Integrated optics and electronics
An optoelectronic device includes a submount and a lid. The submount includes a substrate and a lens and a laser above the substrate. The lid defines a cavity having a surface coated with a reflective material to form a 45 degree mirror. The mirror reflects a light from the laser to the lens and the light exits the optoelectronic device through the submount.
This invention relates to a wafer-scale package for an optoelectronic device, and more specifically to a wafer-scale submount for an optoelectronic device.
DESCRIPTION OF RELATED ARTOptoelectronic (OE) devices are generally packaged as individual die. This means of assembly is often slow and labor intensive, resulting in higher product cost. Thus, what is needed is a method to improve the packaging of OE devices.
SUMMARYIn one embodiment of the invention, an optoelectronic device includes a submount and a lid. The submount includes a lens and a laser above a substrate. The lid defines a cavity having a surface coated with a reflective material to form a 45 degree mirror. The mirror reflects a light from the laser to the lens and the light exits the optoelectronic device through the submount.
BRIEF DESCRIPTION OF THE DRAWINGS
Use of the same reference symbols in different figures indicates similar or identical items. The cross-sectional figures are not drawn to scale and are only for illustrative purposes.
DETAILED DESCRIPTIONAn optoelectronic package may include a submount wafer, a ring wafer bonded to the submount wafer, and a lid wafer bonded to the ring wafer. The lid wafer typically includes an integrated lens. The submount wafer typically includes an edge-emitting laser and interconnects for powering the laser. The ring wafer is typically processed using RIE etching to form straight walls. An additional component that has a 45 degree surface is placed next to the laser inside the ring. This component acts as a mirror to reflect the light from the laser up through the lens in the lid. Alternatively, the ring wafer is processed to form an integrated 45 degree mirror that reflects the light from the laser up through the lens in the lid wafer.
The optoelectronic package requires two hermetic seals: one between the ring wafer and the lid wafer, and another between the ring wafer and the submount wafer. In the case where there is an additional mirror component, that too has to be aligned and attached. The optoelectronic package also requires three wafers to be processed. Additionally, to maintain the correct path length, it is necessary to use two thin wafers (e.g., 275 microns) for the ring wafer and the lid/lens wafer. Thus, what is needed is an apparatus that addresses these disadvantages of the optoelectronic package.
Integrated Optics and Electronics
In step 12, as illustrated in
To form the stack, an amorphous silicon layer is first formed on substrate 54. The amorphous silicon layer can be deposited by low pressure chemical vapor deposition (LPCVD) at 550° C. or by plasma enhanced chemical vapor deposition (PECVD). The thickness of the amorphous silicon layer can be determined by the following formula:
In the above equation, t is the phase shifting lens layer, λ is the target wavelength, N is the number of the phase shifting lens layer, and Δni is the difference in the refractive index (ni) between phase shifting lens material and it's surrounding. In one embodiment where λ is 1310 nm, N is eight, ni of amorphous silicon is 3.6, and ni of silicon dioxide is 1.46, the amorphous silicon layer has a typical thickness of 765 angstroms.
A silicon dioxide (SiO2) layer is next formed on the amorphous silicon layer. The silicon dioxide layer can be thermally grown on the amorphous silicon layer in steam at 550° C. Alternatively, silicon dioxide layer can be deposited by PECVD. The silicon dioxide layer has a typical thickness of 50 angstroms. The process of depositing amorphous silicon and low temperature thermal oxidation of the amorphous silicon is repeated for the desired number of phase shift layers.
Once the stack is formed, each layer is masked and etched to form the desired diffractive lens. The silicon dioxide layer on the top amorphous silicon layer is first dipped off using a diluted water/hydrofluoric acid (HF) solution (typically 50:1). A photoresist is next spun, exposed, and developed on the amorphous silicon layer. The amorphous silicon layer is then plasma etched down to the next silicon dioxide layer, which acts as the etch stop. The process of masking and etching is repeated for the remaining phase shifting layers.
In one embodiment, lens 52 is a bifocal diffractive lens that converts laser light into a small angle distribution that is spread uniformly throughout a volume. The volume's dimensions are large relative to the size of the input face of an optical fiber so the components can easily align. The bifocal diffractive lens has a surface with ridges that provide two focal lengths f1 and f2. A design process for the bifocal diffractive lens can begin with determining the first phase function that defines a surface contour for a conventional diffractive lens having focal length f1. Any conventional techniques for diffractive lens design can be used. In particular, commercial software such as GLAD from Applied Optics Research, Inc. or DIFFRACT from MM Research, Inc. can analyze the phase functions of diffractive elements. A second phase function is similarly generated, wherein the second phase function is such that if the second phase function were multiplexed together with the first phase function, the combination would provide a diffractive lens having the second focal length f2. The second phase function is then scaled so as to provide a partially efficient diffractive lens that focuses a percentage (e.g., 50%) of the incident light but passes the remainder (e.g., 50%) of the incident light unperturbed. The first phase function and the scaled second phase function are multiplexed together to form a final bifocal lens design.
In another embodiment, lens 52 is a hybrid diffractive/refractive element. The hybrid diffractive/refractive element spreads the light from over a volume to expand the alignment tolerance for an optical fiber as described above. The hybrid diffractive/refractive lens has at least one surface with a curvature for one focal length, e.g., f2. Further, diffractive features of a partially efficient diffractive lens are superimposed on one or both surfaces of hybrid diffractive/refractive lens so that the combination provides two focal length f1 and f2 for separate fractions of the incidence light
In step 14, as illustrated in
In step 16, as illustrated in FIGS. 4 to 6, a metal layer 1 is formed over oxide layer 56 and then patterned. In one embodiment, metal layer 1 (
In step 20, as illustrated in
In step 22, as illustrated in
In step 24, as illustrated in FIGS. 11 to 13, a metal layer 2 is formed over dielectric layer 64. In one embodiment, metal layer 2 is titanium-platinum-gold (TiPtAu) sequence deposited by evaporation. Titanium has a typical thickness of 0.1 micron, platinum has a typical thickness of 0.1 micron, and gold has a typical thickness of 0.5 micron. Metal layer 2 is formed to form contact pads and bonding pads. In one embodiment, a photoresist is spun, exposed, and developed to form a liftoff mask 72 (
Metal layers 1 and 2 can be patterned to form 2 interconnect levels. The two interconnect levels can be connected by plugs in between the two levels.
Contact pads 82 and 84 provide electrical connections to laser die 122. Contact pads 82 and 84 are connected by respective buried traces 90 and 92 to respective contact pads 94 and 96 located outside seal ring 106. Contact pads 82 and 84 are part of metal layer 2 formed and patterned in step 24. Traces 90 and 92 are part of metal layer 1 formed and patterned in step 16.
Contact pads 86 and 88 provide electrical connection to monitor photodiode die 124. Contact pads 86 and 88 are connected by respective buried traces 98 and 100 to respective contact pads 102 and 104 located outside seal ring 106. Contact pads 86 and 88 are part of metal layer 2 formed and patterned in step 24. Traces 98 and 100 are part of metal layer 1 formed and patterned in step 16.
In step 28, as illustrated in
In step 30, as illustrated in
In one embodiment, lid 130 has a <100> plane at a 9.74 degree offset from a major surface 138. Lid 130 is wet etched so that surface 132 forms along a <111> plane of the silicon substrate. As the <100> plane of lid 130 is at a 9.74 degree offset from major surface 138, then the <111> plane and mirror 135 are oriented at a 45 degree offset from major surface 138.
In step 32, as illustrated in
As can be seen, a light 152 (e.g., 1310 nm) is emitted by laser die 122. Light 152 is reflected from mirror 135 downwards to lens 52. Lens 52 then focuses light 152 so it can be received by an optical fiber at a specified location. As insulator layer 64, oxide layer 56, and substrate 54 are transparent to light 152, light 152 can exit optoelectronic device 150 through submount 80.
In step 34, as illustrated in
As one skilled in the art understands, the process described above can be performed on a wafer-scale so that numerous OECEs 150 are formed simultaneously. These OECEs 150 are then singulated to form individual packages.
OECE 150 offers several advantages over the conventional optoelectronic package. First, only two wafers are needed to make OECE 150 instead of three wafers for a conventional package. Second, the wafers can be of standard thickness (e.g., 675 microns) instead of two thin wafers for a conventional package. Third, only one hermetic seal is needed between lid 130 and submount 80 instead of two for a conventional package.
Alignment Post for Optical Subassemblies
OSA 212 is not complete and testable until its elements have been aligned and fixed into their proper positions. This alignment is usually accomplished by powering OE device 220 and moving TO can 228 in X, Y, and Z directions relative to port 224. This alignment is then “fixed,” generally either with a polymer adhesive or by a laser-welding process.
OSA designs vary significantly from product to product but they usually involve a packaged device (e.g., OE device 220), a lens (e.g., lenses 222), and a fiber alignment feature (e.g., port 224). The fiber alignment feature is usually a precision hole fabricated with injection-molded plastic or ceramic to accept a ceramic ferrule (e.g., ferrule 216).
There is a continuous push to manufacture smaller and cheaper OSAs. There are many good reasons related to cost, quality, and functionality for wanting a small OSA. However, the small OSA is not complete until it includes an alignment feature. Thus, what is needed is an alignment feature for small OSAs.
It is important to note that although alignment post 304 may look similar to port 224 (
The concept of aligning to an OD (i.e., to a post) is subtly different from aligning to an ID (i.e., to a hole) but offers two key advantages: cost and size.
Cost—It is very easy and economical to manufacture posts with a precision diameter. This is because a long rod can be made by grinding the OD and then many parts can be made by simply slicing off pieces of the rod. The cost of making a precision feature with perhaps a micron or two of tolerance is important to keeping the cost of OSA 306 to a minimum. The cheapest precision feature one can make is a sphere (e.g., a ball bearing) and probably the second cheapest precision feature is a cylinder.
Size—OECE 302 can be manufactured in a two-dimensional array of parts. This manufacturing method would produce hundreds or even thousands of OSAs 306 complete except for the alignment features. Ideally, the alignment features would be added while the OSAs 306 are still in array form but this will only be possible if the alignment feature is smaller than the footprint of OECE 302.
Another advantage of a small OSA 306 is that it can be aligned closer together with another OSA 306 to mate with new smaller FO connectors. In fact, one of the historical reasons for the current size of duplex connectors (such as the duplex LC connectors) goes back to how close together two TO cans can be aligned into ports. OSA 306 would thus allow for smaller connectors and smaller transceivers.
Various other adaptations and combinations of features of the embodiments disclosed are within the scope of the invention. For example, submount 80 can include additional active and passive circuitry. Specifically, submount 80 can be processed to form passive circuitry such as resistors and capacitors, and active circuitry such as transistors. Submount can also be processed to include a bipolar CMOS (BiCMOS) integrated circuit. Numerous embodiments are encompassed by the following claims.
Claims
1. A laser submount, comprising:
- a substrate;
- a lens above the substrate; and
- a laser above the substrate.
2. The laser submount of claim 1, wherein the substrate is selected from the group consisting of silicon, quartz, sodium borosilicate glass, sapphire, gallium arsenide, silicon carbide, and gallium phosphide.
3. The laser submount of claim 1, further comprising:
- a planarization layer covering the lens; and
- an interconnect layer above the planarization layer.
4. The laser submount of claim 3, wherein the planarization layer is an oxide layer.
5. The laser submount of claim 3, further comprising:
- a dielectric layer above the interconnect layer; and
- a contact pad above the dielectric layer, wherein the laser is electrically connected to the contact pad.
6. The laser submount of claim 5, further comprising:
- a sealing ring above the dielectric layer and surrounding the contact pad and the laser.
7. The laser submount of claim 1, further comprising:
- at least one of a passive integrated circuit and an active integrated circuit.
8. A method for forming a laser submount, comprising:
- forming a lens above a substrate; and
- mounting a laser to the laser submount above the substrate.
9. The method of claim 8, wherein the substrate is selected from the group consisting of silicon, quartz, sodium borosilicate glass, sapphire, gallium arsenide, silicon carbide, and gallium phosphide.
10. The method of claim 8, further comprising, subsequent to said forming a lens and prior to said mounting a laser:
- forming a planarization layer covering the lens; and
- forming an interconnect layer above the planarization layer.
11. The method of claim 10, wherein the planarization layer is an oxide layer.
12. The method of claim 10, further comprising, subsequent to said forming an interconnect layer and prior to said mounting a laser:
- forming a dielectric layer covering the interconnect layer; and
- forming a contact pad above the dielectric layer, wherein the laser is electrically connected to the contact pad.
13. The method of claim 12, further comprising, subsequent to said forming a dielectric and prior to said mounting a laser:
- forming a sealing ring above the dielectric layer and surrounding the contact pad and the laser.
14. The method of claim 8, further comprising:
- forming at least one of a passive integrated circuit and an active integrate circuit above the substrate.
Type: Application
Filed: Sep 19, 2003
Publication Date: Mar 24, 2005
Inventors: Kendra Gallup (Marina del Rey, CA), James Matthews (Milpitas, CA)
Application Number: 10/666,442