and with the geometric parameters comprising w.sub.L=3.95 μm, w.sub.M=10 μm, w.sub.R=5.9 5 μm, and d=1.1 μm), all related to mode analysis for TRW1 laser where the main waveguide width is equal to 10 μm;

[0034] FIG. 2B, on the left side illustrates the sweeping result for the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.1.sup.M and TE.sub.0.sup.L as a function of the left waveguide width w.sub.L and trench width, while the right side thereof illustrates the sweeping result for coupling between TE.sub.2.sup.M and TE.sub.1.sup.R as a function of the right waveguide width w.sub.R and trench width d, again all still related to mode analysis for TRW1 laser where the main waveguide width is equal to 10 μm;

[0035] FIG. 2C illustrates the net modal gain of the modes in the conventional RW laser (solid circles) and the modes in the TRW laser structure (hollow triangles), again for the case of the TRW1 laser where the main waveguide width is equal to 10 μm;

[0036] FIG. 3A on the left side illustrates electric field component Ex distributions of the TE modes supported by the isolated RW structures, while the right side thereof illustrates Ex distributions of the TE modes in the TRW structure sorted from the lowest-order (top) to the highest-order (bottom) mode (and with the geometric parameters comprising w.sub.L=4.9 μm, w.sub.M=12 μm, w.sub.R=7.25 μm, and d=0.9 μm), all related to mode analysis for TRW2 laser where the main waveguide width is equal to 12 μm;

[0037] FIG. 3B, on the left side illustrates the sweeping result for the smaller modal discrimination of the supermodes formed by the coupling between i). TE.sub.1.sup.M and TE.sub.0.sup.L modes; ii). TE.sub.3.sup.M and TE.sub.1.sup.L modes; iii). TE.sub.2.sup.M and TE.sub.1.sup.R modes, and iv). TE.sub.3.sup.M and TE.sub.2.sup.R modes, again all still related to mode analysis for TRW2 laser where the main waveguide width is equal to 12 μm;

[0038] FIG. 3C illustrates the net modal gain of the modes in the conventional RW laser (solid circles) and the modes in the TRW laser structure (hollow triangles), again for the case of the TRW2 laser where the main waveguide width is equal to 12 μm;

[0039] FIG. 4A illustrates the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.1.sup.M and TE.sub.0.sup.L while FIG. 4B illustrates the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.2.sup.M and TE.sub.1.sup.R, when the additional loss is assumed to be 15 cm.sup.−1; and

[0040] FIGS. 5A and 5B illustrate the effect of wavelength variation A and etching depth deviation B on the gain margin Δg for single-mode operation in the TRW1 design.

[0041] Repeat use of reference characters in the present specification and drawings is intended to represent the same or analogous features or elements or steps of the presently disclosed subject matter.

DETAILED DESCRIPTION

[0042] Reference now will be made in detail to embodiments, one or more examples of which are illustrated in the Figures. Each example is provided by way of explanation of the embodiments, not limitation of the present disclosure. In fact, it will be apparent to those skilled in the art that various modifications and variations can be made to the embodiments without departing from the scope or spirit of the present disclosure. For instance, features illustrated or described as part of one embodiment can be used with another embodiment to yield a still further embodiment. Thus, it is intended that aspects of the present disclosure cover such modifications and variations.

[0043] High-power semiconductor lasers that operate in the fundamental spatial mode are highly desired in numerous applications.sup.[1-4], including optical communication, pumping sources, laser surgery, and direct-diode material processing. Typically, an index-guided ridge waveguide (RW) structure is employed to maintain single spatial mode operation. Conventional single-mode RW lasers require a narrow lateral waveguide to suppress or cut off higher-order modes.sup.[5]. However, the small modal sizes and small emitting apertures give rise to high-power density and poor heat dissipation. Because of the restriction of the catastrophic optical mirror damage (COMD) and/or thermal rollover effect, the maximum available output power from a conventional RW laser is very limited.sup.[6]. To address these challenges, the width of the RW laser needs to be increased to scale up the single-mode output power, which requires a weak effective index contrast between the core and cladding. For example, relatively weak index waveguides are used in both Slab Coupled Optical Waveguide Lasers (SCOWLs).sup.[7,8] and large core lases based on strongly asymmetric epi-designs.sup.[9,10].

[0044] However, the weak index guide makes the waveguide structure susceptible to the carrier induced index suppression, thermal lensing and spatial hole burning effects. It tends to support multiple transverse modes at high drive currents, resulting in mode instabilities manifesting as kinks in the light-output (IL) curves, beam steering and beam-quality degradation.sup.[11]. Clearly, it would be desirable to introduce additional mechanisms in RW laser structure to avoid the onset of higher-order modes and simultaneously obtain high-power. Several methods have been investigated to realize mode-filtering in the lateral direction, such as angled gratings lasers.sup.[12,13] and anti-resonant reflective optical waveguide (ARROW) lasers.sup.[14]. Photonics crystal resonators are proposed in surface-emitting-type semiconductor lasers for the realization of single-mode operation with larger emission area.sup.[15-17]. In recent years, by harnessing notions from the selective breaking of parity-time (PT) symmetry, a novel PT symmetric design approach is established to achieve single-mode lasing in transversely multi-mode microring lasers.sup.[18-20]. To further scale up the output power, super symmetric coupled laser arrays based on the principle of unbroken supersymmetry (SUSY) are proposed, showing great promise in generating high-power diffraction-limited lasing output in strip waveguide arrays.sup.[21,22], microring resonator arrays.sup.[23], and most recently proposed, higher-dimensional microlaser arrays.sup.[24].

[0045] In light of the SUSY principle, we here propose a high-power single transverse mode laser by constructing a triple-ridge waveguide (TRW) structure, schematically shown in FIG. 1A. Instead of using the laser arrays in the previous SUSY arrangement, our proposed TRW structure consists of a main broad waveguide in the middle and a pair of auxiliary partner waveguides on both sides. We allow the main waveguide to support high-order modes so that the waveguide width is increased without reducing the index contrast. The pump current is injected in the main waveguide to provide optical gain, while extra optical loss is introduced in the two auxiliary partner waveguides. The auxiliary waveguides are designed to support guided modes that only couple with the high-order modes of the main waveguide, which effectively filters out these high-order modes in the main waveguide. By appropriately manipulating the modal discrimination of the modes in the TRW structure, one can effectively suppress all the undesired modes below the lasing threshold while keeping the fundamental one almost intact, thereby ensuring the single-mode lasing with a larger emitting aperture, and accordingly, a higher output power than a conventional RW laser. Compared with SUSY laser array, the proposed TRW is simpler in design, and thus, experimental implementation.

[0046] Additionally, avoiding etching in the main ridge waveguide is beneficial to achieving lower optical internal loss and better electrical characteristics.

[0047] Moreover, the proposed TRW structure does not require a small effective index difference, making it less susceptible to the carrier-induced index suppression, thus providing the stable single-mode operation at high injection currents.

[0048] Moreover, the proposed strategy has no specific requirements for the epitaxial structure and is compatible with the fabrication of a conventional RW laser, making it promising for low-cost and mass production. It is worth noting here that our proposed strategy is different from the reported evanescent spatial filtering in RW or broad area lasers.sup.[5, 25-27] which introduces selective losses for higher-order modes by applying the fact that higher-order modes experience greater spatial extent outside the guide (inside the lossy region). These methods do not require phase matching and thus provide relatively low modal discrimination. In addition, since it is difficult for them to keep the fundamental mode intact, the lasers usually show increased thresholds and decreased slopes. By virtue of the novel mode engineering in our proposed TRW structure, more efficient and stable mode filtering with much better modal discrimination can be realized in practice.

[0049] FIG. 1A depicts an example schematic diagram of an example currently disclosed triple-ridge waveguide (TRW) laser structure, representing fundamental principles of single-mode lasing in the currently proposed triple-ridge waveguide (TRW) laser. FIG. 1B graphically depicts propagation indices of guided modes in a single-ridge waveguide structure as a function of the ridge width. FIG. 1C illustrates a cross-section of the presently disclosed FIG. 1A TRW laser structure, together with illustration of mode coupling mechanism.

[0050] Based on the SUSY quantum field theory in one-dimensional (1D) Schrödinger problems, the unbroken SUSY regime refers to the case when all eigenvalues of the primary infinite potential well are exactly matched to those of its superpartner except for the ground state. The notions of SUSY transformations can be adapted from quantum physics to photonics by exploiting the mathematical isomorphism between the Schrödinger and the Helmholtz equation. In this context, the refractive index distribution plays the role of the potential. The optical guided mode profiles and the corresponding propagation constants are respectively analogous to the eigenfunction and eigenvalues.

[0051] As shown in the schematic and cross-section views of the proposed TRW structure (FIGS. 1A and C), the primary (main) waveguide is formed by a broad ridge which supports three transverse electric (TE) modes. To implement the unbroken optical SUSY framework, we construct a lossy superpartner (two lossy auxiliary partner waveguides in the TRW structure) with propagation constant values that match those of guided modes associated with the main waveguide apart from the fundamental mode.

[0052] To roughly obtain the ridge widths of the two auxiliary waveguides, we sweep the propagation constants of TE modes versus the ridge width in a single ridge waveguide, as shown in FIG. 1B. The propagation constants β are generally complex in a laser cavity. In this Figure, only the real parts of propagation constant are considered. As indicated by the vertical dotted line in FIG. 1B, the main waveguide supports three modes, TE.sub.0.sup.M, TE.sub.1.sup.M and TE.sub.2.sup.M. As indicated by two horizontal dotted lines, the TE.sub.1.sup.M and TE.sub.2.sup.M modes can respectively find a match in {β} with the narrower ridge waveguides. In a weakly coupled waveguide system, according to the coupling mode theory, the coupling strength between two modes is determined by the coupling coefficient, phase detuning (difference in {β}) and gain-loss contrast. Efficient coupling will split the coupled modes into symmetric (in-phase) and anti-symmetric (out-of-phase) supermode pairs. Therefore, by engineering the two lossy narrow auxiliary waveguides to phase match with the higher-order modes in the main waveguide, all the higher-order modes will split into supermode pairs except for the fundamental mode.

[0053] Since extra optical loss has been added in the auxiliary waveguides, the supermodes will experience higher modal loss and accordingly higher lasing thresholds. In this regard, we should appropriately control the coupling coefficient in the TRW laser structure in such a way that the amplitude distributions of the supermodes have a large spatial overlap with the lossy region. On the other hand, the fundamental mode TE.sub.0.sup.M in the main waveguide, which is decoupled with the lossy auxiliary waveguides, will hold the lowest threshold in the cavity and becomes the lasing mode in the mode competition. Therefore, the proposed TRW structure can effectively suppress the unwanted higher-order modes and ensure the single-mode lasing in a broad ridge waveguide. Note, the main waveguide in the TRW structure is not restricted to support only three modes.

[0054] The TRW laser design in this paper uses a conventional InGaAs/GaAs epitaxial wafer, which is comprised of a compressively strained InGaAs single quantum well (SQW), located in a symmetric GaAs waveguide layer with total thickness of 800 nm, 1360 nm-thick Al.sub.0.25Ga.sub.0.75. As n-cladding, 940 nm-thick Al.sub.0.25Ga.sub.0.75. As p-cladding, and 50 nm-thick GaAs p-contact layer on the top. The thicknesses of insulation layer (SiO.sub.2) and p metal contact are assumed to be 200 nm and 300 nm, respectively. The peak photo-luminescence wavelength of the InGaAs SQW is around 1 um, and thus, the operation wavelength λ in the following simulation is assumed to be this value.

[0055] The widths of the three ridges are respectively denoted as w.sub.L (left), w.sub.M (main) and w.sub.R (right), as depicted in FIG. 1C. The two trench widths between the three ridges are set equal, and thus both denoted as d. The Finite Difference Eigenmode (FDE) solver in Lumerical is used to perform the mode analysis. The semiconductor materials are modeled with a complex refractive index n=n.sub.r+i.Math.n.sub.i, where the real index n.sub.r is calculated by the material composition. The imaginary part n.sub.i=−gλ/4π, where g represents the gain (positive) or loss (negative). The value of background internal loss depends on the epitaxial wafer used, which is not related with the calculated modal discrimination. The enhancement of modal discrimination in the TRW laser structure can be realized by either applying selective mode pumping or introducing additional optical loss in the auxiliary waveguides. We follow the latter approach in this work. To reflect this in the simulation model, a uniform gain is applied to the entire structure and a uniform loss of 20 cm.sup.−1 is assumed in the auxiliary waveguides. Since we mainly focus on illustrating the operation principle of the TRW laser structure, the method to introduce extra loss will not be discussed in detail. General approaches include ion implantation, selective area regrowth, depositing metallic films, etc. In addition, the charge transport and thermal effects in diode lasers are ignored in this manuscript. But it is worth mentioning that the thermal lens effect under continuous wave (CW) mode can be well addressed in the TRW structure by predesign adjustment and proper heat control.

[0056] FIG. 1 illustrates the operation principle of the proposed TRW laser. As shown in the schematic and cross-section view of the TRW structure (FIGS. 1A and C), the main waveguide is formed by a broad ridge which supports TE modes, and gain is provided in this region by current injection. In order to achieve stable single-mode lasing, the two higher-order modes must be suppressed. FIG. 1B shows the calculation results of the propagation constants of TE modes versus the ridge of width in a single ridge waveguide.

[0057] FIG. 1 shows fundamental principle of single-mode lasing in the proposed TRW laser. FIG. 1A shows the schematic of the TRW laser structure; FIG. 1B shows the propagation indices of guided modes in a single ridge waveguide structure as a function of the ridge width; and FIG. 1C shows the cross-section of the TRW laser structure with the illustration of mode coupling mechanism.

[0058] The propagation constants β are generally complex in a laser cavity. In this Figure, only the real parts of propagation constant are considered. As indicated by the vertical dotted line in FIG. 1B, the waveguide supports three modes TE.sub.0.sup.M, TE.sub.1.sup.M and TE.sub.2.sup.M. As indicated by two horizontal dotted lines, the TE.sub.1.sup.M and TE.sub.2.sup.M modes can respectively find a match in {β} with the narrower ridge waveguides. Therefore, we construct the TRW structure with two lossy narrow auxiliary waveguides engineered to be phase-matched with the higher-order modes in the main waveguide. The loss in the auxiliary waveguides can be introduced easily by blocking the electrical pumping or applying ion implantation. In a weakly coupled waveguide system, according to the coupling mode theory, the coupling strength between two modes is determined by the coupling coefficient, phase detuning (difference in {β}), and gain-loss contrast. Efficient coupling will split the coupled modes into symmetric (in-phase) and anti-symmetric (out-of-phase) supermode pairs. Since strong loss has been added in the auxiliary waveguides, the supermodes (formed due to the coupling between high-order modes in the main waveguide and guided modes in the auxiliary waveguides) will experience higher modal loss, and accordingly, higher lasing thresholds. In this regard, we should appropriately control the coupling coefficient in the TRW laser structure in such a way that the amplitude distributions of the supermodes have a large spatial overlap with the lossy region. On the other hand, the fundamental mode TE.sub.0.sup.m in the main waveguide, which is decoupled with the lossy auxiliary waveguides, will hold the lowest threshold in the cavity and becomes the lasing mode in the mode competition. Therefore, the proposed TRW structure can effectively suppress the unwanted higher-order modes and ensure the single-mode lasing in a broad ridge waveguide. It is worth mentioning that the main waveguide in the TRW structure is not restricted to support only three modes. In the subsequent paragraphs, two specific designs will be presented to better demonstrate the proposed scheme and elucidate the underlying mechanism and design essentials.

[0059] We first consider a TRW laser design with the main waveguide supporting three modes. FIG. 2A, on the left side, illustrates electric field component Ex distributions of the TE modes supported by the isolated RW structures, while the right side thereof, illustrates Ex distributions of the TE modes in the TRW structure sorted from the lowest-order (top) to the highest-order (bottom) mode (and with the geometric parameters comprising w.sub.L=3.95 μm, w.sub.M=10 μm, w.sub.R=5.95 μm, and d=1.1 μm), all related to mode analysis for TRW1 laser where the main waveguide width is equal to 10 μm. FIG. 2B, on the left side, illustrates the sweeping result for the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.1.sup.M and TE.sub.0.sup.L as a function of the left waveguide width w.sub.L and trench width, while the right side thereof illustrates the sweeping result for coupling between TE.sub.2.sup.M and TE.sub.1.sup.R as a function of the right waveguide width w.sub.R and trench width d, again all still related to mode analysis for TRW1 laser where the main waveguide width is equal to 10 μm. FIG. 2C illustrates the net modal gain of the modes in the conventional RW laser (solid circles) and the modes in the TRW laser structure (hollow triangles), again for the case of the TRW1 laser where the main waveguide width is equal to 10 μm.

[0060] A conventional InGaAs/GaAs epitaxial wafer is employed in this paper. It is comprised of a compressively strained InGaAs single quantum well (SQW), located in a symmetric GaAs waveguide layer with total thickness of 800 nm, 1360 nm-thick Al.sub.0.25Ga.sub.0.75. As n-cladding, 940 nm-thick Al.sub.0.25Ga.sub.0.75As p-cladding and 50 nm-thick GaAs p-contact layer on the top. The peak photo-luminescence wavelength of the InGaAs SQW is around 1 μm, and thus, the operation wavelength λ in the following simulation is assumed to be this value. The widths of the three ridges are respectively denoted as w.sub.L(left), w.sub.M (main) and w.sub.R (right), as depicted in FIG. 1C.

[0061] The two trench widths between the three ridges are set equal, and thus, both denoted as d. The Finite-Difference Eigenmode (FDE) solver in Lumerical is used to perform the mode analysis. The semiconductor materials are modeled with a complex refractive index n=n.sub.r+i.Math.n.sub.i, where the real index n.sub.r is calculated by the material composition. The imaginary part n.sub.i=−gλ/4π, where g represents the gain (positive) or loss (negative). The background internal loss is assumed to be 2 cm.sup.−1. An additional loss of −20 cm.sup.−1 is assumed in the lossy regions to represent the quantum adsorption and other introduced losses. In the simulation, only TE modes are considered because the gain for a compressively strained QW is much higher for the TE polarization.sup.[16].

[0062] In this design, the etching depth is set to be 900 nm, and accordingly, the effective index difference Δn.sub.eff between the ridge and trench is about 3.5×10.sup.−3. This relatively large built-in Δn.sub.eff can effectively suppress the carrier induced anti-waveguide effects. At this etching depth, the isolated main RW with w.sub.M=10 μm supports three TE modes, as shown in FIG. 2A. As depicted in the schematic diagram of FIG. 1C, the TE.sub.0.sup.L and the TE.sub.0.sup.R modes are respectively chosen to couple with the TE.sub.1.sup.M and the TE.sub.2.sup.M modes. It can be found in FIG. 1B that the bottom horizontal dotted line (associated with the TE.sub.2.sup.M mode) also has an intersection with the TE1 dispersion curve at a relatively larger width, which can also be utilized to couple with the TE.sub.2.sup.M mode. Considering that the higher-order modes are less confined and thus typically have larger coupling coefficients than lower order ones, TE.sub.1.sup.R mode is chosen to couple with TE.sub.2.sup.M in the design instead of TE.sub.0.sup.R.

[0063] In order to find the optimum auxiliary waveguide widths and trench widths, parameter sweeping based on two waveguide coupling is performed. The left panel of FIG. 2B plots the sweeping result for the modal discrimination of the two coupled waveguides (the main waveguide and left side waveguide), where the modal discrimination is defined as the modal loss difference between the lowest-loss fundamental mode and the second lowest loss high-order mode. It can be found that a larger trench width indicates a higher modal discrimination at the optimum auxiliary waveguide width but with a smaller margin of error. In practice, it is better to have a larger parameter margin to compensate the fabrication errors, and therefore, a moderate trench width of 1.1 μm is chosen in this design. The right panel of FIG. 2B shows the sweeping result for the coupling between the main waveguide and right waveguide. Owing to the larger coupling coefficient, a larger bright area, i.e., larger margin of error, is clearly observed. The triangle in FIG. 2B indicates the final selected values for the waveguide and trench widths.

[0064] FIG. 2C shows the comparison of the net modal gain of the modes in the designed TRW laser structure (triangles) with those in a conventional RW laser (dots). Clearly, the modal discriminations in the conventional RW structure are very small. All the three modes are easy to reach their thresholds, leading to the multimode lasing and degraded beam quality. After introducing two lossy auxiliary waveguides, apart from the fundamental mode, the higher-order modes couple with lossy modes and split into supermodes with much enhanced losses. With a large modal discrimination, the stable single-mode lasing can be obtained. In the TRW laser, as indicated in the FIG. 2C, a high modal discrimination Δg=9.8 cm.sup.−1 is achieved for the single-mode operation. Moreover, it is found that the loss of fundamental mode in the TRW structure is only 0.39 cm.sup.−1 larger than that of the conventional structure, implying that the proposed scheme barely introduces additional loss for the fundamental lasing mode.

[0065] We first consider a TRW laser design with the main waveguide supporting three modes. In the simulation, only TE modes are considered because the gain for a compressively strained QW is much higher for the TE polarization.sup.[28]. In this design, the etching depth is set to be 900 nm, and accordingly, the effective index difference Δn.sub.eff between the ridge and trench is about 3.5×10.sup.−3. This relatively large built-in Δn.sub.eff can effectively suppress the carrier induced anti-waveguide effects. At this etching depth, the isolated main RW with w.sub.M=10 um supports three TE modes, as shown in FIG. 2A.

[0066] As depicted in the schematic diagram of FIG. 1C, the TEL 0 and the TER 0 modes are respectively chosen to couple with the TEM 1 and the TEM 2 modes. It can be found in FIG. 1B that the bottom horizontal dotted line (associated with the TEM 2 mode) also has an intersection with the TE1 dispersion curve at a relatively larger width, which can also be utilized to couple with the TEM 2 mode. Considering that the higher-order modes are less confined and thus typically have larger coupling coefficients than lower order ones, TER 1 mode is chosen to couple with TEM 2 in the design instead of TER 0.

[0067] In order to find the optimum auxiliary waveguide widths and trench widths, parameter sweeping based on two waveguide coupling is performed. The left panel of FIG. 2B plots the sweeping result for the modal discrimination of the two coupled waveguides (the main waveguide and left side waveguide), where the modal discrimination is defined as the modal loss difference between the lowest-loss fundamental mode and the second lowest loss high-order mode. It can be found that a larger trench width indicates a higher modal discrimination at the optimum auxiliary waveguide width but with a smaller margin of error. In practice, it is better to have a larger parameter margin to compensate the fabrication errors, and therefore, a moderate trench width of 1.1 um is chosen in this design. The right panel of FIG. 2B shows the sweeping result for the coupling between the main waveguide and right waveguide. Owing to the larger coupling coefficient, a larger bright area, i.e., larger margin of error, is clearly observed. The triangle in FIG. 2B indicates the final selected values for the waveguide and trench widths.

[0068] FIG. 2C shows the comparison of the net modal gain of the modes in the designed TRW laser structure (triangles) with those in a conventional RW laser (dots). Clearly, the modal discriminations in the conventional RW structure are very small. All the three modes are easy to reach their thresholds, leading to the multimode lasing and degraded beam quality. After introducing two lossy auxiliary waveguides, apart from the fundamental mode, the higher-order modes couple with lossy modes and split into supermodes with much enhanced losses. With a large modal discrimination, the stable single-mode lasing can be obtained. In the TRW laser, as indicated in the FIG. 2C, a high modal discrimination Δg=9.8 cm.sup.−1 is achieved for the single-mode operation.

[0069] As mentioned before, the proposed TRW structure is capable of filtering out more than two higher-order modes. In what follows, a TRW2 laser design with w.sub.M=12 μm is investigated.

[0070] FIG. 3A on the left side illustrates electric field component Ex distributions of the TE modes supported by the isolated RW structures, while the right side thereof illustrates Ex distributions of the TE modes in the TRW structure sorted from the lowest-order (top) to the highest-order (bottom) mode (and with the geometric parameters comprising w.sub.L=4.9 μm, w.sub.M=12 μm, w.sub.R=7.25 μm, and d=0.9 μm), all related to mode analysis for TRW2 laser where the main waveguide width is equal to 12 μm. FIG. 3B, on the left side illustrates the sweeping result for the smaller modal discrimination of the supermodes formed by the coupling between i). TE.sub.1.sup.M and T.sub.0.sup.L modes; ii). TE.sub.3.sup.M and TE.sub.1.sup.L modes; iii). TE.sub.2.sup.M and TE.sub.1.sup.R modes, and iv). TE.sub.3.sup.M and TE.sub.2.sup.R modes, again all still related to mode analysis for TRW2 laser where the main waveguide width is equal to 12 μm. FIG. 3C illustrates the net modal gain of the modes in the conventional RW laser (solid circles) and the modes in the TRW laser structure (hollow triangles), again for the case of the TRW2 laser where the main waveguide width is equal to 12 μm.

[0071] The etching depth is also assumed to be 900 nm. In this case, the isolated main waveguide supports four modes, denoted as TE.sub.0.sup.M to TE.sub.3.sup.M, as shown in FIG. 3A. Same as the first design, a TE.sub.0.sup.L and a TE.sub.1.sup.R dissipative mode are utilized in the second design to couple with the TE.sub.1.sup.M and the TE.sub.2.sup.M mode, respectively.

[0072] The panel i in FIG. 3B shows the modal discrimination of the supermodes formed by the TE.sub.1.sup.M-TE.sub.0.sup.L coupling as a function of the auxiliary waveguide and trench widths. We can see that the modal discrimination drops sharply at a larger trench width around 1.35 μm, indicating a significant decrease in the coupling efficiency. This happens because the gain-loss difference is getting same as the coupling coefficient and thus results in a very weak effective coupling. A continual increase in the trench width will drive it into the so-called parity-time (PT) symmetry broken region.sup.[22, 29], i.e., no coupling at all. It can be found that the critical trench width where PT-breaking occurs in this design is less than that in the TRW1 design because the guided modes are better confined in a broader waveguide, leading to a smaller coupling coefficient at the same trench width. Therefore, a smaller trench width of 0.9 μm is selected in this design, and accordingly, w.sub.L is determined to be 4.9 μm. In addition to the TE.sub.0.sup.L mode, this left auxiliary waveguide also supports a higher-order mode TE.sub.1.sup.L, whose real propagation constant is close to that of the TE.sub.3.sup.M mode. Panel ii plots the parameter sweeping result for the TE.sub.3.sup.M-TE.sub.1.sup.L coupling. There is a very large parameter variation margin due to strong mode coupling. This is because both TE.sub.3.sup.M and TE.sub.1.sup.L have a long evanescent tail extending outside waveguide. The sweeping result for TE.sub.2.sup.M-TE.sub.1.sup.R coupling is plotted in Panel iii with the red triangle indicating the selected w.sub.R and d values. Despite that the highest order mode TE.sub.2.sup.R supported by the right waveguide is far from the phase matching with TE.sub.3.sup.M, it helps to enhance the losses of the supermodes formed by TE.sub.3.sup.M and TE.sub.1.sup.L. The out-of-phase TE.sub.3.sup.M and TE.sub.1.sup.L supermode has a smaller {β}, which is close to that of TE.sub.2.sup.R, and thus shows a higher loss than the in-phase supermode.

[0073] The net modal gain of the modes of TRW2 is shown as the hollow triangles in FIG. 3C. Compared with the conventional RW (solid circles), the modal discrimination Δg for the single-mode lasing is increased by 9.5 cm.sup.−1.

[0074] FIGS. 5A and 5B illustrate the effect of wavelength variation A and etching depth deviation B on the gain margin Δg for single-mode operation in the TRW1 design.

[0075] The lasing wavelength is determined by the longitudinal modes in the cavity. Since the semiconductor quantum well have a broadband gain spectrum, the lasing wavelength might vary among different cavities. FIG. 5A depicts the effect of the wavelength variation on the modal discrimination Δg. It can be seen that Δg is only changed by 0.5 cm.sup.−1 as the wavelength increases by 60 nm, demonstrating the robustness against the wavelength variation.

[0076] Another parameter that affects the modal discrimination in practice is the etching depth. The effect of etching depth deviation on modal discrimination is shown in FIG. 5B. As the etching depth varies ±30 nm, Δg changes about 2.2 cm.sup.−1 in the TRW1 design while still exhibiting a large modal discrimination for the stable single-mode operation.

[0077] In summary, we have theoretically introduced a novel TRW structure in edge-emitting laser diodes capable of high-power single-transverse-mode operation. A pair of lossy auxiliary waveguides is equipped in the TRW laser to suppress the unwanted higher-order modes in the laser cavity. In this work, we demonstrate two TRW laser designs, the main broad waveguides of which respectively support two and three higher-order modes besides the fundamental mode. By carefully engineering the TRW structural parameters, both designs achieve large modal discrimination for stable single-mode lasing. It is also shown that the proposed TRW laser is robust to deviations of lasing wavelength and structural parameter, and therefore, very favorable for high yield and mass production.

[0078] As mentioned before, the proposed TRW structure is capable of filtering out more than two higher-order modes. In what follows, a TRW2 laser design with w.sub.M=12 um is investigated. The etching depth is also assumed to be 900 nm. In this case, the isolated main waveguide supports four modes, denoted as TE.sub.0.sup.M to TE.sub.3.sup.M, as shown in FIG. 3A. Same as the first design, a TE.sub.0.sup.L and a TE.sub.1.sup.R dissipative mode are utilized in the second design to couple with the TE.sub.1.sup.M and the TE.sub.2.sup.M mode, respectively. The panel i in FIG. 3B shows the modal discrimination of the supermodes formed by the TE.sub.1.sup.M-TE.sub.0.sup.L coupling as a function of the auxiliary waveguide and trench widths. We can see that the modal discrimination drops sharply at a larger trench width around 1.35 um, indicating a significant decrease in the coupling efficiency. This happens because the gain-loss difference is getting same as the coupling coefficient, thus resulting in a very weak effective coupling. A continual increase in the trench width will drive it into the so-called parity-time (PT) symmetry broken region.sup.[29], i.e., no coupling at all. It can be found that the critical trench width where PT-breaking occurs in this design is less than that in the TRW1 design because the guided modes are better confined in a broader waveguide, leading to a smaller coupling coefficient at the same trench width. Therefore, a smaller trench width of 0.9 um is selected in this design, and accordingly, w.sub.L is determined to be 4.9 um. In addition to the TE.sub.0.sup.L mode, this left auxiliary waveguide also supports a higher-order mode TE.sub.1.sup.L, whose real propagation constant is close to that of the TE.sub.3.sup.M mode. Panel ii plots the parameter sweeping result for the TE.sub.3.sup.M-TE.sub.1.sup.L coupling. There is a very large parameter variation margin due to strong mode coupling because both TE.sub.3.sup.M and TE.sub.1.sup.L have a long evanescent tail extending outside waveguide. The sweeping result for TE.sub.2.sup.M-TE.sub.1.sup.R coupling is plotted in Panel iii with the triangle indicating the selected w.sub.R and d values. Despite that the highest order mode TE.sub.2.sup.R supported by the right waveguide is far from the phase matching with TE.sub.3.sup.M, it helps to enhance the losses of the supermodes formed by TE.sub.3.sup.M and TE.sub.1.sup.L. The out-of-phase TE.sub.3.sup.M and TE.sub.1.sup.L supermode has a smaller {β}, which is close to that of TE.sub.2.sup.R, and thus shows a higher loss than the in-phase supermode. The net modal gain of the modes of TRW2 is shown as the hollow triangles in FIG. 3C. Compared with the conventional RW (solid circles), the modal discrimination Δg for the single-mode lasing is increased by 9.5 cm.sup.−1.

[0079] FIG. 4A illustrates the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.1.sup.M and TE.sub.0.sup.L, while FIG. 4B illustrates the smaller modal discrimination of the supermodes formed by the coupling between TE.sub.2.sup.M and TE.sub.1.sup.R, when the additional loss is assumed to be 15 cm.sup.−1.

[0080] The lasing wavelength is determined by the longitudinal modes in the cavity. Since the semiconductor quantum well has a broadband gain spectrum, the lasing wavelength might vary among different cavities. FIG. 4A depicts the effect of the wavelength variation on the modal discrimination Δg. It can be seen that Δg is only changed by 0.5 cm.sup.−1 as the wavelength increases by 60 nm, demonstrating the robustness against the wavelength variation. Another parameter that affects the modal discrimination in practice is the etching depth. The effect of etching depth deviation on modal discrimination is shown in FIG. 4B. As the etching depth varies ±30 nm, Δg changes about 2.2 cm.sup.−1 in the TRW1 design while still exhibiting a large modal discrimination for the stable single-mode operation.

[0081] FIG. 5 indicates the effect of wavelength variation A and etching depth deviation B on the gain margin Δg for single-mode operation in the TRW1 design.

[0082] In summary, we have theoretically introduced a novel TRW structure in edge-emitting laser diodes capable of high-power single-transverse-mode operation based on the principle of unbroken SUSY. A pair of lossy auxiliary waveguides is equipped in the TRW laser to suppress the unwanted higher-order modes in the laser cavity. In this work, we demonstrate two TRW laser designs, the main broad waveguides of which respectively support two and three higher-order modes besides the fundamental mode. By carefully engineering the TRW structural parameters, both designs achieve large modal discrimination for stable single-mode lasing. It is also shown that the proposed TRW laser is robust to deviations of lasing wavelength and structural parameter, and therefore, very favorable for high yield and mass production.

[0083] One may also use ion implantation during fabrication as an additional way to introduce dopants/loss in the auxiliary waveguides. This may further include implanted ions associated with the auxiliary waveguides for providing additional loss in the auxiliary waveguides. Still further, one may include secondary cross-sectional trenches respectively within the auxiliary waveguides and absorbing layers around the trenches. Such absorbing layers may comprise at least one of Germanium (Ge), Chromium (Cr), alloys of Ge and Cr, and other material with high absorption for the lasing wavelengths of the high-power single transverse mode laser.

[0084] While the present subject matter has been described in detail with respect to specific example embodiments thereof, it will be appreciated that those skilled in the art, upon attaining an understanding of the foregoing may readily produce alterations to, variations of, and equivalents to such embodiments. Accordingly, the scope of the present disclosure is by way of example rather than by way of limitation, and the subject disclosure does not preclude inclusion of such modifications, variations and/or additions to the present subject matter as would be readily apparent to one of ordinary skill in the art.

REFERENCES

[0085] .sup.1U. Brauch, P. Loosen, and H. Opower, High-power Diode Lasers, 303-368 (2000). [0086] .sup.2F. Bachmann, Applied surface science 208, 125-136(2003). [0087] .sup.3W. Horn, Laser Technik Journal, 4(3), 62-65 (2007). [0088] .sup.4H. Nasim, and J. Yasir, Optics & Laser Technology 56, 211-222(2014). [0089] .sup.5H. Wenzel, F. Bugge, M. Dallmer, F. Dittmar, J. Fricke, K. H. Hasler, and G. Erbert, IEEE Photon. Technol. Lett. 20(3), 214-216 (2008). [0090] .sup.6M. Ziegler, J. Tomm, U. Zeimer, and T. Elsaesser, J. Electron. Mater., 39(6), 709-714 (2010). [0091] .sup.7J. Fan, L. Zhu, M. Dogan, and J. Jacob, Opt. Express 22(15), 17666-176722(2014). [0092] .sup.8Smith, G. M., R. K. Huang, J. P. Donnelly, L. J. Missaggia, M. K. Connors, G. W. Turner, and P. W. Juodawlkis, In 2010 23rd Annual Meeting of the IEEE Photonics Society, pp. 479-480. IEEE, 2010. [0093] .sup.9M. Wilkens, H. Wenzel, J. Fricke, A. Maaßdorf, P. Ressel, S. Strohmaier, A. Knigge, G. Erbert, and G. Tränkle, IEEE Photon. Technol. Lett. 30(6), 545-548 (2018). [0094] .sup.10M. Wilkens, G. Erbert, H. Wenzel, A. Knigge, P. Crump, A. Maaßdorf, J. Fricke, P. Ressel, S. Strohmaier, B. Schmidt, and G. Tränkle, In High-power Diode Laser Technology XVI 2018 (International Society for Optics and Photonics, 2018) p. 105140E. [0095] .sup.11H. Wenzel, M. Dallmer, and G. Erbert, Optical and quantum electronics, 40(5), 379-384 (2008). [0096] .sup.12R. E. Bartolo, W. W. Bewley, I. Vurgaftman, C. L. Felix, J. R. Meyer, and M. J. Yang, Appl. Phys. Lett. 76(22), 3164-3166 (2000). [0097] .sup.13L. Zhu, P. Chak, J. K. Poon, G. A. DeRose, A. Yariv, and A. Scherer, Opt. Express, 15(10), 5966-5975 (2007). [0098] .sup.14M. Kanskar, M. Nesnidal, S. Meassick, A. Goulakov, E. Stiers, Z. Dai, T. E. Earles, D. Forbes, D. Hansen, P. Corbett, and L. Zhang, International Society for Optics and Photonics, 4995, 196-208 (2003). [0099] .sup.15K. Hirose, Y. Liang, Y. Kurosaka, A. Watanabe, T. Sugiyama, and S. Noda, Nat. Photonics 8(5), 406-411(2014). [0100] .sup.16M. Yoshida, M. De Zoysa, K. Ishizaki, Y. Tanaka, M. Kawasaki, R. Hatsuda, B. Song, J. Gelleta, and S. Noda, Nat. Mater. 18(2), 121-128 (2019). [0101] .sup.17A. Y. Song, A. R. K. Kalapala, R. Gibson, K. J. Reilly, T. Rotter, S. Addamane, H. Wang, C. Guo, G. Balakrishnan, R. Bedford, W Zhou, and S. Fan, Appl. Phys. Lett., 119, 031105 (2021). [0102] .sup.18Hodaei, Hossein, Mohammad-Ali Miri, Matthias Heinrich, Demetrios N. Christodoulides, and Mercedeh Khajavikhan, Science 346(6212), 975-978 (2014). [0103] .sup.19H. Hodaei, M. Miri, A. U. Hassan, W. E. Hayenga, M. Heinrich, D. N. Christodoulides, and M. Khajavikhan, Laser Photonics Rev. 10(3), 494-499 (2016). [0104] .sup.20W E. Hayenga, H. Garcia-Gracia, E. S. Cristobal, M. Parto, H. Hodaei, P. LiKamWa, D. N. Christodoulides, and M. Khajavikhan, Proc. IEEE 108(5), 827-836 (2019). [0105] .sup.21R. El-Ganainy, L. Ge, M. Khajavikhan, and D. N. Christodoulides, Phys. Rev. A 92(3), 033818 (2015). [0106] .sup.22M. P. Hokmabadi, N. S. Nye, R. El-Ganainy, D. N. Christodoulides, and M. Khajavikhan, Science 363(6427), 623-626 (2019). [0107] .sup.23B. Midya, H. Zhao, X. Qiao, P. Miao, W. Walasik, Z. Zhang N. M. Litchinitser, and L. Feng, Photonics Res. 7(3), 363-367 (2019). [0108] .sup.24X. Qiao, B. Midya, Z. Gao, Z. Zhang, H. Zhao, T. Wu, J. Yim, R. Agarwal, N. M. Litchinitser, and L. Feng, Science 372(6540) 403-408 (2021). [0109] .sup.25M. Buda, H. H. Tan, L. Fu, L. Josyula, and C. Jagadish. IEEE Photon. Tech. Lett. 15(12), 1686-1688(2003). [0110] .sup.26J. P. Leidner, and J. R. Marciante, IEEE J. Quan. Electron. 48(10), 1269-1274(2012). [0111] .sup.27H. Wenzel, P. Crump, J. Fricke, P. Ressel, and G. Erbert, IEEE J. Quan. Electron. 49(12), 1102-1108(2013). [0112] .sup.28C. H. Gao, H. Y. Ong, W. J. Fan, and S. F. Yoon, Computational Materials Science 30, 6-302 (2004). [0113] .sup.29C. E. Rüter, K. G. Makris, R. El-Ganainy, D. N. Christodoulides, M. Segev, and D. Kip, Nat. Phys. 6(3), 192-195 (2010).