VCSEL with anisotropic curved mirror

Abstract

The present disclosure provides new and innovative VCSEL devices and systems. In an example, a VCSEL device includes an endoscope having a curved mirror surface of the VCSEL and an anisotropic radius of curvature (ROC) of the curved mirror surface, where the ROC includes four directions that are + x, + y,-x,-y, in at least one direction, the ROC is within a range of predefined ROC values that are greater than a cavity length of the VCSEL and less than a standard beam width (ROCUL), and in at least one other direction, the ROC is outside the range.

Description

Background Art

[0001] The present disclosure generally relates to light emitting elements and light emitting devices including vertical cavity surface emitting lasers (VCSELs). The VCSELs described herein may be vertical cavity surface emitting lasers or vertical external cavity surface emitting lasers. In a VCSEL, laser light is allowed to resonate between two light reflecting layers (usually distributed Bragg reflector (DBR) layers, also referred to as mirror layers). The VCSEL structure including the DBR layers, cavity length, aperture size and shape affects the shape and characteristics of the emitted laser beam. Summary of the invention

[0002] The present disclosure provides new and innovative VCSEL devices and systems. In an embodiment, a VCSEL cavity mirror includes a curved mirror surface of the VCSEL and a radius of curvature (ROC) of the curved mirror surface having an anisotropy, wherein the ROC includes four directions, the four directions being +x, +y, -x, -y, and in at least one direction, the ROC is within a range of predefined ROC values ​​greater than the cavity length of the VCSEL and less than a standard beam width (ROCUL), and in at least one other direction, the ROC is outside the range.

[0003] Additional features and advantages of the disclosed methods and apparatus are described in and will be apparent from the following detailed description and accompanying drawings. The features and advantages described herein are not all-inclusive, and in particular, many additional features and advantages will be apparent to one of ordinary skill in the art in view of the drawings and description. Furthermore, it should be noted that the language used in this specification is primarily selected for readability and instructional purposes and not for the purpose of limiting the scope of the inventive subject matter. BRIEF DESCRIPTION OF THE DRAWINGS

[0004] Figure 1 is a diagram of a curved mirror of a VCSEL according to an embodiment of the present disclosure.

[0005] Figure 2 is a cross-sectional view of a VCSEL showing embodiments of a radius of curvature according to various embodiments of the present disclosure.

[0006] Figure 3 is a flow chart of a method for forming a curved mirror according to an embodiment of the present disclosure.

[0007] Figure 4 FIG. 4 is a schematic diagram of a cross section of a VCSEL device in a first direction according to an embodiment of the present disclosure.

[0008] FIG. 5A is a schematic diagram of a cross section of a VCSEL device in a second direction according to the first exemplary embodiment of the present disclosure.

[0009] 5B is a schematic diagram of a cross section of the VCSEL device of FIG. 5A in a plan view according to the first exemplary embodiment of the present disclosure.

[0010] FIG. 6A is a schematic diagram of a cross section of a VCSEL device in a second direction according to a second exemplary embodiment of the present disclosure.

[0011] 6B is a schematic diagram of a cross section of the VCSEL device of FIG. 6A in a plan view according to a second exemplary embodiment of the present disclosure.

[0012] 7A-7D are schematic diagrams of cross-sections and corresponding plan views of various example VCSEL devices in the second direction according to example embodiments of the present disclosure.

[0013] Figure 8 is a diagram of a process for reducing the length to form a saddle point according to a second exemplary embodiment of the present disclosure.

[0014] 9A to 9C are cross-sectional views of a VCSEL showing examples of mirror footprints according to various examples of the present disclosure.

[0015] Fig.10 A schematic diagram of various apertures for a VCSEL device in plan view according to an exemplary embodiment of the present disclosure is provided.

[0016] Fig.11 is a schematic diagram of individual emitter arrays on an interconnected mirror according to an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0017] VCSELs may have one or more curved mirrors or DBRs that can be configured to focus or adjust laser beam characteristics. In VCSELs with curved mirrors, the beam characteristics are heavily influenced by the shape of the mirror, including the radius of curvature (ROC) of the mirror, which may also be referred to as a lens. Typically, the mirrors in VCSELs are circularly symmetric; however, anisotropy in the mirrors can have beneficial effects in controlling the mode of the VCSEL, such as achieving stable polarization.

[0018] In recent years, VCSELs with curved mirrors have been demonstrated to have good performance using III-nitride material systems. In addition, the ability to control the shape of the lateral modes by changing the mirror parameters has been shown by controlling the lateral modes with the mirror ROC. However, due to the symmetry of devices such as circularly symmetric curved mirrors and the symmetry of the crystal orientation, these devices lack polarization control. This lack of polarization preference can lead to several problems, such as: the polarization of the device switches with lifetime and current; has a low polarization ratio; the polarization ratio changes with lifetime and current; the polarization direction is random and must be measured individually for each device to be determined.

[0019] Although some limited work has been done on VCSELs with curved mirrors, there are three main approaches to controlling polarization in VCSELs: using materials with polarization-dependent gain, using gratings to induce polarization dependence on reflection, or breaking the symmetry of the cavity net gain / loss, such as etching the mirrors with elliptical patterns.

[0020] To obtain polarization dependence of the gain, this is often achieved through anisotropic strain in the active region. This strain can be structurally induced or intrinsic due to the crystal orientation. Exploiting the structure to induce strain can make processing very complex and often requires a complete redesign of the process. For materials such as III-nitrides, polarization can be induced by using a non-basal plane orientation. In this polar structure, anisotropic biaxial strain in the semi-polar and non-polar oriented wells leads to a splitting of the valence band such that the recombination preferentially produces a single polarization. However, this approach often requires the use of non-basal plane oriented substrates. These substrates are typically more expensive and of lower quality than conventional c-plane oriented substrates.

[0021] Gratings have become the typical method for controlling polarization in arsenide-based devices. Simply etching lines or depositing strips of dielectric on top of the DBR, the mirror reflectivity becomes polarization-dependent, thus promoting lasing of polarizations with higher reflectivity. However, the reflectivity is strongly affected by the size of the grating, so delicate process control is required. Even more problematic is that the shorter the wavelength of the design, the smaller the grating size. This means that for VCSELs emitting in the visible spectrum, electron beam lithography is usually required to form the grating, significantly increasing manufacturing complexity and time.

[0022] Finally, the mirrors or structures can be modified to introduce anisotropic losses in the cavity. This is accomplished by etching anisotropic shapes (e.g., ellipses) into the mirror structure. While this is a simple variation or addition to the standard fabrication process, it is very sensitive to misalignment and can also increase the losses seen by the desired lasing mode, degrading device performance.

[0023] The present disclosure solves the above problems by providing a VCSEL cavity with improved beam characteristics using curved mirrors with varying radii of curvature along different axes whose relationship to the radii of curvature is defined by electrical aperture parameters along those axes.

[0024] Typically, as discussed in this disclosure, the top of a lens or curved mirror is defined as the centroid of the optical mode intensity of the lasing mode, or the "highest" point of the lens, or can be considered as the point within 2 μm of the centroid of the electrical aperture.

[0025] When referring to an axis direction as discussed in this disclosure, the axis will be specified so that the minimum ROC of the lens is aligned with the x-axis or the y-axis. In other words, the direction of taking the highest second-order derivative of the lens profile at the "top of the lens" is defined as parallel to the x-axis or the y-axis. Therefore, the direction with the maximum curvature is set to x or y. For example, a lens with height z(x, y):

[0026] [Mathematical formula 1]

[0027] or or

[0028] Figure 1 A curved mirror 102 with different radii of curvature (ROC) 106 and a corresponding electrical aperture 104 are shown. In this case, the ROC -x is the minimum ROC 106 of the lens and is aligned parallel to the x-axis. When defining the sign of the ROC 106, a positive value corresponds to a convex lens 102a, an infinite or near-infinite value refers to a flat mirror 102b, and a negative value refers to a concave lens 102c, such as Figure 2 It should be understood that the curved mirror or lens (e.g., DBR) having the ROC 106 may have multiple layers ( Figure 2 ), as will be described and discussed below. In addition, as discussed herein, unless the convex, nearly flat, or concave shape of the curved mirror is specifically mentioned, the curved lens 102a, 102b, or 102c may be generally referred to as the curved mirror 102.

[0029] Figure 2 Different notation conventions are shown for the ROC 106, as described in the present disclosure, in relation to a cavity 208 and a planar reflector or mirror 210 (eg, a DBR) configured to emit a laser beam from a VCSEL. Figure 1 As shown, the curved mirror will have a ROC value {ROC x , ROC -x , ROC y , ROC -y}, so that the sign of the curvature of the mirror in some directions can be different. For example, the ROC 106 of a curved mirror can be convex in some directions and concave or flat in other directions (e.g., 102a in the x-direction and 102b or 102c in the y-direction).

[0030] In defining the difference in ROC values, a threshold called ROCUL is used in the present disclosure. In an exemplary embodiment, ROCUL is the ROC 106 for which 1 / e 2 The beam width is equal to the electrical aperture diameter of the device defined by the aperture at the vertical position in that direction. For a device with an aperture 104 defined by oxide, this would correspond to the beam waist in cavity 208 at the level of oxide 212. x , r -x , r y , r -y} aperture 104, with a corresponding ROCUL in each direction {ROCUL x ,ROCUL -x ,ROCUL y ,ROCUL -y Mathematically, in this embodiment, for the intensity distribution of a Gaussian beam in direction x, g(x) is given by:

[0031] [Mathematical formula 2]

[0032]

[0033] 1 / e 2 The strength is given as

[0034] [Mathematical formula 3]

[0035] x=2σ

[0036] Thus, ROCUL occurs when:

[0037] [Formula 4]

[0038] x = 2σ = radius 孔径 (r a )

[0039] For a Gaussian beam in cavity 208 with curved mirror 102 and plane mirror 210, the beam width σ at plane mirror 210 is given by:

[0040] [Formula 5]

[0041]

[0042] Where λ is the optical mode wavelength, n is the refractive index, L is the cavity length (between 210 and 102), and R is the radius of curvature of the curved mirror 102. Assuming that the aperture 104 is defined near the plane mirror 210, these two equations can be combined to give the following definition of ROCUL in the direction x:

[0043] [Mathematical formula 6]

[0044]

[0045] In another exemplary embodiment, ROCUL is a ROC 106 for which the 1 / e beamwidth in a direction is equal to the electrical aperture diameter of the device in that direction at the vertical position defined by the aperture. In this exemplary embodiment, the 1 / e intensity is given by

[0046] [Formula 7]

[0047]

[0048] The different relationships are obtained by the above mathematical relationship. For ROCUL, any other suitable value may be used, for example, for 1 / e to 1 / e 2 Thus, it should be appreciated that the above relationship may vary depending on the ROCUL set for a particular implementation, as the threshold set for ROCUL may depend on the structure at hand. Furthermore, ROCUL is set so that the ROC anisotropy is sufficiently different in different directions to produce the desired effect (e.g., polarization control) for a given structure.

[0049] As discussed herein in various embodiments below, the directions of the ROC 106 may be described using the following terms: the "S" direction corresponds to the direction in which loss is introduced to the mode, with values ​​in that direction belonging to the set S; and the "Q" direction corresponds to the direction in which the reflector is intended to act as a normal, with values ​​in that direction belonging to the set Q.

[0050] As will be appreciated by those skilled in the art, the aperture 104 is defined by lateral current confinement, and the size of the aperture 104 is determined by the effective width of the current flow introduced into the layer or gain region of greatest loss. Typically, the layer of greatest loss for this mode is the active region, however, this is not necessarily the case. Metal contact layers, implant layers, or oxide layers are all other examples of layers that may be used to define the aperture 104.

[0051] For the specific examples of reflector shape variations shown below, unless otherwise stated, the values ​​given may generally correspond to the VCSEL cavity 208 described herein, although this description is not intended to limit the scope of the present disclosure, but only to provide suitable and typical exemplary embodiments. The growth substrate is composed of GaN, and the substrate orientation corresponds to the c-direction, with a bevel cut within ±3°. The epitaxial structure is composed of n-type material, followed by an active region of 3±1 quantum wells with a peak gain of 450nm. There is a p-type layer above the active region. The current aperture 104 is formed by ion implantation of the p-type layer with B or Al ions. In order to define the shape and size of the aperture 104, photolithography can be used to define the pattern of an injection barrier layer such as a thick photoresist or metal. The barrier material can be placed around the aperture location to prevent the area from being implanted and electrically isolated. For embodiments where the aperture size and shape are not specified, a circular aperture 104 with a diameter of 4μm can be used. Then, another n-type region with a tunnel junction contact or a transparent conductive oxide such as ITO is used to grow or deposit the current spreading layer, respectively. Metal contacts are deposited on the current spreading layer and to the n-type layer to enable electrical injection through the hole 104. To form the optical cavity 208, the substrate is thinned and a flat mirror 210 is deposited on one side and a curved mirror 102 is placed on the other side. For example, the total cavity length can be 20 μm. The reflector 102 can be composed of alternating layers of Ta2O5 and SiO2 with a thickness of 1 / 4 the wavelength of light in the material, corresponding to a free space wavelength of 450 nm. These layers form a dielectric distributed Bragg reflector.

[0052] In an example, a curved lens or reflector 102 is formed by spin coating a photoactive polymer. The polymer is then photoexcited through a mask and developed to form a designed pattern according to the desired lens shape. The top of the reflector 102 is aligned with the center of the aperture 104. The sample is then heated to cause the polymer to reflow and form a curved shape (e.g., 102a). The curved pattern is then transferred to the GaN surface below the polymer by dry etching. Finally, a dielectric layer is deposited on the curved surface to form the reflector 102.

[0053] The different curvatures of the reflectors 102a, 102b, 102c can be achieved by changing the shape of the polymer and the fluid properties (such as viscosity or surface tension) before reflow. Therefore, choosing the right resin is often an important parameter in determining the optimal manufacturing conditions. As described herein, the long side of a rectangular resist pattern is generally referred to as the length of the pattern and the short side is referred to as the width of the pattern. For the case of describing an array, the same general process is used, although each reflector 102 defines multiple apertures. Figure 3A general process 300 is shown for forming the curved mirror 102 on a substrate 302 by forming a shape using a photoresist 304, which is etched according to typical processes to produce the desired shape of the curved mirror 102, as discussed above. Figure 3 In the embodiment shown, the mirror 102 has a nearly infinite ROC in the middle and a ROC of approximately 70 μm at both ends.

[0054] Figure 4 A schematic diagram of a cross section of a VCSEL device 400 is shown, wherein the curved mirror 102a has a symmetrical ROC 106 and is Q-concentrated in the cross section shown, meaning that it shows a direction that is not intended to introduce additional losses. For the following description of embodiments of the present invention, this direction will be parallel to the y-axis. The approximate thresholds of the edges of the resonant mode shapes are depicted by lines 402s and 402b (e.g., half-power beamwidth). In the illustrated example of device 400, the ROCUL is approximately 250 μm. In an exemplary embodiment, the cavity 208 can be a VCSEL integrated cavity structure or an external cavity structure. In an exemplary embodiment, the laser cavity 208 is made of GaN, InP, GaAs, or another semiconductor material. It should be understood that the same reference numerals may be omitted where the inclusions are unnecessary for the explanation or description of the structure.

[0055] FIG. 5A is a schematic diagram of a cross section of a VCSEL device 500 in a second direction (e.g., the x direction). FIG. 5B is a schematic diagram of a cross section of the VCSEL device of FIG. 5A in a plan view. In a first exemplary embodiment shown in FIGS. 5A and 5B , the reflector 102 b has a varying ROC 106 along the x direction and a constant ROC 106 along the y direction, wherein the profile 510 of the reflector shape is longer in the x direction (e.g., 150 μm) than in the y direction. The ROC 106 along the x direction approaches infinity in the middle of the reflector 102 b where the aperture 104 is located. In other words, the reflector 102 b is substantially flat along the x direction and curved along the y direction, such as would be described by a three-dimensional cylindrical shape. This curvature can be achieved by creating a rectangular pattern with the polymer prior to reflow so that the long direction has sufficient length relative to the short side so that the lens is flattened along the top. For example, using an n-type photoresist with a specific viscosity and height after spinning, a mirror 102b with a total length of 150um and a width of 50um allows the ROC 106 along the x-direction to be stabilized by placing the aperture 104 in the middle third of the mirror length. The aperture 104 is placed in a stable region, for example with its center at 75μm along the length of the mirror 102b. The ROC 106 along the y-direction can be set to a constant value of 70μm. The approximate thresholds for the edges of the resonant mode shapes are depicted by lines 502a and 502b (e.g., half-power beamwidth).

[0056] FIG. 6A is a schematic diagram of a cross section of a VCSEL device 600 in a second direction (e.g., the x direction). FIG. 6B is a schematic diagram of a cross section of the VCSEL device of FIG. 6A in a plan view. In a second exemplary embodiment as shown in FIG. 6A and FIG. 6B , the ROC 106 in the x direction may not be stable in the first third and last third (i.e., the left and right sides) of the reflector 102b. In the case where it is desired to reduce the preference for a single polarization, for example, when it is desired to have a device with a known polarization direction and a low polarization ratio, the position of the aperture 104 can be moved to the end of the reflector 102b. Thus, the number of directions in which loss is introduced can be reduced. For example, if the aperture 104 is placed at the left third of the lens 102b, as shown in FIG. 6A and 6B , the axial direction of the loss is introduced in the +x direction. However, the loss introduced by the ROC 106 in the -x direction is less than when the aperture 104 is placed at the center of the lens 102b, thereby reducing the loss introduced in this direction and reducing the polarization preference in the y direction.

[0057] In other exemplary embodiments, the difference in ROC 106 between the Q direction and the S direction may be varied to advantageously produce a desired level of polarization preference. y =70μm and ROC x =infinity), as similarly presented in the embodiments of FIGS. 5A to 6B, this structure is considered as a reference structure for the following discussion regarding FIGS. 7A to 7D. In addition to adjusting the arrangement of aperture 104 on reflector 102 to change the number of axes where loss is introduced, the ROC 106 of reflector 102 in the loss direction can be adjusted as discussed further below.

[0058] 7A-7D are schematic diagrams of cross-sections of various example VCSEL devices, with corresponding plan views, having different ROC values ​​in the second direction. x It can be reduced from infinity as shown in FIG. 7A to a ROCUL value of ˜250 μm as shown in FIG. 7B to smoothly modulate the loss difference between the x-polarization and the y-polarization. Relative to the structure of FIG. 7A having reflector lens shapes 702a and 710a, the reflector lens shapes 702b and 710b shown in FIG. 7B can advantageously provide a reduced polarization preference. In practice, this can be achieved by reducing the length of the polymer pattern so that the reflector forms a smooth curved surface with a nearly constant second-order derivative near the top of the lens 102, thereby defining a finite positive ROC 106. This embodiment structure of FIG. 7B can be formed by making a rectangular pattern with a length of 80 μm and a width of 50 μm.

[0059] Alternatively, in another exemplary embodiment, the polarization preference can be further increased by using a reflector with a negative ROC value in the x-direction, as shown in Figure 7C. This can be described as a saddle point geometry for the reflector lens shapes 702c and 710c. Using a concave ROC 106 in the x-direction will expand the beam and result in additional losses in absorption and diffraction. As a result, the losses in the x-direction can be greatly increased, which may be advantageous for situations where there is low loss outside the aperture and therefore a significant beam expansion is required to achieve sufficient loss difference to achieve stable polarization control. Alternatively, if the ROCUL value is large enough that it is difficult to form a reflector 102 with a ROC 106 greater than ROCUL, such as in the case of a shorter cavity 208, the concave mirror 102c provides a further increase in the loss difference between the directions. The creation of saddle points can be created by careful consideration of the polymer length. For certain surface tension values, such as Figure 8 As shown, after reflow, the end of the polymer will tend to produce a bulge in the cross section. By reducing the cavity length so that the active areas of the two convex parts 810a and 810b begin to interact with each other, a smoothly curved shape 800 can be formed. For example, a rectangle with a length of 100μ0 and a width of 400 degrees can produce this shape.

[0060] Finally, in another exemplary embodiment, the ROC 106 in the direction of loss can be changed to introduce significant losses through a different mechanism. Instead of increasing the ROC 106 to make the beam wider and have a greater overlap with the lossy region, the ROC 106 can be reduced to less than the cavity length, as shown in FIG7D . By making the ROC 106 smaller than the cavity length, it can be considered as a mirror "too focused" and leading to an unstable resonant mode. Therefore, in the ROC y The standard is 70 μm and ROC x Below the cavity length, for example 15 μm, only the components of the resonant mode aligned along the y direction are stable. This can be achieved by making the length of the polymer rectangle corresponding to the Q direction at a standard value of 40 μm, while the width corresponds to the S direction and can be 18 μm.

[0061] For the above exemplary embodiments of FIGS. 7A-7D , all of these example designs may have a ROC of 70 μm. y , while Figure 7A has an infinite ROC x ; FIG. 7B has a ROC less than infinity and equal to or greater than ROCUL (eg, 250 μm); FIG. 7C has a negative ROC x , such as -500 μm; and FIG. 7D has a ROC greater than zero but less than the cavity length x , such as 15μm.

[0062] As briefly mentioned above, Figure 8is a diagram of the process used to reduce the length to form a saddle point. Figure 8 It is shown how reducing the length of the polymer rectangle can give a saddle point, for example, comparing a first length 802, a shorter length 804, a shorter length 806 that creates a saddle point, and an even shorter length 808. It should be understood that if the length is too long (e.g., 802) or too short (e.g., 808), a saddle point may not be formed. The range of acceptable lengths can be controlled by changing the properties of the resist.

[0063] It should be noted that the ROC value near the aperture 104 is a relevant aspect of the mirror shape for controlling beam characteristics and mode control. So far, the mirror 102 has been depicted as having a footprint with curved edges. However, the mirror 102 may also have one or both ends truncated, as shown in Figures 9A-9C. All three structures have ROC values ​​that are not intended to introduce additional losses. y , such as 70μm, and infinite ROC x (e.g., >5,000 μm). In an exemplary embodiment, a ROC 106 greater than 5,000 μm may be considered to be substantially infinite and provide a substantially flat surface because the ROC 106 may be provided to be at least one order of magnitude larger in one direction than the other, and preferably two orders of magnitude larger. As shown in Figures 9A-9C, the difference in footprints 910a, 910b, 910c may be due to mask design affecting the reflow properties at the ends of the reflector, or it may be due to additional process steps etching specific portions of the reflector. Figure 9A shows a footprint pattern 910a in a plan view and a reflector shape 902a in a cross-sectional view for the previously described reflector shape. Figure 9B shows a footprint pattern 910b in a plan view and a reflector shape 902b in a cross-sectional view with one side truncated. Figure 9C shows a footprint pattern 910b in a plan view and a mirror shape 902b in a cross-sectional view with both sides truncated.

[0064] In addition to changing the radius of curvature of the mirror, the aperture shape can be changed to affect the loss profile in the cavity 208. Since the main spatial dependence of the loss profile is usually determined by the overlap of the transverse mode profile and the aperture shape, in some embodiments, the shape of the aperture 104 relative to the mode defined by the curved mirror 102 can be quite important. Similar to changing the relative ROC values ​​in the x and y directions, the shape of the aperture 104 can be changed to modulate the polarization preference. x Infinity and ROC y Larger than cavity length and smaller than ROCUL y The mirror 102 will again be used as a reference device for the following discussion of another exemplary embodiment, but with the aperture shape altered in a manner that may be used in conjunction with any of the aforementioned changes in mirror curvature.

[0065] Fig.10 Plan views of various example apertures for VCSEL devices according to example embodiments of the present disclosure are provided.Aperture shape 104a is circular as previously described with reference to the above embodiments. Fig.10 Also shown are elliptical aperture shapes 104b, rectangular aperture shapes 104c, and diamond aperture shapes 104d. The illustrated set of different aperture shapes 104a-d can be realized with the same reflector parameters (e.g., 102a and 102b). Changing the aperture x and y dimensions affects the ROCUL in that direction, and therefore affects the loss in that direction.

[0066] To increase the preference for polarization directions parallel to the y-direction, an elliptical aperture shape 104b may be used, where the radius of the aperture in the x-direction may be reduced relative to the radius in the y-direction. This will reduce the ROCUL x and increase the loss observed in this direction. For example, an aperture 104b with a radius of 1.5 μm in the x-direction and a radius of 2 μm in the y-direction may be used. Alternatively, Fig.10 As shown, by increasing the aperture radius in the y direction (r y ) to increase ROCUL y , the loss in the y direction can be reduced. For example, using r x 2μm, r y The 3μm elliptical lens can be ROCUL y Increased to about 1200μm.

[0067] In addition to using smoothly curved aperture shapes, edge shapes or polygonal shapes, such as rectangular aperture shapes 104c, diamond aperture shapes 104d, or triangular shapes, can be used. These shapes can have different side lengths to modulate the loss curve, similar to the use of elliptical aperture shapes 104b, but also have more potential for modulating the gain curve. In the previous exemplary embodiment, the VCSEL structure is configured to adjust the overlap of the mode with the net gain / loss curve by changing the shape of the mode relative to the lateral loss curve, while assuming that the gain curve remains unchanged. However, in the absence of completely uniform injection, as is the case in most practical systems, the gain characteristic curve will also change. Typically, due to current diffusion problems, the current may be more concentrated at the edge of the aperture. For circularly symmetric apertures, this aspect does not provide polarization preference, although it may have limited effects on elliptical apertures. However, relative to the mode shape produced by the anisotropic reflector, controlling the current injection non-uniformity can lead to further preference for a specific polarization. For example, the electric field is typically concentrated around the corners, and the use of a diamond aperture may result in an increase in the injected current around the corners. By changing the angle of the corner, the electric field profile can be controlled to increase the injected current along the desired polarization direction to increase the polarization preference in that direction. Alternatively, the net gain / loss can be controllably reduced by this approach. For example, referring to aperture shape 104d, where a y 2μm, a x A mirror radius of 1.5 μm will provide a similar loss preference to an elliptical mirror of the same radius, but may have an enhanced gain difference between the two directions due to imperfect current injection.

[0068] Using a curved mirror with a non-circular aperture is advantageous because the beam shape will be heavily influenced by the shape of the reflector. This means that even if, for example, a rectangular aperture shape is used, the reflector shape can be chosen to keep the output beam relatively circularly symmetric. For example, for a rectangular aperture shape 104b, its r x 2μm, and r y 3μm, corresponding to ROCUL x About 250μm, ROCUL Y is about 1200 μm, a mirror 102 with a constant radius of curvature of 500 μm may be selected. This will result in losses in the x-direction because the ROC x >ROCUL x , but still allows the laser to be parallel to the y direction, because L <ROC y <ROCUL y In this case, even though the aperture shape 104b is rectangular, the output beam can be circularly symmetric, thereby retaining one of the many advantages of VCSEL devices.

[0069] For diamond-shaped apertures, different shapes of the diamond (e.g., narrow, normal, wide) can provide different effects on the current distribution in the aperture 104d. Higher current regions can be located in narrow corners and, therefore, can have higher gain, which can enhance or reduce overlap with the preferred mode. Wide corners can be lower current regions than narrow corners, and inner regions away from any corners can be even lower current regions. Therefore, to provide neutral gain alignment, as Fig.10 The aperture shape 104d shown in FIG. 1 may have a x =a y Thus, the center region of the aperture 104d is the lowest current region, the center of the edge is the higher current region, and the four corners are the higher current regions than the center of the edge of 104d. To provide enhanced gain alignment, the aperture shape 104d may have a x >a y , so that the top and bottom corners have the greatest current concentration, thereby enhancing gain. To provide reduced gain alignment, the aperture shape 104d may have a x <a y , so that the left and right corners have the maximum current concentration, thereby reducing the gain.

[0070] Thus, in the case of enhanced gain alignment, the top and bottom of the aperture shape 104d have larger currents, while the middle has relatively smaller currents, and conversely, in the case of reduced gain alignment, the right and left sides of the aperture shape 104d have larger currents, while the middle has relatively smaller currents. As will be appreciated by those skilled in the art, apertures of other shapes will be similarly affected with respect to the effect of corners on current distribution.

[0071] The disclosed method of using anisotropic mirrors and / or aperture parameters to control patterns is not limited to a single emitter, but is well suited for multiple emitters or arrays of devices using the same lens. Fig.11 is a diagram of different exemplary embodiments of an array of emitters on interconnected reflectors. An example for implementing a multi-emitter arrangement is a 1-D array 111a with VCSEL devices along the same reflector, such as Fig.11 As shown. Integrating all devices on the same reflector allows for smaller spacing between devices because the footprint of the end of the reflector does not need to be considered for minimum device spacing. This implementation can be extended from 1-D to 2-D emitter arrays. The symmetry of the array can be changed depending on the desired emission characteristics of the array. Fig.11 An array 111b with triangular cells is shown, with regular distances between emitters. Similarly, Fig.11An array 111c with rectangular cells is shown, with transmitters spaced at regular intervals from each other. The triangular array 111b can induce three different polarization directions, each 60 degrees apart, while the rectangular array 111c can produce two directions, 90 degrees apart. Well-defined polarization directions and ratios can be advantageous for applications where relatively equal emissions from known polarizations are desired, such as for polarization multiplexed information transmission. Furthermore, the number of polarizations can be increased by placing the transmitters around a circular reflector. Fig.11 An array 111d is shown with a circular shape. If the emitters are relatively evenly spaced, with a spacing of ≦π / 2 radians, the total emission of the group of emitters can be independent of polarization. This embodiment can reduce speckle of the emitted light and is highly advantageous for display applications. However, the chips do not need to be regularly spaced, as array 111e shows multiple emitters with different spacings. Fig.11 These shaped arrays 111 can be tiled to create larger device arrays than those shown, and the device pitch can be varied depending on the application. For example, tiling a triangular base can be used to create a hexagonal lattice structure.

[0072] Finally, the orientation of the mirror anisotropy can be aligned with the relevant crystal orientation of the device. Semi-polar and non-polar III-nitride based devices have a polarization preference parallel to the a-direction that increases with tilt angle. By aligning the mirrors so that the additional losses are perpendicular to the a-direction, the preference for a single polarization can be enhanced. This may be useful for low tilt angle semi-polar orientations or where the device structure induces strain in the holes, changing the band structure. Alternatively, if it is desired to suppress the polarization direction, the mirror losses can be aligned parallel to the a-direction, although this may result in reduced device performance.

[0073] The main effect of the present invention is to allow control of the cavity mode loss / gain curve in VCSELs with curved mirrors, which can then be used to advantageously implement polarization control of the device. The mode parameters of VCSELs are determined by the overlap of different modes defined by their transverse, longitudinal and polarization characteristics with the net gain / loss in the cavity. By increasing the loss of certain modes, those modes can be suppressed from lasing. Thus, the desired mode can be selected by accurately aligning the gain spectrum and increasing the loss of all other modes. In the curved mirror VCSEL cavity 208, the size of the beam, and therefore its overlap with the loss region outside the aperture 104, is controlled by the mirror ROC 106. However, by using an asymmetric mirror, the asymmetry of the beam shape can be introduced. This asymmetry gives the loss curve directionality, thereby promoting the attenuation of modes with intensity oscillations in the low loss direction by suppressing the mode with large overlap in the loss direction. Thus, a mode with a single oscillation direction can be generated, wherein the polarization direction is determined based on the asymmetry in the curved mirror.

[0074] The methods and structures disclosed herein use relatively simple changes in processing to introduce anisotropic losses into the laser cavity, which promotes single-polarization lasing parallel to the low-loss direction. This minor change in processing should make it a simple adjustment to current manufacturing lines, thereby achieving high production volumes and reducing implementation costs.

[0075] The technique is material system and wavelength independent, allowing for broad applicability. Additionally, by maintaining the ROC of the preferred polarization, there should be no additional losses in that mode, allowing for retention of polarization-insensitive device performance.

[0076] Changing the reflector shape allows fine tuning of the polarization preference and can be combined with adjustments to the aperture shape to enhance this preference. By aligning the reflector loss direction with a non-circular aperture, the losses can be maximized beyond what simple changes to the reflector would allow. For example, aligning the ROC=infinity direction with the smaller radius of an elliptical aperture will result in more polarization preference than a circular aperture.

[0077] Finally, the mirror fabrication method lends itself well to making arrays of devices. These devices can be aligned along a single curved mirror or grid to give different numbers of known polarizations. Setting the polarization direction can be beneficial for applications such as polarization-dependent multiplexing. If, instead of a straight mirror, a curved mirror is used, then the polarization along the lens can be smoothly rotated. Furthermore, an array of devices around a ring-shaped reflector will have a total output that is relatively independent of polarization. This can reduce speckle noise, which is very beneficial for displays.

[0078] In an exemplary embodiment, the curvature radius anisotropic VCSEL cavity mirror is configured to provide mode control. Considering the four directions +x, +y, -x, -y, the curvature radius in at least two directions (+x, -x or +y, -y) is greater than the cavity length and less than ROCUL, and at least one of the other directions is outside this range.

[0079] In another exemplary embodiment, the aperture may have a circular shape, an elliptical shape, a rectangular shape, or a triangular shape.

[0080] In another exemplary embodiment, the aperture has a shape that combines (adds or subtracts) any of the previous shapes.

[0081] In another exemplary embodiment, the mirror has a radius of curvature along one axis that is nearly infinite (>5,000) or substantially flat, and a radius of curvature in another axis that is finite and greater than the cavity length.

[0082] In another exemplary embodiment, the aperture is in the middle third of the lens length.

[0083] In an alternative embodiment, the aperture is in the first third of the lens length.

[0084] In another exemplary embodiment, the mirror has a radius of curvature along at least one axis that is smaller than the cavity length and larger than the cavity length in other directions and is finite.

[0085] In another exemplary embodiment, the radius of curvature along at least one axis is concave and convex in another direction, which corresponds to at least one direction having a positive ROC and at least one direction having a negative ROC.

[0086] In another exemplary embodiment, the devices are arranged by using multiple apertures of the same lens.

[0087] In yet another exemplary embodiment, the holes are in a 1-D array, a triangular lattice, a rectangular lattice, a circular pattern, and / or are placed at varying pitches.

[0088] In an exemplary embodiment provided in another perspective illustrated using the mathematical relationship shown below, the "S" direction corresponds to the direction in which loss is introduced into the mode, the values ​​in this direction belong to the set S, and the "Q" direction is the direction in which the mirror is intended to work properly, the values ​​in this direction belong to the set Q.

[0089] [Mathematical formula 8]

[0090]

[0091] In this exemplary embodiment, the opening may have any shape as discussed above.

[0092] In another exemplary embodiment, the ROC is expressed as follows:

[0093] [Mathematical formula 9]

[0094]

[0095] In an example, the aperture may be in the middle third of the lens length. In another embodiment, the aperture may be in the first third of the lens length.

[0096] In another exemplary embodiment, the ROC is expressed as follows:

[0097] [Formula 10]

[0098]

[0099] In these embodiments, the device may use the same lenses in multiple aperture arrangements, including, for example, 1-D arrays, triangular lattices, rectangular lattices, circular patterns, and / or placed at varying pitches.

[0100] Therefore, the present disclosure can generally be achieved by setting anisotropic ROC values ​​in different directions of the curved mirror to provide the above-mentioned advantageous effects and improve the beam characteristics, wherein the curves in different directions are different from each other, for example as shown in the above mathematical relationship. In addition or alternatively, the beneficial effects of the present invention can be achieved by setting the aperture shape differently in multiple directions to provide anisotropic ROCUL values ​​in different directions. Thus, in an alternative embodiment, an isotropic circular curved mirror with anisotropy introduced in the aperture shape (e.g., rectangular, rhombus, elliptical or a combination of shapes) can satisfy the above mathematical relationship and similarly provide the advantageous effects on the beam properties as discussed above. In this example, the VCSEL includes an isotropic arc mirror and an anisotropic aperture, wherein the cavity length is less than the ROC, and in at least one direction (e.g., the Q direction along the y-axis) the ROC is less than the ROCUL, and in another direction (e.g., the S direction along the x-axis) the ROC is greater than the ROCUL or less than the cavity length. In this embodiment, for example, any combination of the above-described aperture shapes, aperture arrangements, and / or beam shaping properties may exist when the above-described mathematical relationships are met. In addition, as described above, any combination of beam shaping properties (including mirror curvature and aperture shape) may be combined to achieve favorable anisotropic beam properties (e.g., focusing and loss, polarization control) for a VCSEL device with a curved mirror.

[0101] It should be understood that the relative sizes of components within the drawings may not be to scale and are presented for illustrative and illustration purposes only.

[0102] A list of numbered exemplary embodiments is provided below:

[0103] Embodiment 1. A VCSEL device, comprising: a curved mirror of the VCSEL; and a radius of curvature (ROC) of the anisotropic curved mirror, wherein the ROC includes four directions, the four directions being +x, +y, -x, and -y, and in at least one direction, the ROC is within a range of predefined ROC values ​​greater than a cavity length of the VCSEL and less than a standard beam width (ROCUL), and in at least one other direction, the ROC is outside the range.

[0104] Embodiment 2. The VCSEL device of embodiment 1, wherein the aperture of the VCSEL has a circular shape.

[0105] Embodiment 3. The VCSEL device of embodiment 1, wherein the aperture of the VCSEL aperture has an oval shape.

[0106] Embodiment 4. The VCSEL device of embodiment 1, wherein the aperture of the VCSEL has a rectangular shape.

[0107] Embodiment 5. The VCSEL device of embodiment 1, wherein the aperture of the VCSEL has a triangular shape.

[0108] Embodiment 6. The VCSEL device of embodiment 1, wherein the aperture of the VCSEL has a shape comprising a combination of at least two of circular, elliptical, rectangular, or triangular.

[0109] Embodiment 7. The VCSEL device of embodiment 1, wherein the ROC along a first axis is nearly infinite such that the cavity mirror is substantially flat along the axis, and the ROC along a second, different axis is greater than the cavity length and is finite.

[0110] Embodiment 8. The VCSEL device of embodiment 7, wherein the aperture is located in the middle third of the first axis.

[0111] Embodiment 9. The VCSEL device of embodiment 7, wherein the aperture is located in the first third of the first axis.

[0112] Embodiment 10. The VCSEL device of embodiment 1, wherein the ROC along a first axis is smaller than the cavity length, and the ROC along a second, different axis is larger than the cavity length and finite.

[0113] Embodiment 11. The VCSEL device of embodiment 10, wherein the ROC along the first axis is negative and the ROC along the second axis is positive.

[0114] Embodiment 12. The VCSEL device of embodiment 1, wherein the plurality of apertures are arranged on the curved mirror surface.

[0115] Embodiment 13. The VCSEL device of embodiment 12, wherein the plurality of apertures are in a one-dimensional array.

[0116] Embodiment 14. The VCSEL device of Embodiment 12, wherein the plurality of apertures are arranged in a triangular lattice.

[0117] Embodiment 15. The VCSEL device of embodiment 12, wherein the plurality of holes are arranged in a rectangular grid.

[0118] Embodiment 16. The VCSEL device of Embodiment 12, wherein the plurality of apertures are placed in a circular pattern.

[0119] Embodiment 17. The VCSEL device of embodiment 12, wherein the plurality of apertures are placed at varying pitches.

[0120] Embodiment 18. The VCSEL device of embodiment 1, wherein ROCUL is the half power beam width.

[0121] Embodiment 19. The VCSEL device of embodiment 1, wherein the ROC is configured to control a mode of the VCSEL.

[0122] Embodiment 20. The VCSEL device of embodiment 1, wherein the laser cavity is made of GaN, InP, GaAs or other semiconductor materials.

[0123] Embodiment 21. A VCSEL cavity mirror, comprising: a curved mirror of the VCSEL; and a radius of curvature (ROC) of an anisotropic curved mirror surface, wherein the ROC includes four directions, the four directions are +x, +y, -x, -y, the ROC in at least one direction is within a range of predefined ROC values ​​greater than the cavity length of the VCSEL and less than a standard beam width (ROCUL), and in at least one other direction, the ROC is outside the range.

[0124] Embodiment 22. A light emitting device comprising a VCSEL cavity mirror according to embodiment 21. It should be understood that various changes and modifications to the exemplary embodiments described herein will be apparent to those skilled in the art. Such changes and modifications may be made without departing from the spirit and scope of the subject matter and without diminishing its intended advantages. Therefore, these changes and modifications are intended to be covered by the appended claims.

Claims

1. A vertical cavity surface emitting laser (VCSEL) device, comprising: cavity mirror, having a curved mirror surface; as well as The radius of curvature (ROC) of the anisotropic curved mirror surface, wherein, The ROC includes four directions, which are +x, +y, -x, and -y. In at least one direction, the ROC is within a range of predefined ROC values ​​greater than the cavity length of the VCSEL and less than a standard beam width (ROCUL), and In at least one other direction, the ROC is outside the range.

2. The VCSEL device according to claim 1, wherein: The aperture of the VCSEL has a circular shape.

3. The VCSEL device according to claim 1, wherein: The aperture of the VCSEL hole has an elliptical shape.

4. The VCSEL device according to claim 1, wherein: The aperture of the VCSEL has a rectangular shape.

5. The VCSEL device according to claim 1, wherein: The aperture of the VCSEL has a triangular shape.

6. The VCSEL device according to claim 1, wherein: The aperture of the VCSEL has a shape including at least two combinations of circular, elliptical, rectangular or triangular shapes.

7. The VCSEL device according to claim 1, wherein: The ROC along a first axis is nearly infinite, such that the cavity mirror is substantially flat along that axis, and the ROC along a second, different axis is greater than the cavity length and is finite.

8. The VCSEL device according to claim 7, wherein: The aperture is located in the middle third of the first axis.

9. The VCSEL device according to claim 7, wherein: The aperture is located in the front third of the first axis.

10. The VCSEL device according to claim 1, wherein: The ROC along a first axis is less than the lumen length, and the ROC along a second, different axis is greater than the lumen length and is finite.

11. The VCSEL device according to claim 10, wherein: The ROC along the first axis is negative and the ROC along the second axis is positive.

12. The VCSEL device according to claim 1, wherein: A plurality of apertures are arranged on the curved mirror surface.

13. The VCSEL device according to claim 12, wherein: The plurality of apertures are in a one-dimensional array.

14. The VCSEL device according to claim 12, wherein: The plurality of apertures are arranged in a triangular lattice.

15. The VCSEL device according to claim 12, wherein: The plurality of apertures are arranged in a rectangular grid.

16. The VCSEL device according to claim 12, wherein: The plurality of apertures are positioned in a circular pattern.

17. The VCSEL device according to claim 12, wherein: The plurality of apertures are placed at varying intervals.

18. The VCSEL device according to claim 1, wherein: The ROCUL is the half power beam width.

19. The VCSEL device according to claim 1, wherein: The ROC is configured to control a mode of the VCSEL.

20. The VCSEL device according to claim 1, wherein: The laser cavity is made of GaN, InP, GaAs or other semiconductor materials.

21. A VCSEL cavity mirror, comprising: The curved mirror of the VCSEL; as well as The radius of curvature (ROC) of the anisotropic curved mirror surface, wherein, The ROC includes four directions, which are +x, +y, -x, and -y. In at least one direction, the ROC is within a range of predefined ROC values ​​greater than a cavity length of the VCSEL and less than a standard beam width (ROCUL), and In at least one other direction, the ROC is outside the range.

22. A light emitting device comprising the VCSEL cavity mirror according to claim 21.

Images