US20020110328A1 - Multi-channel laser pump source for optical amplifiers - Google Patents

Abstract

An optical assembly, such as a multiple output diode laser pump source for EDFAs, is formed by pressing an optical array emitter chip against a standoff structure protruding from a submount such that the emitter chip deforms to match the curvature of the standoff structure. An IO chip is also juxtaposed against the standoff structure such that its optical receivers can receive optical energy from the emitter chip. The IO chip can provide various optical functions, and then provide an optical array output for coupling into an optical fiber array. The standoff structure preferably contacts the emitter chip over an aggregate contact area much smaller than the area by which the emitter chip overlaps the submount. The materials used for bonding the emitter chip and the IO chip to the submount are disposed in the recesses between standoffs and not on the contact surfaces of the standoff structure.

Description

BACKGROUND OF TIIE INVENTION
• 1. Field of the Invention[0001]
• This invention relates generally to semiconductor diode laser based pump sources, and more specifically to techniques for construction of planar multi-channel diode laser pump sources.[0002]
• 2. Description of Related Art[0003]
• Optical amplifiers are an accepted part of long-haul telecommunications systems. They are used to amplify signals after optical fiber propagation losses over long transmission distances typical of such as the nationwide networks. A typical system uses a plurality of Erbium doped fiber amplifiers (EDFAs) pumped by semiconductor diode lasers. Semiconductor diode laser pump sources for EDFAs typically operate at wavelengths of 980 nanometers (nm) or 1480 nm. The EDFA is capable of amplifying wavelengths over a wide bandwidth with a gain spectrum that peaks at about 1530 nm and that typically extends to 1570 nm, or 1620 nm in advanced configurations. Usable output or optical gain is achievable over this 40 to 90 nm region. This wide bandwidth provides the opportunity for the optical signal to be carried on a large number of wavelength channels that can be independently and all-optically amplified by an EDFA. This technique of wavelength division multiplexing (WDM) is currently driving the expansion of modem telecommunications.[0004]
• The EDFA has been available for about 10 years, during which time the performance of the device has increased markedly, benefitting in particular from improved performance of semiconductor diode laser pump sources. However, this improved performance is typically accompanied by increased cost. The increased cost is readily tolerated in the high value, relatively low fiber count long haul system because each fiber is able to carry many data channels via WDM, each data channel being amplified simultaneously within a single EDFA. Thus the cost of the high performance EDFA is shared amongst many revenue generating data streams and subscribers.[0005]
• To achieve continued expansion of data carrying bandwidth to the office and home, the fiber optic transmission system must be extended from the point-to-point long-haul network to Metropolitan (Metro) and access networks. The more diffuse nature of the Metro network, and the need to service users on a more individual basis means that less data is carried onto a single fiber, generally causing proposed Metro networks to be characterized as having fewer data channels per fiber at lower transmission rates and more individual fiber transmission paths and shorter span lengths than long haul systems. The lower data channel count per fiber means less revenue per fiber, and the increased number of separate fibers and decreased span length means increased component numbers. The combination of these two factors leads to a requirement of lower cost components to enable profitable Metro network implementation.[0006]
• EDFAs play a key role in Metro networks, just as they do in the long-haul backbone. The use of EDFAs enable longer ring or mesh propagation distances within the network, and also enable the use of lossy all-optical components such as wavelength demultiplexers and multiplexers, or optical cross connects without the need for costly detection, electrical regeneration, and reemission/modulation of the data signals. Thus a lower cost implementation of the current EDFA found in long-haul networks is required to drive the installation and commissioning of Metro networks.[0007]
• A typical long-haul network EDFA is composed of a number of subsystems or components including one or more erbium doped fiber sections, optical isolators to eliminate back reflection, and one or more semiconductor laser diode pumps with their associated wavelength couplers to combine them with the 1550 nm data signal on the network fiber. A significant proportion of the overall cost of the amplifier results from the semiconductor diode laser pumps, which typically cost many thousands of dollars each. Thus, a lower cost implementation of the diode laser pump source would enable lower cost EDFAs for application in Metro networks.[0008]
• Currently, two main types of diode laser pump sources exist: those that operate at 980 nm and those that operate at 1480 nm. These two wavelengths are absorbed quite efficiently by the erbium ions in the fiber core and offer different performance characteristics for the overall amplifier system. Pumping at 980 nm is usually chosen for pre-amplifiers where low noise amplification is important, as the 980 nm pumping may lead to a more complete population inversion of the emitting erbium state and to a correspondingly lower amplifier noise figure as compared to 1480 nm pumping. Diode lasers operating at 1480 nm are often chosen for high output power amplifiers as the optical-optical conversion efficiency is higher and the dollar cost per mW of output power from the diode laser is generally lower than for 980 nm diode lasers.[0009]
• Prior art semiconductor diode laser pump sources operating at 980 nm (a very similar device configuration is used at 1480 nm, simply utilizing a different semiconductor material system to generate the different wavelengths) generally consist of a number of individual components, shown symbolically in FIG. 1. A[0010]diode laser chip105 is soldered to asubmount110 to provide thermal heatsinking and electrical connection. Eachdiode laser chip105 has a single active laser region that is capable of generating and emitting several hundred mW of output power in a single transverse optical mode. Submount110 is positioned inside abutterfly package115 that has the capability of achieving a hermetic seal on final closure. A single-modeoptical fiber120 is fed through a ferrule and opening inbutterfly package115 and then brought into alignment with the output aperture ofdiode laser chip105. To achieve a high coupling efficiency betweendiode laser chip105 and single-modeoptical fiber120, a lens or chisel shaped tip may be formed on the end face offiber120 so that the rapidly diverging optical mode ofdiode laser chip105 is efficiently converted into the relatively much larger and slower diverging mode offiber120.
• In addition to the need for the lensed or chisel ended[0011]fiber120, there are also very tight constraints placed on the positioning of the tip offiber120 relative to the emitting aperture ofdiode laser chip105. In fact, it is necessary to control the position offiber120 to sub-micron accuracy to achieve optimum coupling. This precise control is typically achieved by holdingfiber120 via a computer controlled multi-axis micropositioner.Diode laser chip105 is energized to generate output light, and the output fromfiber120 is monitored using a photodiode or power meter. The micropositioners then movefiber120 to optimize for maximum signal transmitted therethrough, after whichfiber120 is fixed in position, typically by laser welding of a metallizedfiber ferrule125 to aholder clip130 mounted to the package orsubmount110. Often it is necessary to tweak the alignment offiber120 after initial fixing with further laser-assisted or mechanical bending ofholder clip120.
• The fully active fiber alignment process described above is both cumbersome and slow, and although it can result in remarkably good coupling efficiency between the diode laser and fiber (in excess of 60%), it does not lend itself well to high volume, high yield and low cost manufacturing. It is this fully active alignment step that accounts for a significant portion of the cost of constructing a diode laser pump source for an EDFA.[0012]
• In the prior art, attempts have also been made to construct multi-channel integrated laser arrays and to align them with integrated optoelectronic chips or directly with fiber arrays. The use of a self-aligned solder assembly with mechanical stops and misaligned solder joints is reported to provide three-dimensional passive alignment between the laser diode axis of each diode and a corresponding optical axis of an optical fiber with lateral misalignment of ±2 microns and vertical misalignment of ±0.75 microns, with coupling losses of about 4 dB per channel reported for a 4 diode array.[0013]
• Such prior art techniques do not provide the level of precision obtained using the active fiber alignment described above, and which is conventionally required to achieve efficient coupling between a diode laser and an optical fiber. Difficulties include height variation (that is spacing between the diode laser chip and the substrate) across the lateral array dimension due to solder thickness variations or bonding pressure variations. Thus, co-planarity of the bonding surfaces is difficult to achieve to the degree desired for very efficient coupling (typically about 0.2 microns over 10 mm for a high channel count laser diode array).[0014]
• Bowing of the diode laser chip also gives rise to misalignments across an array of emitters. Fabrication of the diode laser structure using epitaxial growth and planar surface lithography often results in a laser array chip with a residual bow or warpage.[0015]
• In addition, misalignment can also be caused by particles trapped between the diode array and the mounting surface. Keeping large planar surfaces free of particles is difficult. Even a single sub-micron particle is sufficient to cause severe misalignment along a large array. In one prior art approach, the presence of foreign particles is accommodated by supporting the laser array on a pair of standoffs, above the large substrate area, one standoff at each end of the array. This may help to minimize the effect of foreign particles, but does not address the absolute positioning of the multiple emitters across the array, and cannot alleviate misalignment arising from laser array curvature due to warped or bowed wafers.[0016]
• In view of the problems associated with prior art techniques for manufacturing pump sources, it is an object of this invention to provide a semiconductor laser array pump source for optical amplifiers which may be manufactured relatively easily and inexpensively, and which enables precise optical alignment of components even in the presence of foreign particles and/or component warpage or bowing.[0017]
SUMMARY OF THE INVENTION
• Roughly described, a multiple output diode laser pump source is manufactured by an array scaleable optical alignment process, which can achieve simultaneous highly efficient coupling between each of the emitters in a laser diode array and a respective receiving waveguide array. The waveguide array may be fabricated (without limitation) in a lithium niobate, glass substrate, or other integrated optics chip. The simultaneous coupling of all of the emitter/waveguide pairs advantageously reduces the number of manufacturing steps, thereby allowing the pump source to be manufactured more rapidly and inexpensively.[0018]
• According to one aspect of the invention, the simultaneous array scalable alignment process utilizes a standoff structure on a submount to define a reference surface for mounting of a laser array and a receiving integrated optic waveguide chip. Mounting both chips to the single reference surface permits optical alignment in the most critical dimension, perpendicular to the plane of the arrays, to be performed passively, leaving active alignment required only in the transverse and longitudinal dimensions and the yaw axis.[0019]
• In addition, the standoff structure also can overcome other problems in the prior art by, for example, eliminating the effect of solder deposition thickness on emitter position, lessening the impact of foreign particulate defects by decreasing the contact area between the arrays and the submount, and alleviating the effects of non-uniform bonding pressure and chip bow or warp by distributing the reference surface across the entire array and therefore referencing the positions relative to the submount of all emitters directly to the same reference surface.[0020]
• The invention will be better understood upon reference to the following detailed description in connection with the accompanying drawings:[0021]
BRIEF DESCRIPTION OF DRAWINGS
• FIG. 1 is a symbolic diagram of a prior art diode laser pump source.[0022]
• FIG. 2 is a symbolic diagram of a laser array pump source according to an embodiment of the present invention, shown without a protective package.[0023]
• FIG. 3 is a cross-sectional view taken along line A-A in FIG. 2.[0024]
• FIG. 4 is a fragmentary symbolic diagram of a submount according to an embodiment of this invention.[0025]
• FIG. 5 is a symbolic diagram of a second embodiment of the invention, showing in particular an IO waveguide array chip with distributed Bragg reflector structures (DBRs) for wavelength stabilization and waveguide tapers for mode size conversion.[0026]
• FIG. 6 is a symbolic diagram of fused-fiber-optic directional couplers arranged to combine light output at different wavelengths from several emitters of a laser array.[0027]
• FIG. 7 is a symbolic diagram of waveguide directional couplers formed on the IO waveguide chip and arranged to combine the light output at different wavelengths from several emitters of a laser array.[0028]
• FIG. 8 is a symbolic diagram of optical fiber polarization multiplexers arranged to combine the light output from pairs of emitters of a laser array.[0029]
• FIG. 9 is a block diagram illustrating detector array locations for output monitoring and power stabilization.[0030]
• FIG. 10 is a symbolic diagram, show in plan view, of an integrated power monitoring detector array on an IO waveguide array chip.[0031]
• FIG. 11 is a cross-sectional view taken along line A-A of FIG. 10 showing locations of distributed Bragg reflector and optical detector chip (two alternatives) for power monitoring.[0032]
• FIG. 12 is a symbolic diagram, shown in plan view, of an integrated power monitoring detector array using waveguide directional couplers.[0033]
• FIG. 13 is a symbolic diagram showing a redundancy-providing array design in plan view.[0034]
• FIG. 14 is a symbolic diagram, shown in plan view, of an embodiment providing for redundancy using an optical fiber switch matrix.[0035]
• FIG. 15 is a symbolic diagram, shown in plan view, of an embodiment providing for redundancy using a waveguide switch matrix on an IO waveguide array chip.[0036]
• FIG. 16 is a schematic diagram of a passive network of optical fiber couplers and splitters for power sharing among the optical fiber outputs of an array source.[0037]
• FIG. 17 is a symbolic diagram, shown in plan view, of an optical fiber array disposed in v-grooves on a submount.[0038]
• FIG. 18[0039]ais a symbolic diagram of a laser array pump source using two submounts with a standoff structure.
• FIG. 18[0040]bis a symbolic diagram, shown in elevated end and perspective views, of an optical fiber array assembled on a silicon substrate adapted with v-grooves.
• FIG. 18[0041]cis a symbolic diagram of an IO waveguide array chip coupled to a v-groove optical fiber array, on a submount with a standoff structure and relief slots.
• FIG. 19[0042]ais a symbolic diagram, shown in plan and elevated side views, of an alternative embodiment of the invention wherein a laser array is coupled directly to an optical fiber array.
• FIG. 19[0043]bis a symbolic diagram showing a perspective view of the alternative embodiment of the invention depicted by FIG. 19a.
• FIG. 20[0044]ais a symbolic diagram showing a waveguide directional coupler for wavelength multiplexing, and distributed Bragg reflectors (DBRs) for wavelength stabilization positioned before the directional coupler, arranged on an IO array chip.
• FIG. 20[0045]bis a symbolic diagram showing a waveguide directional coupler for wavelength multiplexing, and DBRs for wavelength stabilization positioned in series after the directional coupler, arranged on an IO array chip.
• FIG. 20[0046]cis a symbolic diagram showing a waveguide directional coupler for wavelength multiplexing, and DBRs for wavelength stabilization positioned in series after the directional coupler, with one DBR on the array chip and another DBR on an output fiber.
• FIG. 20[0047]dis a symbolic diagram showing a waveguide directional coupler for wavelength multiplexing, and DBRs for wavelength stabilization positioned in series on the output fiber.
• FIG. 21 is a block diagram of multiple EDFAs pumped by a single laser array pump source of the invention.[0048]
• FIG. 22 is a fragmentary perspective view of the FIG. 2 embodiment, showing in particular the relationship between corresponding ports of the laser array and IO waveguide chip.[0049]
DETAILED DESCRIPTION
• The invention will now be described in reference to various non-limiting embodiments thereof. It should be noted that individual elements or features of different embodiments described below may be combined in various permutations to produce a laser pump source having a set of desired physical and/or operational characteristics, and that such combinations are within the scope of the present invention.[0050]
• FIG. 2 illustrates an embodiment of a semiconductor diode[0051]laser pump source200 in accordance with aspects of the invention. Pumpsource200 is shown to include several major components: asubmount205, a diode laser array (hereinafter “laser array”)210, an integrated optic (IO)waveguide chip215 and anoptical fiber array220. As used herein, an optical “array” is a device that includes two or more optical ports. Thesubmount205,laser array210 andIO chip215 are shown in greater detail in FIG. 22, which illustrates in particular opposingedges250 and255 oflaser array210 andIO waveguide chip215.Laser array210 andIO waveguide chip215 are each provided with a plurality ofoptical ports280 and285 arranged along a respectivesubject edge250 or255. The optical ports may take the form of waveguide inputs and outputs, positioned such that thesubject edges250 and255 oflaser array210 andIO waveguide chip215 are substantially perpendicular to theoptical axes290 of the waveguide inputs and outputs, with thelateral spacing240 in the horizontal dimension parallel to the subject edges being substantially the same in both components, at least at the subject edges. In this manner, the optical ports oflaser array210 and the optical ports ofIO waveguide chip215 can communicate with each other if they are arranged in such a way that a sufficient fraction of the optical energy emitted from one of the ports is captured by the other optical port. A condition of perfect alignment between corresponding ports, wherein the optical axes of the pair of corresponding ports are coincident, is not required. Rather, two ports “can communicate” with each other even if their optical axes are spatially or angularly offset, provided that such spatial and/or angular offset does not exceed a desired tolerance. Corresponding ports also “can communicate” with each other if the ports are aligned (or misaligned by no more than a desired tolerance) with an optical path that includes an optical redirector (such as a reflector, refractor, or re-emitter) positioned intermediate in the optical path between the ports. Similarly, as used herein, one optical port “can receive” optical energy from another optical port if the two optical ports “can communicate” with each other.
• As is known in the art,[0052]laser array210 is configured to emit light responsive to application of electrical signals thereto. The term “light” as used herein is not limited to visible light, but rather includes any optical energy having a wavelength or range of wavelengths suitable for a specified application. Light emitted bylaser array210 is coupled into the receiving end ofIO waveguide chip215, where a waveguide taper array may be used to convert the mode emitted bylaser array210 to that of a single mode optical fiber.IO waveguide chip215 may also include other functionalities as described below, such as incorporation of distributed Bragg reflectors (DBRs) for wavelength stabilization, directional couplers for wavelength multiplexing, and switch networks for redundancy. At the output end ofIO waveguide chip215, the transmitted and mode-converted light is coupled into anoptical fiber array220 to be conveyed, for example, to the amplification regions of EDFAs. In accordance with an aspect of the present invention, assembly of the several components to form thepump source200 is advantageously effected by simultaneous optical alignment of the emitters oflaser array210 to corresponding receiver waveguides fabricated inIO waveguide chip215, followed by (or performed concurrently with) the alignment ofoptical fiber array220 to the output end ofIO waveguide chip215. As is discussed in further detail hereinbelow, simultaneous alignment of the laser emitters to the corresponding receiving waveguides may be achieved by utilization of asubmount205 having astandoff structure222 to whichlaser array210 andIO waveguide chip215 are attached.
• As illustrated in FIG. 2,[0053]standoff structure222 may comprise a set of discreteparallel standoffs230 fabricated on and protruding vertically upward from a firstmajor surface227 ofsubmount205.Standoffs230 terminate at their upper end in upper surfaces, at least a portion of which contact corresponding surfaces oflaser array210 andIO waveguide chip215. As used herein, the term “vertical” is intended to mean substantially perpendicular to the major planes ofsubmount205,laser array210 andIO waveguide chip215. Areas of firstmajor surface227 lying betweendiscrete standoffs230 define an array ofwells229 or recessed regions of a depth suitable for receiving solder balls and/or adhesive (for example a glue, epoxy or other such bonding agent). Thewells229 may also carry circuit metallization traces and contact pads for electrical interconnects.
• [0054]Standoff structure222 may be fabricated on firstmajor surface227 ofsubmount205 by photolithographic and selective etching processes. Alternatively,standoff structure222 may be fabricated by laser ablation, by depositing a layer of material anddefining standoff structure222 by photolithography and a solvent or etchant, by positioning and affixing material preformed in a predetermined thickness and shape, or by a combination of the foregoing techniques.Standoff structure222 may alternatively be formed on the appropriate major surface oflaser array210 or of integratedoptic waveguide chip215, or partially onlaser diode array210 or the integratedoptic waveguide chip215, and partially onsubmount205. As shown, discrete standoffs orribs230 extend primarily longitudinally on submount205 (the longitudinal axis being in the horizontal plane ofsubmount205 and oriented substantially perpendicular to the subject edges oflaser array210 and IO waveguide chip215).
• The function of[0055]standoff structure222 is made clearer with reference to FIG. 3, which represents a cross-sectional view taken along the line A-A of FIG. 2, and with further reference to FIG. 4, which depicts a fragmentary perspective view ofsubmount205 in the absence oflaser array210 and10waveguide chip215.Ribs230 contact laser array at a plurality ofcontact portions232. Contactportions232 collectively define areference surface305.Reference surface305 will typically be substantially planar, although some implementations of the invention may utilize standoff structures which define a curved reference surface (i.e., a reference surface having a finite radius of curvature along the lateral and/or longitudinal axes). It will be appreciated that some portions of the standoff structure may be shorter than others, either intentionally or unintentionally, and those will not form contact portions withlaser array210 and do not count in the definition of thereference surface305.
• It should be noted that[0056]standoff structure222 may be formed in a variety of geometries and is not limited to the parallel discrete rib design depicted in FIGS.2-4. For example,standoff structure222 may be formed as a continuous serpentine-shaped structure, or as a comb-type structure having a plurality of teeth joined at one end by a common spine. Irrespective of the exact shape whichstandoff structure222 takes, it is preferable that at least three ofcontact portions232 occur consecutively along a substantially straight line parallel and close to the subject edge of the chip being supported (i.e.,laser array210 or IO waveguide chip215), and that those threecontact portions232 are mutually isolated from each other along that line inreference surface305. Note thatcontact portions232 may, however, connect with each other at locations not on such line, and therefore may not be entirely isolated from each other; but they should be mutually isolated from each other at least along some line in thereference surface305 parallel to and in close proximity to the subject edge.
• It is to be appreciated that, in the lateral dimension along the subject edge,[0057]contact portions232 should be sufficient in number, spacing and extent to effectively define the curvature of the subject edge for a specified application. For many applications, the subject edge will have an infinite radius of curvature, i.e., will be linear. In the longitudinal dimension,contact portions232 should be sufficient in number, spacing, extent and proximity to the subject edge that they control the curvature of the subject edge adequately for the desired application. It is preferable that at least one contact portion included under theoptical component210,215 is sufficiently close to the subject edge to define the vertical angles of the optical axes of each of the optical ports adequately for the desired application.
• It should be appreciated that the present invention does not require a uniform height for all elements of[0058]standoff structure222. In some embodiments, required curvatures and optical axes' angles are best achieved with non-uniform standoff structure heights. For example, a standoff structure may comprise a stepped standoff structure comprising a first set of standoffs having a first height, and a second set of standoffs having a second height different from the first height. The first set of standoffs may then be utilized to supportlaser array210, and the second set of standoffs may be utilized to supportIO waveguide chip215. In this case, it should be noted that the two reference surfaces defined by the first and second sets of standoffs are not co-planar. The stepped standoff structure design may be useful in achieving vertical alignment of optical ports whenlaser array210 andIO waveguide chip215 have substantially different thicknesses. In addition, the stepped standoff structure may provide a physical stop which enables precise longitudinal positioning ofIO chip215, thus facilitating butt coupling tolaser array210. It will be apparent to one skilled in the art that protrusions functioning as physical stops may be incorporated, where desired, into standoff structures having an otherwise uniform height.
• It should be further noted that[0059]contact portions232 of thestandoff structure222 that are under one optical component, forexample laser array210, can be either continuous with or discontinuous (distinct) from corresponding contact portions located under the second optical array component, for exampleIO waveguide chip215.Standoff structure222 should be substantially rigid, at least along the longitudinal axis thereof to achieve proper alignment of the optical ports oflaser array210 andIO waveguide chip215. In the vertical dimension, standoff structures should have sufficient rigidity to withstand the pressures exerted by the chuck pressing against it, as described elsewhere herein. If the pressures are exerted at elevated temperatures, which might be the case if certain bonding agents such as solder are being used, then the vertical rigidity of the standoff structure should be maintained even at the elevated temperature. In this sense, a solder bump which is liquified during mounting does not qualify as part of a standoff structure.
• As used herein the terms “top,” “bottom,” “lower” and “upper and the like are used solely for convenience in referring to particular levels. The levels they refer to are not intended to change if the structure is turned upside down or tilted.[0060]
• It is further noted that the output light emerging from the emitters of[0061]laser array210 diverges rapidly. The rate of this divergence is inversely related to the dimension of the optical mode within the waveguide cavity. For optimally efficient operation of the laser diode emitters, the multiple layers composing the gain (or diode junction) and waveguiding regions in the semiconductor chip are constructed such that the optical mode is tightly confined, with typical optical mode diameters of 1 μm in the vertical dimension (perpendicular to the plane of the diode laser junction) and 3 μm in the lateral dimension transverse to the optical axis of the diode laser waveguide. Thus the laser diode output diverges faster in the vertical dimension perpendicular to the plane of the laser diode junction, compared to the divergence rate in the lateral dimension. This small vertical dimension and the corresponding rapid divergence makes the vertical alignment between two such waveguides the most critical, i.e. having the smallest allowable misalignment tolerance for efficient coupling.
• To achieve accurate alignment between the emitters of[0062]laser array210 and the corresponding waveguides ofIO waveguide chip215,standoff structure222 defines areference surface305, as illustrated by and described earlier in connection with FIG. 3. Thisreference surface305 is used to accurately align the optical axes of the emitters ofarray210 and the receiving waveguides of the integratedoptical waveguide chip215 in the critical vertical dimension. This approach preferably employs a flip-chip bonding technique, which involves bonding components ofpump source200 with the active surface (also variously referred to as the “top”, “upper” or “circuit” surface) of thecomponent facing submount205. According to typical construction methods of 10 devices (such aslaser array210 or IO waveguide chip215), several devices of a given type are typically fabricated simultaneously by photolithographic and planar processing techniques on a single wafer, which is then separated into identical smaller units referred to as chips, components, or devices. As a consequence of the method of construction, the active and passive optical and electronic circuit structures of the devices, such as waveguides, p-n junctions, and other structures known in the art, are typically disposed near one major surface of the IO component, herein referred to as the “active” side.
• Both[0063]laser array210 andIO waveguide chip215 are preferably designed and fabricated such that the centroids of the optical modes of each are located at a very well known distance from the respective active surfaces. Forlaser array210, this condition may be accomplished by the use of controlled epitaxial growth of substantially planar layers on a planar substrate, which are then processed with lithographic techniques to define the active region. ForIO waveguide chip215, this condition may be achieved by using the carefully controlled in diffusion of a species into the surface of a planar substrate, e.g., diffusion of protons into lithium niobate. Typically, the distance in the vertical dimension between the centroid of the optical mode and the active surface inlaser array210 will be different than the corresponding distance inIO waveguide chip215. To correct for this difference, a thin film material of an appropriate thickness may be deposited on the active surface of one of the components (i.e., the component having the smaller centroid-to-surface distance) such that the two distances are rendered substantially equal. The thin film material may comprise, for example, SiO₂deposited by sputtering, or by ion beam assisted deposition, which technique enables control of the film thickness to an accuracy of a few nanometers. The thin film material may alternatively be a metal such as gold. Preferably, the thin film material is patterned into discrete pads spatially corresponding to some or all ofcontact areas232. Well-known photolithographic techniques, such as photoresist liftoff or etching with photoresist protection, may be used to pattern the pads. Alternatively, the pads maybe patterned by deposition through a shadow mask. The patterning serves to provide pads with sufficient area to support the associated component (e.g., IO waveguide chip215), while minimizing any stress resulting from thermal expansion mismatch between the component and the thin film material. Other techniques which may be used to compensate for differences in the centroid-to-surface distance include interposing, during assembly, preformed metal pads of appropriate thickness between the component having the shorter-centroid-to-surface distance and the underlying portions ofstandoff structure222.
• Assembly of[0064]laser array210 to submount205 may be performed according to the following process. As described above,standoff structure222 is prepared on a firstmajor surface227 ofsubmount205. If desired, asuitable solder405 may be deposited intowells229 defined byribs230 ofstandoff structure222, as illustrated in FIG. 4, to enable bonding of diodelaser array chip210 tosubmount205. In another embodiment the adhesive is placed in recesses laterally outside the standoff structure.Laser array210 andsubmount205 are next placed on respective chucks within a flip chip bonder. At least one of the chucks of the bonder preferably provides a compliant or deformable layer as described in detail below.Laser array210 and thesubmount205 are brought into initial alignment, which may be preformed, for example, by using a combination of optical alignment and fiducial marks for lateral and in-plane (yaw) angular alignment, and an autocollimator to set parallelism (roll and pitch angular alignment). The flip chip bonder is then used to bring the top or active surface oflaser array210 into contact with thereference surface305 defined bystandoffs230. As it is practically impossible to orient thelaser array210 and thesubmount205 in perfectly parallel relation, especially when one considers that the fabrication process of thelaser array210 generally results in a warped or bowed chip, one portion of the active surface oflaser array210 will contact the corresponding area ofstandoff structure222 before remaining portions of the active surface are brought into contact withstandoff structure222.
• Following the establishment of initial contact between[0065]laser array210 andstandoff structure222, force is applied by the flip chip bonder to bring thechucks holding submount205 andlaser array chip210 still closer together, thus pressinglaser array210 ontoreference surface305 ofstandoff structure222. Force is transmitted tolaser array210 through a compliant or deformable layer disposed between thelaser array chip210 and the metal (or ceramic) chuck which holds thechip210. This compliant layer can be compressed and distorted, thereby distributing the applied force over the area of thelaser array chip210 and permitting it to flex and/or rotate such that it becomes substantially uniformly contacted to thereference surface305 across its entire area. The distributed applied force is sufficient to rotate thelaser array chip210 to overcome any remaining nonparallelism.
• In essence,[0066]laser array210 is pressed ontostandoff structure222, optionally deforminglaser array chip210 orsubmount205 or both from its or their originally provided shape, until the resulting curvature of thelaser array chip210 substantially matches the resulting curvature ofreference surface305. Once such uniform contact is achieved the flip chip bonder may be used to heat the two components to a temperature sufficient to causesolder405, which was previously deposited inwells229 between stand-offs230, to ball up andcontact laser array210. On cooling,solder405 solidifies and the force applied by the flip chip bonder may be released. The solder forms a strong bond betweenlaser array210 andsubmount205 which may be used for both mechanical support and, if desired, electrical connection between the laser diode drive circuitry and the laser emitters themselves, and also for providing a good thermal contact for cooling.
• It will be noted that the combination of the[0067]standoff structure222 and the compliant layer disposed betweenlaser array210 and the chuck of the flip chip bonder enables uniform contacting of the laser diode array, to thereference surface305 defined bystandoff structure222, across the entire width oflaser array210. Thus, all the emitters of thelaser diode array210 are accurately located relative to thereference surface305. This location is achieved independently of the presence of curvature or warping in either thesubmount205 or thelaser array chip210, both as originally provided and as finally bonded.
• It will also be noted that the position of the emitters relative to[0068]reference surface305 is determined only by the accuracy of thickness control in the layer deposition and formation processes used to constructlaser array210. There is no solder or adhesive located between the active surface of laser array and thecontact portions232 ofstandoffs230; rather, the solder or adhesive405 is located in thewells229 between thestandoffs230, and so variations in the thickness or volume ofsolder405 deposited on thesubmount205 do not affect the alignment of thelaser diode array210.
• In addition, the use of[0069]standoff structure222 comprising a set ofstandoffs230 arranged across the width oflaser array210, advantageously reduces the actual contact area betweenlaser array210 andsubmount205 compared to an assembly wherein the entire planar surface oflaser array210 is contacted with a corresponding planar surface of thesubmount205. This reduced contact area decreases the probability of a material defect or foreign object (e.g., a dust or debris particle) being located between thelaser array210 and thesubmount205 at a point of contact, and therefore decreases the probability of a resulting misalignment that will adversely affect the coupling efficiency. As shown in FIG. 2, the aggregate contact area between thestandoffs230 and each of the optical array components (i.e.,laser array210 and IO waveguide chip215) is substantially less than the total area by which thesubmount205 overlaps each such optical array component. The overlap area is defined herein as the intersection area of a perpendicular projection of the optical components ontosubmount205. The aggregate contact area is preferably less than 50 percent, and more preferably less than 10 percent, of the corresponding overlap area in order to minimize misalignment arising from defects or the presence of foreign particles.
• Attachment of[0070]IO waveguide chip215 to thesubmount205/laser array210 subassembly (hereinafter “subassembly”) may be performed by a method closely similar to the above-described method for attachinglaser array210. However, active alignment techniques may be employed for positioning ofIO waveguide chip215, as is set forth below.
• During the attachment procedure, the subassembly may be mounted on a fixed or movable stage and provided with electrical connections to driver circuitry to energize the laser emitters.[0071]IO waveguide chip215 is preferably mounted with its active or top surface (at or proximal to which are located the waveguides) facing thestandoffs230 fabricated on the major surface225 of thesubmount205, with the receiving ends of the waveguides facing corresponding emitters oflaser array210.IO waveguide chip215 may be mounted in a suitable holder attached to a computer controlled multi-axis micropositioner capable of submicron positioning accuracy and repeatability, such as an autoalign system.IO waveguide chip215 is initially aligned relative to the assembly using optical alignment and fiducial marks to set lateral and longitudinal position and yaw angle, and an autocollimator to set parallelism to submount (roll and pitch angles).IO waveguide chip215 is then brought into close proximity tostandoffs230, with the receiving end ofIO waveguide chip215 located adjacent to the emitting edge oflaser array210. It will be understood that this alignment approach is based on butt coupling between thelaser array210 andIO waveguide chip215, requiring close approach of the facets of the two components to prevent diffraction losses between the two. Energizinglaser array210 results in the emission of light from the emitters, which is captured by the receiving waveguides inIO waveguide chip215. A photodetector (or array of detectors) is used to monitor the light transmitted throughIO waveguide chip215. The position ofIO waveguide chip215 is then adjusted under computer control to maximize the transmitted light. Importantly, adjustment need be performed only in the lateral, longitudinal and yaw dimensions since the vertical position and the pitch and roll angles are defined bystandoff structure222. Once alignment is optimized (as indicated by maximization of the light transmitted through the waveguide(s) of IO waveguide chip215), the position of the chip may be memorized by the computer control system, and theIO waveguide chip215 withdrawn to allow the dispensing of adhesive (e.g., an epoxy) intowells229 located betweenstandoffs230. Alternatively, solder deposited intowells229 can be used to provide mechanical fixing, if appropriate portions ofIO waveguide chip215 have been metallized. After deposition of the epoxy,IO waveguide chip215 is returned to its exact previous position using the highly accurate repeatability of the positioning stages. Final alignment is confirmed andIO waveguide chip215 is contacted and pressed ontostandoff structure222 using the micropositioners. If an epoxy is employed to provide adhesion betweenIO waveguide chip215 andsubmount205, the epoxy is subsequently cured, e.g., by UV exposure. Note that the mount supportingIO waveguide chip215 on the micropositioner is preferably designed to be compliant or deformable such that whenIO waveguide chip215 is contacted tostandoff structure222,IO waveguide chip215 is allowed to roll or pivot and flex if necessary to conform to the reference surface defined bystandoff structure222, and/or to remove any warpage or bowing inIO waveguide chip215. Alternatively, a passive optical process may be employed.
• By attaching[0072]IO waveguide chip215 in this manner, it can be understood that high coupling efficiency between the emitters oflaser array210 and the waveguides ofIO waveguide chip215 can be achieved, as the crucial vertical alignment of corresponding emitting and receiving optical port is performed passively and with very high accuracy.
• Following alignment and attachment of[0073]IO waveguide chip215, the light transmitted throughIO waveguide chip215 may be coupled into an array of (preferably single mode) optical fibers. The optical fibers may for instance be assembled in a silicon v-groove array as known in the art to accurately define the spacing between the fiber cores, and to position the fiber cores along a substantially straight line in the lateral dimension. Simultaneous alignment of all of theIO waveguide chip215 outputs into respective optical fibers may be achieved in a single active alignment step, where the position of the optical fiber array relative to the output facet of the IO waveguide chip is adjusted to maximize the light transmitted through the fibers to a photodetector. Once the alignment has been completed the position of the fiber array may be secured using, for instance, UV cured epoxy, solder, thermally cured epoxy, laser welding etc. to bond it to submount205 or alternatively, to the output facet ofIO waveguide chip215.
• In general, the alignment tolerances for optical coupling between the output waveguides of[0074]IO waveguide chip215 and the fiber array are considerably looser (i.e., greater) than those required for acceptable coupling between the diode laser and the input waveguides of the IO waveguide chip or between the diode laser and a lensed or chisel-tipped optical fiber. The looser tolerances are due to the difference in mode sizes between those inlaser array210 and the fiber array: in a typical implementation ofpump source200, the optical mode in a single mode optical fiber has a characteristic vertical dimension in the range of approximately 6 μm (for a fiber carrying 980 nm light) to 9 μm (for a fiber carrying 1480 nm light) fiber, as compared to an optical mode having a characteristic vertical dimension of approximately 1 μm inlaser array210. To accommodate the larger mode size of the optical fiber, the mode size is expanded in the IO waveguide chip, from a small size matching the diode laser mode at the input waveguide, to a larger size matching the optical fiber mode, at the output waveguide. A small misalignment (e.g., on a sub-micron scale) reduces the overlap between optical modes that are several microns in size by only a small fraction and, accordingly, has only a minor effect on the coupling efficiency between a fiber array and an IO waveguide chip with matching mode size. In contrast, the same amount of misalignment would have a major effect on the coupling efficiency between a laser diode array and an IO chip or fiber array due to the much smaller mode sizes involved. It is also possible to perform alignment of the fiber array using a standoff designed reference surface as used for the diode laser-IO waveguide chip alignment, as described in detail below.
• An embodiment of the invention will now be described in terms of specific exemplary implementations thereof It will be appreciated that the following implementations are intended to be illustrative rather than limiting.[0075]
• In a first exemplary implementation of diode[0076]laser pump source200,laser array210 is configured to emit light at a wavelength of around 980 μm.Laser array210 maybe fabricated from epitaxially grown layers of AlGaAs or InGaAs as known in the art. As discussed above,laser array210 comprises a plurality of individual emitters arranged in mutually parallel fashion. The total number of emitters may vary over an extended range (between 2 and several hundreds, inclusive) depending on requirements of a specific application, but will typically lie in the range between 4 and 64, inclusive. Each emitter preferably lases in a single transverse mode, known in the art as the (0,0) mode. The position of the centroid of the laser mode relative to the active surface of laser array may be determined either from optical measurements of the output mode pattern or from computations based on the refractive indices and thicknesses of the layers grown to form the laser junction and waveguide structure. This position may readily be determined to fractions of 1 μm.
• [0077]Laser array210 may be fabricated by methods known in the art involving the photolithographic patterning of precisely grown epitaxial layers to define the laser structure. Electrical contacts are applied by other techniques known in the art to ensure ohmic contact to the laser junction and to the backside of the wafer for the current return path. The laser arrays are then cleaved from the wafer to prepare the emission facets and cleaved or diced to size laterally to define the number of emitters per array. Lithographically defined alignment marks may be provided to guide the location of the cleaving steps. Suitable coatings may be applied to the cleaved or etched facets of the laser array to improve lifetime and to provide preferential emission of the output radiation from the output facet. Typically the output facet is coated to provide a relatively low reflectivity (1-30%) to provide the output coupling from the laser diode cavity, while the other, or back facet, is typically coated with a high reflectivity layer (for example, greater than 90%). Preferably the individual diode laser emitters are fabricated at a sufficient lateral spacing to minimize thermal cross-talk between emitters and to enable efficient heat extraction from the laser array without requiring an excessive laser diode junction temperature. The maximum optical output power should be chosen to remain below the threshold for catastrophic optical damage at the output facets of the emitters. In one implementation, the maximum output power of each individual emitter is in the range of 50-500 mW.
• The[0078]submount205 may preferably be fabricated from single-crystal silicon. Alternatively, materials such as lithium niobate and beryllium oxide may be used to provide a closer thermal expansion match to the GaAs material of the laser diode array. The submount material is preferably chosen to exhibit high thermal conductivity in order to enable efficient extraction of heat generated by the laser emitters during operation. In the case of a single-crystal silicon submount material,standoff structure222 may be defined using photolithographic patterning and wet etching of the silicon wafer surface. The upper surface ofstandoff structure222 comprises the unetched regions of the original surface of the silicon wafer and therefore retains the original surface's smoothness and flatness. The height ofstandoff structure222, is determined by the etch processing conditions of time, temperature and etch agent (among others) and may be controlled to sub-micron accuracy. In a typical implementation ofsubmount205, the height ofstandoff structure222 will be around 5 μm.
• Alternatively, for both silicon and the alternative materials (lithium niobate and beryllium oxide)[0079]standoff structure222 may be fabricated using the deposition of thin film materials over the surface of the original wafer. For instance, a layer of SiO₂or SiN may be deposited by RF sputtering, PECVD, ion beam assisted deposition, or other process as known in the art, with very well controlled absolute thickness (±10's of nm) and excellent uniformity across the wafer (<±0.5%).Standoff structure222 may then be formed by photolithographic patterning and etching of the thin film layer(s). For instance, in the case of SiO₂, a photoresist layer may be spun onto the thin film, exposed through a mask carrying the desired standoff pattern and developed. The patterned resist layer may then be used as a mask to etch the SiO₂layer for instance using a wet buffered oxide etch or a dry etch, e.g. reactive ion etching in CHF₃. The height of standoff structure is determined by the thickness of the film layers deposited on the surface. By depositing layers of different materials sequentially with controlled thickness, and then using chemically selective etching it is possible to construct a standoff structure having regions of different heights.
• [0080]Wells229 disposed betweenstandoffs230 maybe utilized as reservoirs for a bonding agent to attachlaser array210 tosubmount215. Electrical traces may also be disposed on the submount surface between thestandoffs230 such that electrical contact may be made to the emitters oflaser array210. In some instances it may be desirable for the electrical traces to enable individual connections for each emitter so that the operation and power output of each emitter may be controlled independently using suitable external drive circuitry. In others it may be desirable to gang several or all of the emitters together on common electrical connections. Either of these situations can readily be accommodated by the wafer scale photolithographic processes used to image and pattern the traces on the submount. For some choices of submount material, such as silicon, it may be necessary to provide an electrical isolation layer beneath the electrical traces to prevent short circuits and current leakage through the bulk of the submount.
• Metal pads may be fabricated in selected wells which correspond to bonding and/or electrical contact areas on[0081]laser array205. These metal pads may be coated with a layer of solder, e.g. and indium based solder deposited by evaporation or electroplating, or a gold-tin solder deposited by evaporation. The thickness and area of this solder is chosen such that it does not completely fill (either vertically or laterally) the well, but so that when it is melted it can ball up due to surface tension and contact the surface of a component being held against the reference surface defined by the standoff array. The type of solder to be used for this attachment is defined by the performance requirements of the assembled device. At a minimum the solder must remain solid and maintain a robust physical and/or electrical attachment throughout the desired operating and storage temperature range of the device, typically about −35° C. to about +85° C.
• The fabrication of[0082]submount205 may readily be performed on a wafer scale using standard planar lithographic processing techniques as known m the art. Once the fabrication processing is complete the individual submounts may be cut from the wafer using a dicing saw. As shown in FIGS.2-3, thesubmount205 may be sized such that it does not extend beyond the output end ofIO waveguide chip215, in order to facilitate the approach of the output fiber array to the output facet of theIO waveguide chip215 as described below.
• Attachment of[0083]laser array210 to submount205 may be performed using a modified flip chip bonder. The prepared submount with standoffs, electrical connections, solder preforms and lithographically patterned optical alignment or fiducial marks is loaded onto the substrate chuck of the bonder and held in place typically with vacuum pull from the backside. The prepared laser array, with electrical and bonding contact pads and facet coatings, is loaded into a specially modified sample chuck on the flip chip bonder.
• The chuck is modified such that interposed between the laser array chip and the metal or ceramic chuck is a layer of compliant or rubber like material. This material, for example, a polyimide, a silicone rubber, or a fluoroelastomer such as Vitong (manufactured by DuPont Dow Elastomers), is deposited on the chuck and subsequently processed to yield a flat surface, e.g., by lapping and polishing on a flat plate with progressively finer polishing compounds. Vacuum holes are formed in the compliant layer to match those present in the standard chuck to provide a holding force to secure the array chip.[0084]
• After loading[0085]laser array210 andsubmount205, an autocollimator is used to align the two components approximately parallel in the pitch and roll axes, and lateral, longitudinal and yaw angle adjustments are made to positionlaser array210 correctly with respect to the alignment marks onsubmount205, such that the electrical and/or attachment contact pads onlaser array210 will align with respective features on submount.
• The two chucks are then brought close together such that[0086]laser array210contacts standoff structure222. In general, the angular alignment oflaser array210 andsubmount205 is imperfect, andlaser array210 often is warped or bowed due to stresses introduced during the fabrication processing, such that one point or line onlaser array210 contacts the standoff structure before remaining portions. At this point, the compliant layer betweenlaser array210 and the sample chuck is compressed or distorted, accommodating angular misalignment, and allowinglaser array210 to rotate and/or flex such that it comes into substantially uniform contact withstandoff structure222. Although the compliant layer allows the rotation/flexing oflaser array210 to contact the standoff structure, it preferably resists lateral displacements or motion that would otherwise cause lateral or longitudinal misalignment oflaser array210 with respect tosubmount205. After contact is achieved the vacuum hold onlaser array210 may be released if desired.
• Subsequent to contact (or if preferred before contact is complete or during the contacting process) the temperature of the components is raised sufficiently high to melt the solder which was previously deposited and patterned in wells. On melting, the solder preferably balls up until it contacts and wets the contact pads on[0087]laser array210. Fluxes and other processes such as forming gas or formic acid vapor may be used as known in the art to improve the solder flow, wetting and adhesion properties. The subassembly is then cooled to harden the solder. If it has not already been released the vacuum hold on the sample may be removed during or after the cool down cycle.
• Once cool, the subassembly may be removed from the flip chip bonder for further processing. Electrical connections to the active side of[0088]laser array210 are preferably made by the solder joints described above. Electrical connection to the backside oflaser array210 to provide the current return path may be made by placing wire bonds from metallized pads on the backside of the semiconductor wafer to bonding pads located on the submount.
• In the presently-described implementation, IO waveguide chip comprises a lithium niobate (LN) crystal with waveguides fabricated therein using an annealed proton exchange (APE) method. Preferably the LN crystal is X- or Y-cut such that the waveguides support a TE polarization mode that matches the output polarization of typical laser emitters. Alternatively, a TM polarized diode laser may be used, coupled to an APE waveguide in Z-cut lithium niobate, which also supports a TM polarized mode. In alternate implementations,[0089]IO waveguide chip215 maybe formed using other suitable optical materials and methods of waveguide fabrication, including without limitation, ion exchanged or indiffused glass, multi-layer polymer stack, or silica-on-silicon structures.
• The receiving or input end of[0090]IO waveguide chip215 is adapted with an array of waveguides that having lateral spacing matched to that of thelaser array210 emitters. In addition, the waveguide dimensions and process parameters (such as anneal time/temperature) are chosen so as to create a very tightly confined optical mode with mode dimensions similar to that of the emitters (typically around 1 μm in the vertical dimension and 3 μm in the lateral dimension). The tightly confined waveguide mode inIO waveguide chip215 assures a high overlap integral between thelaser array205 output mode and theIO waveguide chip215 input mode when the facets of the two chips are brought into close proximity in the butt coupling procedure described below. In practice, the actual mode size in the input waveguides ofIO waveguide chip215 maybe somewhat larger than that of thelaser array210 emitters due to process constraints.
• A primary function of[0091]IO waveguide chip215 is to act as a mode size converter betweenlaser array210 and the single mode optical fibers employed at the output ofpump source200. Tapered waveguide sections are used for this purpose, according to known art, and it is in general possible to create almost any desired mode size by appropriate choice of physical mask dimensions, proton exchange time and temperature, and annealing time and temperature.
• The waveguide taper structure may be constructed in a single processing step, with a single set of exchange and anneal parameters and one lithographically defined mask. Alternatively the taper structure maybe constructed in several sequential steps using multiple exchanges, anneals, and masking steps. The single step construction method may not enable full optimization of the waveguide mode size at the extreme input and output regions of[0092]IO waveguide chip215, but it offers a far less complicated process than the sequential method, which requires several critical alignments. An example of single-step constructed waveguide taper structure formed inIO waveguide chip215 is a 4 μm wide mask dimension at the input/receiving end which tapers down to a 2 μm wide mask dimension at the output end. The wider channel at the input creates a well-confined small mode, while the reduced lateral dimension of the waveguide mask at the output end significantly reduces the total number of protons added to the substrate in the same proton exchange and anneal, and thus provides a waveguide core at the output that has a smaller refractive index. Consequently a much more weakly confined waveguide is produced at the output end of the taper, having a mode that is well matched to that of a single mode optical fiber. If desired, further optimization of the mode sizes in the vertical dimension to maximize mode overlap and coupling efficiency at the input and output may be achieved by the addition of thin film layer sections to the surface ofIO waveguide chip215. For example, a layer of high index Nb₂O₅may be provided at the receiving end to confine the waveguide mode strongly near the surface of the chip; or a layer of SiO₂(or SiON_x), at the output end, to effectively bury the waveguide and force the mode down, away from the surface of the LN crystal, making it more symmetrical. Preferably, the waveguide structure is also designed to be adiabatic, supporting only a single lowest order (0,0) mode throughout its length, so that excess loss due to mode conversion is minimized or eliminated.
• It is known in the art that waveguides with very tight confinement and small mode sizes often exhibit high propagation losses. It is therefore preferable for the tightly confining receiving end section of the waveguide to be as short as possible, and for the waveguide dimensional taper to start very close to the receiving facet. In principle it is possible to reduce the taper length to under 500 μm whilst maintaining low-loss tapering of the waveguide.[0093]
• After fabrication of the waveguide structures, it is desirable to apply well controlled thin film layers to the surface of the wafer corresponding to the top or active surface of[0094]IO waveguide chip215, such that the centroid of the intensity profile of the optical mode ofIO waveguide chip215 is located at the same distance below the effective wafer surface (that is, the surface of the thin film layer deposited on the wafer) as the centroid of the mode of the emitters oflaser array210. Alternatively, the thin film may be applied to merely provide for optimum optical coupling to be achieved irrespective of the optical mode centroids. The required thickness of material may range from 0.1 μm to several microns, preferably controlled to better than 0.1 μm in absolute thickness and in uniformity. A suitable material for this layer is SiO₂, deposited for instance using ion beam assisted deposition which enables exceptionally high uniformity and tight deposition thickness control up to tens of nanometers. The exact layer thickness required to match the centroids of the modes oflaser array210 andIO waveguide chip215 may be determined from experimental measurement or from modeling. The height matching layer described above may also serve the purpose of preventing dirt and debris from coming into direct contact with the surfaces of the optical waveguides.
• In addition to carrying the mode converting tapered waveguide structure,[0095]IO waveguide chip215 may also be provided with alignment and fiducial marks referenced to the waveguide positions, for instance fabricated by planar photolithography and etching of a thin metal layer deposited on the crystal surface. These marks may be used for initial alignment during the attachment process described below. They may also be used to define dicing and polishing markers so that the end faces ofIO waveguide chip215 maybe cut and polished to enable end-face launching of laser light into the waveguide facets. Such cutting and polishing is a routine process that is well understood in the art, with a requirement that the angles of the polished face with respect to both the longitudinal waveguide axes and the active surface be accurately controlled.
• After dicing and polishing, it is preferable that anti-reflection (AR) coatings be provided on the input and output end facets of[0096]IO waveguide chip215. LN has a high refractive index (approximately 2.2) and consequently light incident on a surface of an uncoated LN crystal will experience a significant amount of Fresnel reflectance. To maximize the amount of light coupled into and through the waveguides ofIO waveguide chip215, it is desirable to minimize the amount of reflection. To achieve this objective, an anti-reflective (AR) coating comprising a multi-layer stack of alternating SiO₂and TiO₂layers may be deposited on the appropriate surfaces ofIO waveguide chip215. As is known in the art, the number and thicknesses of the layers in the stack are adjusted according to the refractive index of the substrate being coated. Residual reflections may be thereby reduced to values as low as 0.01-0.1%, depending on the complexity of the coating structure and the optical bandwidth over which it must provide the anti-reflection function. In addition to maximizing the amount of light coupled into and out of the waveguides inIO waveguide chip215, the AR coating also minimizes unwanted back reflections into the emitters oflaser array210, which may otherwise distort the lasing properties and cause unwanted fluctuations in emission wavelength and/or output power.
• The prepared[0097]IO waveguide chip215 may be aligned and attached to thesubmount205/laser array210 subassembly as follows. The submount/laser array subassembly is mounted on a holder and connected to an electrical drive circuit to enable the laser emitters to be energized. This electrical connection may either be temporary with zero-insertion force connectors or the like, or may involve mounting and wirebonding of the submount/laser assembly to some other carrier structure.
• [0098]IO waveguide chip215 is mounted in a specially designed holder and held in place using vacuum force or other suitable technique, with the activesurface facing submount205. Preferably, the chip holder is attached to a six-axis motion control system, such as an autoalign system, providing computer controlled motion and positioning with sub-micron accuracy (≦0.1 μm) in the three orthogonal displacement axes and the three angular axes. Initial alignment ofIO waveguide chip215 tostandoff structure222 is achieved using a combination of an autocollimator to adjust pitch and roll angles, and optical alignment marks to provide longitudinal, lateral and yaw adjustment.
• [0099]IO waveguide chip215 is then brought into close proximity tostandoff structure222, with the receiving facet ofIO waveguide chip215 brought into close proximity to the output or emitting facet oflaser array210 so as to enable butt coupling of the emitted laser light.Laser array210 is energized to emit light by the application of an electrical signal from the above-mentioned drive circuit. At the output end ofIO waveguide chip215 may be located one or more lenses disposed to direct and focus light emitted from one or more waveguides inIO waveguide chip215 to one or more photo detectors configured to monitor the power emitted from the waveguides ofIO waveguide chip215. The position ofIO waveguide chip215 is adjusted using computerized techniques in order to maximize the optical power transmitted through the waveguides. During final coupling optimizationIO waveguide chip215 is preferably contacted tostandoff structure222, allowing final setting of the lateral, longitudinal and yaw dimensions while constraining vertical, pitch and roll motions.
• Note that the holder supporting[0100]IO waveguide chip215 is preferably designed to allow a degree of flex or distortion in the position of the chip in response to pressure applied thereto. In this manner the holder may be designed with roll centers and pivot points such that, asIO waveguide chip215 firstcontacts standoff structure222, it is able to roll or flex sufficiently that itcontacts standoff structure222 substantially uniformly across its entire width, in a similar way as the compliant layer disposed betweenlaser array210 and the flip chip bonder chuck enables uniform contact betweenlaser array210 andstandoff structure222 even in the presence of initial misalignments and bowing or warping of the array itself. This holder may incorporate strain gauges and flexure structures in the mounting arms such that the strain in the holder may be monitored to determine when contact betweenIO waveguide chip215 andstandoff structure222 has occurred.
• Once optical coupling has been optimized,[0101]IO waveguide chip215 must be securely attached to the submount. This function may be performed using a UV curing adhesive, such as an epoxy, in the following manner. First, the position ofIO waveguide chip215 is stored in memory by the computer control system, andIO waveguide chip215 is then withdrawn from contact withsubmount205. An epoxy dispense system may then be moved into position oversubmount205 and controlled volumes of epoxy dispensed into specifiedwells229.IO waveguide chip215 is then repositioned at the stored location, and (if necessary) fine adjustments of the position and orientation ofwaveguide chip215 are effected to re-optimize optical coupling betweenlaser array210 and IO waveguide chip215 (noting that the precise positional repeatability of commercial computer control systems, which is typically around 0.1 μm, generally obviates the need to perform such re-optimization). The epoxy may then be cured using exposure to UV or short wavelength visible light which may be transmitted throughIO waveguide chip215. Those skilled in the art will recognize that it is important to choose an epoxy that exhibits good adhesion to the contact materials, that is the submount, theIO waveguide chip215 surface and/or any thin film materials deposited onIO waveguide chip215 for height adjustment and surface protection. Alternatively, a UV-curing acrylate adhesive may be employed. As another alternative procedure,IO waveguide chip215 may be fastened in place using a solvent-free polymeric adhesive, such as a commercially available solvent-free thermal-curing or UV-curing epoxy. This approach may leave less time to reoptimize alignment (due to generally higher pre-cure viscosity because of the absence of a solvent) but may advantageously reduce or eliminate subsequent outgassing of the adhesive (which may shorten device lifetime if not properly mitigated).
• Alternatively,[0102]IO waveguide chip215 may be fastened in place using solder. In one implementation of this approach, metal pads may be defined photolithographically on the active surface of IO waveguide chip21 (on the surface of any thin film coatings deposited over the waveguides, or on the surface of the LN material itself if no such coatings have been applied). It is noted that the metal pads should lie between the waveguide features so as not to introduce excessive absorption losses. Solder may be deposited and patterned onto these contact pads as known in the art to create solder preforms distributed in some pattern onIO waveguide chip215 matched to respective bonding pads onsubmount205. Preferably, the thickness of the solder as deposited is not sufficient to touch the submount bonding pad whenIO waveguide chip215 is contacted tostandoff structure222, but is sufficient to cause the solder, upon melting, to contact the submount bonding pad thereby producing a robust mechanical joint. In the case of this solder attach approach, after initial alignment optimization the assembly may be heated to the melting point of the solder, which preferably is lower than the melting points of any solders used in eth attachment or wirebonding of the submount/laser assembly. Once the solder has balled up and made contact between the bonding pads on the respective components the temperature is reduced to re-solidify the solder and form the mechanical bond. Preferably the laser is not energized during thermal cycling during solder bonding, as high temperature operation can lead to premature and sometimes instant failure. Note that techniques known in the art to improve solder bonding, such as the use of fluxes, forming gas and formic acid vapor may be applied to improve this process.
• It is noted that a laser welding-based technique may alternatively be utilized to attach[0103]IO waveguide chip215 tosubmount205. Since the LN material ofwaveguide chip215 is substantially transparent to optical energy having wavelengths in the 0.35 μm-4 μm region, a laser beam having a wavelength within this range is able to traverse the material without significant absorption such that it impinges upon and is absorbed by a solder or adhesive target. In this manner an IR laser welding station could be used to attachIO waveguide chip215 tosubmount205. Because this technique does not require heating of large portions ofIO waveguide chip215 and/orsubmount205, its use may be advantageous in connection with component materials having relatively low melting or deformation temperatures.
• The[0104]assembly comprising submount205,laser array210 andIO waveguide chip215 may now be prepared for the alignment and attach of the fiber array, which functions to carry the laser light to the amplification region. It is noted that the task of coupling the fiber array to the mode matched output waveguides ofIO waveguide chip215 requires considerably less precision, relative to the task ofcoupling laser array210 andIO waveguide chip215, due to the significantly greater mode sizes involved. The mode diameters of the output waveguides and the single mode fibers are typically between 3-7 μm. Preferably, the array of fibers is prepared using silicon V-groove technology as known in the art, such that (1) the centers of the optical fiber cores are accurately positioned at the desired pitch to match the output waveguides fromIO waveguide chip215, and; (2) the fiber core centers are aligned along a substantially straight lateral line to match the output waveguide facets ofIO waveguide chip215. It is also preferable that the end faces of the optical fibers be either anti-reflection coated or cleaved/polished at an angle such that the Fresnel reflection from the input fiber core facet is eliminated, or at least not coupled back into the waveguide inIO waveguide chip215, from where it may be coupled back into the emitters ofdiode laser array210 resulting in power and/or wavelength instability in the laser output.
• Active optical alignment and attachment may be performed using any of the methods currently practiced in the art, such as laser welding or UV adhesive bonding. In the case of UV adhesive, the assembly is mounted on a holder and connected to drive circuitry to activate some or all of the diode laser emitters. The light emitted from the diode laser and transmitted through[0105]IO waveguide chip215 exits the output waveguide facets ofIO waveguide chip215.
• The silicon V-groove mounted fiber array is mounted on a multi-axis micropositioner, preferably with computer control. The fiber array is brought into proximity with the output facet of[0106]IO waveguide chip215 and at least coarsely aligned with respect to lithographically defined fiducial or alignment marks disposed onIO waveguide chip215. Light coupled into the fibers of the array is monitored by one or more photodiodes or power meters disposed at the output of the fiber array. Preferably the signal from the monitoring photodiodes is used by the computer control system to adjust the position of the fiber array until maximum optical coupling, and therefore transmitted signal, is achieved.
• Epoxy may then be dispensed either around the silicon V-groove fiber holder and/or between the fibers and the end facet of IO waveguide chip as desired. Bonding of the silicon V-groove fiber holder may be achieved by exposing the epoxy to UV light. Preferably the epoxy is chosen to exhibit little shrinkage on cure such that the position of the fiber array is not significantly altered by the curing process. In some instances it may be preferable to have an epoxy free optical path, that is to ensure that no epoxy is located between the output waveguide facets of[0107]IO waveguide chip215 and the input facets of the optical fibers in the array. In such instances the epoxy should be dispensed only in side areas free of waveguide facets. In this embodiment where the fiber array is bonded directly to the output facet ofIO waveguide chip215, it may be preferable to bond a spacer piece of LN material to the top surface ofIO waveguide chip215 at the output end prior to the cut and polish processes described above. In this way, the end faces ofIO waveguide chip215 and the spacer LN piece are cut and polished simultaneously, effectively forming a single surface which spans both above and below the waveguide output facet. In this case, it is preferable that submount205 does not extend to the output end of the LN chip as the spacer LN piece would interfere with the flip chip bonding process. In other embodiments described below it may be important thatsubmount205 extends at least up to and sometimes beyond the output facet of the LN chip. The spacer LN piece described above serves to provide a larger mounting surface for the attachment of the fiber array, which in turn leads to a more robust and reliable mechanical joint and improved stability and lifetime of the optical alignment between the two structures.
• Once the attachment of the fiber array is performed, the diode laser pump module assembly is completed by the provision of an external protective package such as that shown in FIG. 21, and suitable electrical and control connections are provided to enable the desired laser operation as known in the art. Preferably the diode laser pump source may be enclosed in a hermetically sealed butterfly type package as known in the art.[0108]
• It should be noted that elements of the external package may be incorporated into the above described process flow at any point without interfering with the performance of this invention. For instance, it may be preferable to place and mount the submount/laser diode array/IO waveguide chip assembly inside the open butterfly package before performing alignment and attach of the fiber array, or vice versa. In addition, it will be apparent that, for example, the output ends of the fibers may be attached to the amplifier(s) by connectorizing or fusion splicing.[0109]
• FIG. 5 depicts an[0110]amplifier pump source500 in accordance with a second embodiment of the present invention, wherein increased functionality is provided to pumpsource500 by the inclusion of distributed Bragg reflectors (DBRs)505 in the form of Bragg gratings. For the purpose of clarity, the submount has been omitted from FIG. 5, andlaser array210 andoptical fiber array220 are depicted in phantom.DBRs505 are preferably superimposed on the optical waveguide structures507 ofIO waveguide chip215 and are configured to provide narrow-band, wavelength-selective retroreflection. The light reflected by eachDBR505 is fed back into a corresponding emitter oflaser array210 for the purpose of stabilizing the wavelength of emitted light, in accordance with processes well established in the prior art. The wavelength of light reflected by eachDBR505 is determined by its grating period. It should be noted that the grating period may be varied among theseveral DBRs505 fabricated onIO waveguide chip215, thereby enabling different emitters oflaser array210 to be locked to different wavelengths. With the appropriate choice of DBR bandwidth (determined by the length of the grating), and reflectivity (determined by a combination of the grating order, effective depth, and length), the light output by each emitter can be locked to within the bandwidth of the associatedDBR505 in a manner largely independent of the operating conditions oflaser array210, specifically drive power and ambient temperature. Typically a reflection back into the emitter of a few percent of the total emitter output power is sufficient to lock the emitter output to the wavelength band ofDBR505. The output facet of thelaser array210 may have an appropriate low reflectivity coating, again in the order of a few percent, while the input facet ofIO waveguide chip215 should preferably have a very lowreflectivity anti-reflection coating510 such that the broadband (in wavelength terms) reflectivity from the input facet does not destabilize the emitters oflaser array210. The output facet ofIO waveguide chip215 should also preferably have ananti-reflection coating515 to suppress reflection.
• DBRs[0111]505 may be fabricated using planar lithographic processing methods known in the art. For example, a fine pitch grating suitable for retroreflection of ˜980 nm light in a lithium niobate waveguide requires a period of around 230 nm for a first order structure. This period may be created in a thin photoresist layer (˜0.1 μm) using two-laser-beam holographic interference lithography or using a contact printing approach with a phase mask illuminated either with a UV laser or a mercury lamp. Development of the exposed resist leaves a grating pattern which may be transferred directly into the lithium niobate, or into some surface layer on the lithium niobate, using an approach such as ion beam milling. Alternatively reactive ion etching or laser ablation could be used for pattern transfer. The depth of the grating and thus the reflectivity per grating line is defined largely by the depth, which is simply controlled by the etch time. Different periods may be fabricated using a sequential exposure and etch process with shadow mask material such as chrome covering the areas which do not require a grating of a particular period. Alternatively, using the phase mask exposure process, a number of different periods can be defined simultaneously using a custom designed and fabricated phase mask containing the relevant patterns for transfer. This grating fabrication processing for IO chips would typically be performed after waveguide fabrication on a wafer but before surface protection coating and chip dice (separation into chips) and end face polish. The exact design ofDBRs505 depends heavily on the precise waveguide structure and overlayer/protection materials used and must be adjusted accordingly based on existing modeling capabilities combined with empirical measurements ofDBR505 performance after fabrication.
• In FIG. 5, different locations of[0112]DBRs505 are shown ondifferent waveguides520,525. The placements ofDBRs505 illustrates that the design of the DBR and the design of the waveguide are linked. Onnear waveguide520,DBR505 is shown disposed on the highly confininginput waveguide section530 that supports a very small mode to provide good coupling efficiency to the emitter output mode. In this case the grating needs only to be very shallow in order to achieve a desired low reflectivity as a shallow structure interacts strongly with the small, tightly confined optical mode. However, for some waveguide fabrication processes, the tightly confined input waveguide section may exhibit relatively high propagation losses, making it preferable that the input section be as short as possible. In this case it is preferable to placeDBR505 as shown on thefar waveguide525, superimposed on a lower loss, more loosely confiningwaveguide section535. Preferablywaveguide section535 is intermediate in mode size and confinement between the tight input section and the loose output section which requires a grating of greater depth in order to interact efficiently with the relatively large and weakly guided optical mode which is dimensioned to match efficiently to a single mode optical fiber.
• Alternatively[0113]DBR505 may be the conventional type known in the art formed by lithographically patterned UV exposure of the core and cladding of an optical fiber, and in this case may be located in each of the fibers of output fiber array. In other embodiments, the grating may be superimposed over the taper to increase the bandwidth of the grating, which may be desirable to optimize laser stabilization.
• Another embodiment of the present invention, shown in FIG. 6, adds further functionality to laser array amplifier pump source[0114]600 by increasing the output power coupled into asingle fiber620 by means of wavelength multiplexing the output of several laser emitters from thelaser array610. Increasing the output power infiber620 increases the available performance of an amplifier pumped by laser array amplifier pump source600, while still allowing the individual emitters oflaser array610 to be operated at output powers well below the threshold for failure due to catastrophic facet damage or other effects. Using a set of narrow-band DBRs605, a number of diode laser emitters are stabilized to different wavelengths, separated by a known wavelength interval and defined by the periods of theDBR605 gratings. Preferably all the emitted wavelengths lie within the absorption spectrum of the amplifier material of application, such that they may usefully contribute to the amplification of signal light. For application in EDFAs the emitted wavelengths are preferably within about 15 nm of the peak of the absorption spectrum located near 975 nm. The different wavelength outputs from the laser emitters are then combined using fused-fiber-opticdirectional couplers625 which provide a wavelength selective coupling function and enable the power in several wavelength channels to be combined to asingle fiber620. The design off used fiber directional couplers must be matched to the emitter wavelengths and their spacing. Such fused fiber WDM couplers are commercially available with wavelengths separations down to ˜2 nm to combine 4 or more channels to a single fiber.
• FIG. 7 illustrates an embodiment of the invention similar to the FIG. 6 embodiment but distinguished therefrom by its use of integrated optic waveguide directional couplers[0115]710 (in place of the fused fiber directional couplers of the FIG. 6 embodiment) to perform the wavelength multiplexing operation on theIO waveguide chip715.Directional couplers710 maybe fabricated by defining two optical channel waveguides in close proximity to create a coupling region. In LN, waveguidedirectional couplers710 may be designed using existing modeling capabilities such as beam propagation modeling (BPM) and knowledge of the properties of annealed proton exchange waveguides. The length of the coupling region ofdirectional couplers710 is known to be inversely proportional to the wavelength separation of the two channels to be combined. In FIG. 7 the optical waveguides are shown to have two taper regions: afirst taper region725 situated upstream in the optical path relative todirectional couplers710, and asecond taper region735 situated downstream in the optical path relative to thedirectional couplers710. This arrangement enables independent optimization of the waveguides in the three main sections of the chip input, directional couplers, and output. In this way the waveguide dimensions and confinement can be optimized to achieve high input coupling efficiency fromlaser array705, minimum directional coupler length with low loss within theIO waveguide chip715, and high output coupling efficiency to single modeoptical fiber720. The number of channels which can be practically multiplexed together in this way is limited by the length of the directional couplers.
• For example, a simple coupler designed to multiplex two wavelengths around 980 nm separated by ˜5 nm may have a length of ˜4 mm, 8 mm or 19 mm, for example.[0116]
• FIGS. 20[0117]a-ddepict still further embodiments of the invention, wherein wavelength selective feedback to stabilize the individual emitters of the laser diode array is provided after, and by, one or more wavelength multiplexing directional couplers that combine the outputs of several emitters of the array, as described in detail below.
• The use of narrow-band DBRs for stabilizing diode lasers was described above with reference to FIG. 5. Use of fused fiber couplers and, alternatively, IO waveguide directional couplers, for combining the output of two or more emitters operating at different wavelengths, also called wavelength multiplexing, was described in connection with FIGS. 6 and 7. In those embodiments, the DBRs are positioned upstream in the optical path from the wavelength multiplexing directional couplers as shown in FIG. 20[0118]a. Referring to FIG. 20a, two emitters of a laserdiode array chip2210 are butt-coupled to waveguides in anIO waveguide chip2215. Twowaveguides2231 and2232 receive light from the emitters, and are adapted with DBRs2201 and2202, followed by a waveguidedirectional coupler2240. DBRs2201 and2202 stabilize the two emitters at two different wavelengths λ₁and λ₂. The outputs at the two wavelengths are combined (multiplexed) indirectional coupler2240 and propagate in asingle waveguide2231 to the output facet of the IO waveguide chip, where the combined light beam containing energy at two different wavelengths couples into an outputoptical fiber2220.
• In the embodiments depicted by FIGS. 20[0119]b-c,the stabilizing DBRs are positioned after the directional couplers, in series, one after another. In a first alternative shown in FIG. 20b, twoDBRs2201 and2202 are disposed in series onwaveguide2231, after waveguidedirectional coupler2240, onIO chip2215. According to known art, the waveguidedirectional coupler2240 transmits, in the reverse direction, light of wavelength λ₁preferentially in one input waveguide arm, such as2231, and light of wavelength λ₂preferentially in the other input waveguide arm, such as2232. It should be noted that not all waveguide directional couplers have this property, and therefore this embodiment requires use of appropriate design for waveguide directional couplers to ensure their functioning in the desired manner. In a second alternative, shown in FIG. 20c, two stabilizingDBRs2251 and2252 constructed in the form of fiber Bragg gratings (FBGs), as known in the art, are positioned in series onoutput fiber2220. In a third alternative (not shown) one stabilizing DBR is positioned onwaveguide2231 afterdirectional coupler2240 onIO waveguide chip2215, and the other stabilizing DBR is positioned onoutput fiber2220 coupled to waveguide 2231.20d The DBRs in the alternative positions, two of which are shown in FIGS. 20b-c, are effectively in series in the multiplexed light output path and function in substantially the same manner described in connection with FIG. 20b, whether they are on the IO chip or on the output fiber. Grating fabrication processes in optical fiber may be a more mature technology than grating fabrication processes in IO materials, and thus may provide an economic advantage in some applications. Positioning the DBRs on optical fiber rather than on the IO chip allows wavelength selection of the emitters of a laser diode array independently of wavelength multiplexer design, and this may also provide an advantage in some cases.
• In a fourth alternative form of this embodiment shown in FIG. 20[0120]d, feedback for stabilizing the laser diode array emitters in a wavelength multiplexed configuration is provided after the waveguide directional couplers, in a position substantially similar to those shown in FIGS. 20b-cfor narrow-band DBRs, by a weak spectrally-wide reflector. Preferably, such a reflector is awideband DBR2203 disposed on theoptical waveguide2231, between thedirectional coupler2240 and theoutput facet2217 of theIO waveguide chip2215, and reflections originating from further downstream in the optical path are suppressed by appropriate techniques known in the art, such as by adapting the end of theoptical fiber2220 with a lens or chisel-tip shape and providing anti-reflective (AR) coating thereon. Wavelength selectivity to stabilize the emitters to particular wavelengths is provided by the waveguidedirectional coupler2240 acting together with the wideband reflector. Maximum feedback occurs at the wavelengths of preferential reverse transmission in the waveguide directional coupler, which is appropriately designed with sufficient wavelength selectivity, according to known art. Preferential reverse transmission to oneinput waveguide2231 of the directional coupler occurs in narrow bands peaked at a first set of wavelengths λ₁, λ₃, . . . , and to theother input waveguide2232, at a second set of wavelengths λ₂, λ₄, . . . , which alternate with the first set of wavelengths such that λ₂is intermediate λ₁and λ₃, λ₃is intermediate λ₂and λ₄, and so on. The optical reflection band ofDBR2203 is adapted to lie within the open-loop optical gain band of the laser diode array emitters, and to be sufficiently wide to reflect two adjacent preferential reverse transmission wavelengths (e.g., λ₁and λ₂), one for each of the two input waveguides of thedirectional coupler2240, but to substantially transmit and not reflect the other wavelengths of both sets. Therefore the emitters of thelaser diode array2210 that are coupled to these waveguides are stabilized at the wavelengths λ₁and λ₂, respectively. Alternatively, thewideband DBR2203 may be omitted, and wideband reflection may be provided by a thin film stack optical reflector, with its reflection band and operation substantially similar to that described above forDBR2203, disposed on theoutput facet2217 of IO waveguide chip2215 (in place of an AR coating). Further alternatively, the wideband DBR or thin film stack reflector maybe omitted and wide band reflection may be provided by an uncoated or suitably coated output facet of the IO waveguide chip or the input facet of a cleaved output fiber, and wide-band selectivity provided by adapting the laser diode emitters ofarray2210 to have an open-loop optical gain spectrum that is appropriately wide to contain only the wavelengths λ₁and λ₂, thereby selecting only these directional coupler retroreflection wavelengths for lasing, one for each emitter. This structure and method to employ the narrow-band wavelength selectivity of optical waveguide directional couplers to stabilize the laser emitters is not known in prior art, and provides the considerable advantage of removing the need to have narrow-band DBRs and to match their spectral transmission characteristics to the wavelength multiplexing directional couplers.
• FIG. 8 shows a still further embodiment of the present invention which provides for increasing the output power of the laser array amplifier pump source that is coupled into a single fiber output. As depicted, fiber optic polarization-maintaining (PM) combiners or[0121]multiplexers825 are used to combine the outputs from pairs ofadjacent waveguides802 and804 of theIO waveguide chip815 intosingle output fibers835. Here the output ofwaveguides802 and804 onIO chip815 are coupled into polarization preserving optical fibers, the output fromwaveguide802 coupling intooptical fiber842 and the output fromwaveguide804 coupling intofiber844. Alternate fibers have been twisted by 90° so that the outputs from adjacent waveguides couple into orthogonal polarizations in alternate fibers, as indicated symbolically by alternating butterfly shading at the ends of the fibers. In general, all of the outputs from alaser array810 andIO waveguide chip815 are polarized in the same direction, parallel to the plane of the diode junction. Owing to the twist as indicated by circular arrows in FIG. 8, the polarization of the light becomes physically rotated 90° along with the fiber in alternate fibers, such asfiber844, before joiningpolarization multiplexer825. The physical900 rotation or twist of alternate fibers may be readily accomplished by alignment when the polarization preserving fibers are fabricated into an array using a silicon V-groove holder. Thus the outputs of two adjacent waveguides may be combined onto the same output fiber but in orthogonal polarizations to effectively increase, by almost a factor of two, the output power in the single fiber, with the penalty of only a small loss associated with the polarization multiplexer itself. Polarization multiplexers capable of combining orthogonally polarized light around 980 nm are known in the art and are commercial available.
• It is noted that the application of the polarization multiplexer may be independent of, or combined with, the wavelength multiplexing of the FIG. embodiment described above in order to further increase the power coupled into an output fiber. It is further noted that the functions of polarization rotation and combination/multiplexing can be performed on the integrated optics chip itself. In LN, it is known to be possible to fabricate frequency selective polarization rotators or TE-TM converters, as well as broadband polarization combiners. From these basic building blocks it is possible to construct a multiple stage polarization combiner. In such a device it is necessary to fabricate a waveguide structure that supports both TE and TM polarization modes, whereas the commonly employed annealed proton exchange(APE) process supports only extraordinarily polarized modes (TE in X and Y-cut LN, TM in Z-cut LN). Other waveguide fabrication techniques in LN include metal indiffusion, which generally creates a polarization insensitive waveguide (although in some instances, such as titanium indiffusion, it often creates waveguides that are susceptible to photorefractive damage and have low power handling capability). A suitable fabrication process to achieve the integrated polarization multiplexer may be zinc indiffusion, which has been shown to produce low loss, polarization-insensitive and photorefractively robust optical waveguides. A transition region from the input, tightly confined annealed proton exchange waveguide is required to match the input light to the relatively larger mode dimensions of the waveguide formed by zinc-indiffusion, which then matches well to the single mode optical fiber output. Alternatively, waveguides formed by titanium indiffusion may be employed, but may be subject to the operational problems alluded to above.[0122]
• It is further noted that multiplexed laser array pump sources are preferably designed such that the specified maximum optical output power in each output fiber may be achieved at less than the maximum possible drive current applied to each laser emitter. Thus, should one or more individual laser emitters fail or suffer from reduced power output in any of the emitter-multiplexed embodiments described above, the remaining fully operational emitter(s) may be driven to produce higher output in order to compensate and maintain the overall design output power without over-driving individual elements, which otherwise would significantly increase the chances of further emitter failures due to junction or facet damage.[0123]
• FIG. 9 depicts another embodiment of the present invention, which adds detectors to monitor the output power of the laser emitters, allowing control of the pump source output in response to control signals. FIG. 9 illustrates several possible alternative locations for the detectors, which preferably take the form of photodiodes sensitive at the emission wavelength of the laser array, e.g., silicon photodiodes which are sensitive at around 980 nm. Detectors in[0124]location940 preferably represent an array of silicon photodiodes, disposed near the back facet of the emitters oflaser array910, and aligned to receive radiation emitted from the back facet on a one-to one basis, wherein each emitter is uniquely associated with one photodiode. The detector array may be monolithic or it may comprise individual photodiode chips. The photodiodes may be attached to the submount either prior or subsequent to the laser array, provided that appropriate care is taken to select processes that do not interfere with those described above for the mounting ofdiode laser array910 andIO waveguide chip915. If required, the submount may incorporate features such as turning mirrors located adjacent to the back facet oflaser array910 to redirect the light emitted from the back facet perpendicular to the plane of the major surface of the submount. In this case the photodiode array may be mounted over the turning mirrors (either face down or in a back-lit configuration) to receive the light emitted from the back facet, and secured in place without contact with any electrical interconnection traces disposed on the submount surface for activation of thelaser array910 emitters. Electrical contact to the photodiode array may be made by solder joints or wire bonding or conductive epoxy. The output signals from the photodiode array are preferably transferred to electrical connection traces on the submount surface and from there may be wirebonded to output pins in the external package for connection to the control electronics andlaser drivers970, as indicated symbolically by the dashedline944. Alternatively, the output signal may be directly wirebonded to traces in the external package. The actual optical output power in each channel may be calibrated with respect to the photodiode signal using a power meter to measure the output power in each output fiber, and the calibration constants stored in memory within the driver circuitry.
• The detectors may alternatively be disposed at[0125]location950. Detectors inlocation950 preferably represent an arrangement whereby the optical power in each output channel of the pump source is monitored via a structure incorporated intoIO chip915. FIGS. 10 and 11 show an embodiment of an integrated power monitoring array incorporated into the present invention. In this embodiment the power in each channel is monitored by aphotodiode1052 disposed to receive light scattered out ofoptical waveguide1080 by a distributed Bragg reflection (feedback) grating1056. Alternatively, a waveguide discontinuity, pit or other suitable scattering structure which extends into the evanescent tail of the waveguide mode may be substituted for grating1056. Some scattering is inherent in the practical implementation of a surface etched relief grating (as described in an embodiment above), even for a first order grating. The proportion of the input beam that is scattered out fromwaveguide1080 is dependent on the mode intensity profile that interacts with the grating features. As the mode intensity profile does not change with power, the scattering proportion should be constant, independent of input power. Thus a measure of the scattered light may be used as an indicator of the power incident (and transmitted through) the grating. Alternatively, the grating structure may be designed specifically to provide an out of plane diffracted beam, for instance as a second or third order grating, as is known in the art.
• The photodiodes, which may be discrete devices or a monolithic array, are disposed to receive the light scattered or diffracted from the waveguide by the grating. They may for instance be bonded face down to the backside of the integrated optic chip at[0126]1152, located substantially over the grating regions on the front face, as shown symbolically in FIG. 11. Alternatively the photodiodes may be located at1153, recessed at least partially inwells1110 in thesubmount1105. These wells may be fabricated by lithographic pattering and etching during the submount fabrication process. The detectors (or detector array) may be mounted into the well using solder or conductive epoxy and may be either front illuminated or back lit. Alternatively the photodiodes may be mounted directly to the front face of the integrated optics chip. The electrical signals from the photodiodes may be carried by conductive traces defined over the surface of the submount to transfer the signals to the edge of the submount for connection to output pins in the external package, and from there to the control electronics andlaser drivers970 identified in FIG. 9 as indicated symbolically by954. The mechanical and electrical attach processes must of course be compatible with the processing outlined earlier for the laser array and integrated optic chip attach.
• FIG. 12 illustrates an alternate embodiment of an integrated power monitoring array with photodiode[0127]power monitoring detectors1252 located at the outputs in integrated optic directional coupler taps1210. The directional coupler taps are preferably designed to couple only a small proportion (preferably around 2-4%) of the power out of the primary waveguides. Couplers taps1210 for this purpose may be designed in a similar manner to those described above for wavelength multiplexing of the laser emitter outputs. If desired, a scattering site or out-of-plane reflector may be fabricated at the end of the directional coupler, e.g. by laser ablation, to more efficiently deflect the light into the photodiode, which may be located on or near either the back or front surface ofintegrated optics chip1215. As described above with reference to FIG. 11, the photodiodes (or monolithic array of photodiodes) may be located at least partially within wells in the submount, under the front surface of the IO waveguide chip.
• Detectors in[0128]location960 in FIG. 9 may represent a number of photodiodes illuminated by light from an array of fiber-optic taps980, each coupling a small percentage (typically 2-4%) of the light from a respective one of the fibers in the output fiber array. Such taps are commercially available and widely used for monitoring of fiber optic signals as their coupling ratio (or proportion of power removed from the main fiber) is constant with wavelength and power, and is largely undisturbed by fluctuations in environmental conditions. The fiber taps may be located inside or outside the overall package of the laser array amplifier pump source and may be butt coupled directly to monitor photodiodes either individually or as an array.
• FIG. 13 depicts another embodiment, wherein further functionality is added to the laser array pump source in the form of redundancy to provide protection against failure of individual laser emitters. In this embodiment supernumerary (extra)[0129]laser emitters1314 are fabricated in thelaser array1310. For example, if the laser array pump source is designed to have 8output fibers1322 accepting light from 8regular laser emitters1312, an extra 2-4 laser emitters may be provided in the array. These extra emitters are coupled toextra waveguides1334 inIO chip1315 simultaneously with theregular emitters1312 as described above. The extra waveguides in the IO waveguide chip may be provided with all the functionality of theregular waveguides1332, e.g. DBR, monitor photodiode, integrated optic directional coupler tap, etc. From the output of the IO waveguide chip the extra waveguides are coupled intoextra fibers1324 located in the output optical fiber array simultaneously with the regular output waveguides and fibers as described above.
• In operation of the laser array pump source, the regular laser emitters are energized to provide optical output power in the regular output fibers for transmission to the amplifier region. Should one of the regular laser emitters fail, for instance due to defect induced failure of the laser diode junction or catastrophic optical damage at the output facet, one of the extra emitters may be energized and the extra output fiber coupled to the now energized extra emitter connected to the amplifier region in place of the output fiber coupled to the failed laser emitter. The act of connection of the extra output fiber to the amplifier region may take the form of a fiber fusion splice or the use of fiber optic connectors as known in the art. Note that preferably the diode laser emitters in the laser array are separated by sufficient lateral distance such that a defect occurring in one emitter junction or on the output facet of one emitter does not substantially affect the operation of adjacent emitters.[0130]
• FIG. 14 illustrates a further improvement to the present invention providing the addition of a switch network or[0131]fabric1425 in theoutput fiber array1420 to provide redundancy. This switch network enables dynamic and remote reconfiguration of the output from the laser array emitters into the fibers in the output fiber array in response to control signals. Thus, should aregular emitter1412 on thelaser diode array1410 fail during operation of the laser array pump source, the output from one of theextra emitters1414 may be switched into the appropriate output fiber for transmission to the amplifier region. The act of switching may preferably be accomplished using a control signal to activate a desired switch from a remote location, allowing computer control of the laser pump source. Such switching control may make use of output channel power monitors to determine whether a particular laser emitter is operating and providing power to a particular amplifier region, and the switching control signals may be provided directly by the laser array pump source drive and control electronics. The switching network may be composed of components known in the art such as thermo-optic switches, opto-mechanical fiber switches or micro-electro-mechanical (MEMs) switches. Redundancy structures and techniques may also be employed during the manufacturing process to deselect or reroute inoperative or poorly performing channels and thereby increase device yields. See U.S. Pat. No. 6,049,641, incorporated by reference herein.
• FIG. 15 shows an alternate embodiment of the present invention where an optical switch fabric or[0132]network1525 is integrated into a waveguide network on anintegrated optic chip1515. The integrated switch network in the integrated optic chip may be comprised of thermo-optic or electro-optic switches as known in the art, such as total internal reflection switches or switched directional couplers, which respond to a control signal in order to switch light between predetermined paths.
• FIG. 16 shows an alternate embodiment where protection against laser emitter failure is provided by a[0133]passive network1600 coupled to theoutput fiber array220 identified in FIG. 2. Thepassive network1600 preferably comprises a series of 50-50directional couplers1670 andpower splitters1680 which share the power entering theinput fibers1625 of the network, among the fibers in theoutput fiber array1635. The passive power splitter and coupler network is preferably composed of fused fiber type components that offer efficient power splitting with very little excess loss which would otherwise decrease the available output power from the pump source.
• With the incorporation of the passive network, the laser array pump source output becomes substantially tolerant of the failure of individual emitters. The failure of a single emitter simply decreases the power in each[0134]output fiber1635 by a small amount as a consequence of the power sharing provided by the passive network, rather than resulting in a total loss of power in a single output fiber which results from a single emitter failure in the absence of either the passive network or another redundancy arrangement.
• Note that in all of these redundancy and failure protection embodiments the phrases “extra emitters”, “extrawaveguides” and “extra fibers” may refer to single elements, or if desired, sets of elements, such as the wavelength multiplexed elements of the embodiment described above. Thus if one emitter from the wavelength multiplexed set of emitters should fail, a complete new set of emitters may if desired, be energized and switched to the appropriate output fiber.[0135]
• FIG. 17 shows yet another embodiment of a pump source[0136]1700 according to the present invention, which provides for a different approach to the attachment of the output fiber array. Asubmount1705 carrying alaser array1710 and an IOwaveguide array chip1715 is preferably fabricated from single crystal silicon. Towards the output end of the submount are fabricated v-grooves1740 as known in the art for alignment ofoptical fibers1720. Such v-grooves are fabricated using photolithographic exposure and patterning of a suitable mask material disposed on the surface of the silicon submount wafer, e.g. silicon dioxide, followed by wet etching of the single crystal silicon. The orientation of the silicon wafer is selected appropriately, for example 100/110. The etching is typically performed using potassium hydroxide (KOH), which provides a selective etching capability such that the wet etching process produces v-grooves1740. The width of v-groove1740 at the surface of the wafer is very accurately defined by the width of the lithographically patterned mask. Consequently the position of an optical fiber resting in v-groove1740 is accurately defined in both the lateral and vertical dimensions via the lithographic patterning process combined with the selective or preferential etching performance of the single crystal material. Preferably the width of v-groove1740 is defined such that the center of the core ofoptical fiber1720 is positioned substantially at the same vertical position above the aforementioned reference surface as the optical waveguide output facet of the IOwaveguide array chip1715.
• The alignment and attachment process of the laser array pump source is modified from that described above as follows. During the attachment of the laser array,[0137]laser array chip1710 is accurately aligned relative toalignment marks1735 defined on the surface of the submount, which are themselves accurately located with respect to the etched v-grooves. The alignment oflaser array chip1710 are performed accurately in the lateral and yaw angle dimensions such that the multiple laser emitters are substantially centered on theaxis1745 of respective v-grooves1740. The longitudinal alignment may be less precise as the length ofIO chip1715 and the longitudinal extent of v-grooves1740 are preferably chosen such that there is some overlap of the v-groove1740 andIO chip1715 at theoutput end1717 ofIO chip1715.
• In the preparation of[0138]IO chip1715 it is important to ensure that the end faces of the chip are polished accurately perpendicular to the axes of waveguides in the lateral dimension.IO chip1715 is aligned to the diode laser and attached to the submount substantially as described above. The yaw angle alignment is particularly important to ensure that the output waveguides ofIO chip1715 are substantially centered over the axes of respective v-grooves1740.
• The output optical fibers may then be aligned and attached using v-[0139]grooves1740 fabricated in the output end of the submount. Several possible techniques exist for aligningfibers1720. Eachfiber1720 may be individually prepared either with a perpendicularly cleaved and anti-reflection coated end face or with a 4 or 8 degree angle cleave or polish to suppress back reflections. Theindividual fibers1720 may then be placed in respective v-grooves1740 and adjusted for optimum longitudinal position by monitoring the light coupled into the fiber from the activated laser emitter(s) using a photodiode or power meter. The fibers may then be mechanically attached to the submount using, for instance, solder or UV or thermally cured epoxy. Alternatively, the fibers may be pre-fabricated into an array and aligned and assembled to the submount in a single process step.
• The use of the above described v-[0140]grooves1740 integrated intosubmount1705 for alignment ofoptical fibers1720 is complicated by the precise lateral and angular positioning accuracy required during the mounting of the first component (laser array1710) tosubmount1705 to ensure that the laser emitters are substantially aligned to the centers of respective v-grooves1720. In addition, the length ofIO chip1715 is potentially limited by the thermal expansion coefficient mismatch between the submount and IO chip materials. If the expansion mismatch is sufficiently great andIO chip1715 sufficiently long, the differential expansion generated during operation of pump source1700 may be severe enough to fracture the mechanical bonds betweensubmount1705 andIO chip1715.
• FIGS. 18[0141]a-cshow another embodiment of a pump source1800 according to the present invention. This embodiment avoids the differential expansion problems noted above in connection with the FIG. 17 embodiment and enables packaging of relatively long IO chips having a thermal expansion coefficient which differs from that of the submount. Alaser array1810 andIO waveguide chip1815 are aligned and attached using a reference surface defined by the array ofstandoffs1830 on afirst submount1805. As shown in FIGS. 18aand18b, a second andseparate submount1806 is used at the output end ofIO waveguide chip1815 to facilitate alignment and attachment of an outputoptical fiber array1820. Thesecond submount1806 is aligned toIO waveguide chip1815 using lithographically defined alignment and/or fiducial marks.Second submount1806 may be attached using substantially the same techniques as described previously for attaching the IO waveguide chip to the laser array/submount sub-assembly.Second submount1806 is provided with a standoff structure which may comprise a second array of standoff features1836 defining a second reference surface to whichIO waveguide chip1815 is contacted and attached. The second submount is also preferably adapted with a set ofrelief slots1840, the centers of which are substantially aligned laterally with the centers of respective output waveguides inIO waveguide chip1815 after the attachment process is complete.Relief slots1840, which may for example be fabricated in single crystal silicon using wet etching or deep RIE processing, are preferably large enough to accept a standard single mode optical fiber cladding without contact with the side walls or bottom of the slot.Relief slots1840 are preferably defined such that they extend under the output end ofIO waveguide chip1815, as shown in FIG. 18a.
• The[0142]output fiber array1820 is preferably fabricated using a single piece of v-groove silicon1824, as shown in FIG. 18b. Such an assembly may for instance be fabricated by bonding thefibers1822 into single v-groove piece1824, followed by the temporary attachment of a second piece of v-groove silicon1826 (indicated in phantom) to provide a “sandwich” around the fibers to support them during a standard optical polishing process, as known in the art. After completion of polishing, which may be perpendicular or at some predefined angle to the optical axis offibers1822, anti-reflection coatings may be applied if desired and the temporarily attached second piece of v-groove silicon1826 removed to leave the structure of FIG. 18b. Preferably the v-grooves are fabricated such that their lithographically defined width places the fiber core the same vertical distance from the surface of the silicon v-groove wafer as the IO waveguide chip optical waveguide is located beneath the surface of the IO waveguide chip.
• The fiber assembly may then be mounted on a computer controlled stage as described earlier and positioned such that the[0143]fibers1822 face therelief slots1840 fabricated insecond submount1806. The v-groove silicon is then brought down into contact with the second reference surface defined by second array of standoff features1836 onsecond submount1806. Using the techniques described earlier, substantially uniform contact between the silicon v-groove material and the reference surface is achieved, substantially aligning the axes10 waveguide and the optical fiber in the vertical dimension. After optimization of the longitudinal and lateral positions of the fiber array relative toIO waveguide chip1815, the two components may be mechanically attached, for example, using V cured epoxy or solder to produce the assembly shown in FIG. 18c. Those skilled in the art will recognize that care must be taken to ensure that the attachment methods used in subsequent steps in the assembly process are compatible with prior steps; for, example, the melting temperature of a solder used in a subsequent step should be lower than that used in a prior step in order to prevent undesired melting and reflow of solder employed for attachment in the prior step.
• FIGS. 19[0144]aand19bdepict another embodiment of a pump source1900 in accordance with the present invention, which utilizes a modification of the single sided silicon v-groove mounted fiber array depicted in FIGS. 18a-c. In contradistinction to the previous embodiment, wherein the optical fibers are polished normal to their longitudinal axes (or at a shallow angle thereto in order to suppress back reflections),optical fibers1922 may be polished at a relatively sharp angle from two opposing sides to form chisel shaped fiber ends1918, with the chisel shape preferably substantially centered on and symmetric about the center of acore1928 ofoptical fiber1922 as shown in plan and side views in FIG. 19a. Again, the polishing process may be facilitated by the addition of a temporarily attached second silicon v-groove piece, which is subsequently removed after polishing. The lithographically defined width of the v-grooves is preferably chosen such that the centers of thecores1928 ofoptical fibers1922 are accurately located at the same vertical distance from the surface of the v-groove silicon wafer as the laser array emitters are located beneath the major surface of the diode laser wafer. The separation of the v-grooves is chosen to match the lateral spacing of the emitters inlaser array1910.
• [0145]Laser array1910 is coupled directly to anarray1920 of chisel or lens endedfibers1922 assembled in the single-sided v-groove holder1924, omitting the IO waveguide chip shown in the above embodiments while still providing simultaneous coupling of all the emitters inlaser array1910 using astandoff structure1930 defining a reference surface to ensure accurate, passive, alignment in the critical vertical dimension.Submount1905 may also be adapted withrelief slots1940 as described above in connection with FIGS. 18aand18c. In the presently described embodiment it is preferable thatlaser array1910 is mounted using the techniques described above, and accurately positioned such that the emitters are substantially centered with respect torelief slots1940. In the longitudinal dimension,laser array1910 is preferably located in close proximity to the proximal end of thecorresponding relief slot1940 such that the slot does not substantially undercut the area ofarray1910 that would otherwise compromise the electrical, mechanical and thermal properties of the bond betweenlaser array1910 andsubmount1905.
• Optical alignment in this embodiment is preferably performed by mounting v-[0146]groove fiber array1920 on a computer controlled micropositioner withfibers1922 facingrelief slots1940 insubmount1905. V-groove fiber array1920 is brought substantially into uniform contact with the reference surface using the techniques described above. The lateral and longitudinal positions offiber array1920 are then optimized to maximize the power coupled from the energized laser emitters tooptical fibers1922. After final alignment, mechanical attachment offiber array1922 to submount1905 may be achieved using solder, thermally or UV cured epoxy, or other suitable adhesive to yield pump source1900, as shown in FIG. 19b.
• FIG. 21 is a block diagram depicting yet another embodiment of the present invention, which provides for pumping of multiple EDFAs from a single laser array pump source of this invention. This can provide advantages such as significant simplification in construction of optical fiber systems where multiple EDFAs are required at a given physical location, and compact design for signal amplification without WDM, where desired.[0147]
• A laser[0148]array pump source2310 is optically connected by itsoutput fibers2320 toEDFAs2390. While connection by three output fibers to three EDFAs is illustrated for purposes of clarity in the figure, it is apparent that such connection may be made to a greater or lesser number of EDFAs as desired for different applications, for example to 8 EDFAs, by 8 output fibers, as limited by the number of outputs ofpump source2310. EachEDFA2390 is shown to have asignal input fiber2330, an amplifiedsignal output fiber2340, aforward pumping port2370 and abackward pumping port2380.Pump output fibers2320 are shown to be connected to theforward pumping ports2370, but alternatively, the connection may be made to backward pumping ports2380 (as indicated by dashed lines), instead. Various EDFAs are known in the art and are available from commercial suppliers, such as the PureGain™ 2500C optical amplifier available from Corning Incorporated (Corning, N.Y.).
• Other alternate, known pump connection schemes may be employed. For example, simultaneous forward and backward pumping may be implemented according to this embodiment by connecting two different output fibers of the laser array pump source to each EDFA, one to the forward and one to the backward pumping port. A laser array pump source of this invention can provide such pumping to a number of EDFAs that is one half the number of its outputs.[0149]
• As used herein, a given event is “responsive” to a predecessor event if the predecessor event influenced the given event. If there is an intervening processing element, step or time period, the given event can still be “responsive” to the predecessor event. If the intervening processing element or step combines more than one event, the signal output of the processing element or step is considered “responsive” to each of the event inputs. If the given event is the same as the predecessor event, this is merely a degenerate case in which the given event is still considered to be “responsive” to the predecessor event. “Dependency” of a given event upon another event is defined similarly.[0150]
• It will be recognized by those skilled in the art that, while the invention has been described above in terms of preferred embodiments, it is not limited thereto. For example, any and all variations described, suggested or incorporated by reference in the Background section of this patent application are specifically incorporated by reference into the description herein of embodiments of the invention. In addition, various features and aspects of the above described invention may be used individually or jointly. Further, although the invention has been described in the context of its implementation in a particular environment and for particular applications, e.g., telecommunications, those skilled in the art will recognize that its usefulness is not limited thereto and that the present invention can be beneficially utilized in any number of environments and implementations. Accordingly, the claims set forth below should be construed in view of the full breadth and spirit of the invention as disclosed herein.[0151]

Claims (146)

1. A method for making an optical assembly, comprising the steps of:

providing a first submount having a standoff structure protruding from a first surface thereof;

pressing an optical emitter chip against the standoff structure such that at least said emitter chip deforms and said emitter chip contacts said standoff structure in a first plurality of contact portions of said standoff structure, said emitter chip having at least a first emitter;

bonding said emitter chip to said first submount;

juxtaposing an integrated optics chip against said standoff structure such that a first optical receiver of said integrated optics chip can receive optical energy emitted by said first emitter; and

bonding said integrated optics chip to said first submount.

2. A method according toclaim 1, wherein said first emitter is disposed on a subject edge of said emitter chip,

and wherein said standoff structure comprises a plurality of at least three segments which are mutually isolated from each other at least under said subject edge of said emitter chip, and which are arranged such that each segment includes a respective first portion which contacts said emitter chip and a respective second portion which contacts said integrated optics chip.

3. A method according toclaim 1, wherein said step of bonding said emitter chip to said first submount comprises the steps of:

applying an epoxy between two of said contact portions on said first submount; and

after said step of pressing, curing said epoxy.

4. A method according toclaim 1, wherein said step of bonding said emitter chip to said first submount comprises the steps of:

flowing solder between two of said contact portions on said first submount, and

after said step of pressing, cooling said solder.

5. A method according toclaim 1, further comprising the step of forming solder bumps between said contact portions on said first submount,

wherein said step of bonding said emitter chip to said first submount is performed as part of said step of pressing said emitter chip against said standoff structure.

6. A method according toclaim 1, wherein said step of bonding said emitter chip to said first submount comprises the step of applying an electrically conductive bonding material to an electrical trace disposed on said first submount between said contact portions said bonding material being in electrical contact with an electrical connection pad on said emitter chip after said step of bonding said emitter chip to said first submount.

7. A method according toclaim 1, wherein said step of bonding said integrated optics chip to said first submount comprises the step of applying an electrically conductive bonding material to an electrical trace disposed on said first submount between said contact portions, said bonding material being in electrical contact with an electrical connection pad on said integrated optics chip after said step of bonding said integrated optics chip to said first submount.

8. A method according toclaim 1, wherein said first plurality of contact portions includes all points on said standoff structure which contact said emitter chip after said step of pressing,

wherein said emitter chip includes a plurality of optical emitters arranged along a first edge of said emitter chip,

and wherein at least three consecutive ones of said contact portions along said first edge of said emitter chip are mutually isolated from each other along said first edge.

9. A method according toclaim 1, wherein said emitter chip is an optical emitter array chip having a plurality of optical emitters including said first emitter.

10. A method according toclaim 1, further comprising the step of pressing said integrated optics chip against said standoff structure after said step of juxtaposing, such that at least said integrated optics chip deforms and said integrated optics chip contacts said standoff structure in a second plurality of contact portions of said standoff structure.

11. A method according toclaim 10, wherein said integrated optics chip comprises a plurality of optical ports, and wherein said step of attaching further comprises the steps of:

activating said first optical emitter to emit optical energy and monitoring optical energy captured by said integrated optics chip; and

repositioning said integrated optics chip laterally relative to said first submount in response to said step of monitoring.

12. A method according toclaim 11, wherein said step of repositioning includes the step of repositioning said integrated optics chip longitudinally relative to said first submount in response to said step of monitoring.

13. A method according toclaim 11, wherein said steps of monitoring and repositioning are performed prior to said steps of pressing said integrated optics chip against said standoff structure and bonding said integrated optics chip to said first submount.

14. A method according toclaim 10, wherein said emitter chip has a plurality of emitters including located on a first edge of said emitter chip,

and wherein said integrated optics chip has a second plurality of optical receivers located on a second edge of said integrated optics chip,

further comprising the step of forming additional material on at least a portion of a first major surface of said emitter chip prior to said step of pressing an optical emitter chip against said standoff structure, such that the optical emitters in said plurality of emitters are located on said first edge at the same distance away from said first major surface as the optical receivers are located away from a major surface of said integrated optics chip.

15. A method according toclaim 10, wherein said emitter chip has a plurality of emitters including located on a first edge of said emitter chip,

and wherein said integrated optics chip has a second plurality of optical receivers located on a second edge of said integrated optics chip,

further comprising the step of forming additional material on at least a portion of a first major surface of said integrated optics chip prior to said step of juxtaposing an integrated optics chip against said standoff structure, such that the optical emitters in said plurality of emitters are located on said first edge at the same distance away from said first major surface as the optical receivers are located away from a major surface of said integrated optics chip.

16. A method according toclaim 1, wherein said emitter chip includes first and second opposite major surfaces, and wherein said step of pressing comprises the steps of:

affixing said first surface of said emitter chip to a compliant surface of a chuck;

moving said chuck such that said second surface of said optical component contacts said standoff structure; and

pressing said chuck toward said first submount until said emitter chip deforms to contact said standoff structure in said first plurality of contact portions.

17. A method according toclaim 1, further comprising the step of providing a first optical fiber disposed such that a receiving end of said first optical fiber can receive optical energy output from a first optical output of said integrated optics chip.

18. A method according toclaim 17, further comprising the step of connecting an output end of said first optical fiber to provide pump optical energy to an optical amplifier.

19. A method according toclaim 1, wherein said integrated optics chip comprises a mode converter in an optical path from said first optical receiver to a first optical output of said integrated optics chip.

20. A method according toclaim 19, wherein said mode converter comprises a waveguide taper.

21. A method according toclaim 1, wherein said integrated optics chip comprises:

a first waveguide segment in a first optical path from said first optical receiver to a first optical output of said integrated optics chip; and

a first wavelength stabilizer disposed in said first waveguide segment.

22. A method according toclaim 21, wherein said wavelength stabilizer comprises a first feedback grating superimposed on said first waveguide segment.

23. A method according toclaim 1, wherein said emitter chip further has a second optical emitter,

wherein said step of juxtaposing includes the step of juxtaposing said integrated optics chip against said standoff structure such that a second optical receiver of said integrated optics chip can receive optical energy emitted by said second emitter,

and wherein said integrated optics chip further comprises a directional coupler having a first input in a first optical path downstream of said first optical receiver, a second input in a second optical path downstream of said second optical receiver, and an output in an optical path upstream of said first optical output.

24. A method according toclaim 1, wherein said integrated optics chip comprises a polarization rotator disposed in a first optical path from said first optical receiver to a first optical output of said integrated optics chip.

25. A method according toclaim 1, wherein said integrated optics chip comprises a power monitor disposed to monitor optical power received by said first optical receiver.

26. A method according toclaim 1, wherein said integrated optics chip comprises a controllable optical switch.

27. A method according toclaim 1, wherein said integrated optics chip has a plurality of optical outputs, further comprising the step of attaching an optical fiber array having a plurality of optical fibers, such that receiving ends of each of said fibers can receive optical energy output from a respective one of the optical outputs of said integrated optics chip.

28. A method according toclaim 27, wherein said step of attaching an optical fiber array comprises the step of attaching said optical fiber array to said first submount.

29. A method according toclaim 28, further comprising the step of forming a plurality of longitudinally-oriented v-grooves in said first submount downstream of said integrated optics chip,

and wherein said step of attaching said optical fiber array to said first submount comprises the step of affixing each of said fibers in a respective one of said v-grooves.

30. A method according toclaim 29, wherein said step of affixing occurs after said steps of bonding said emitter chip to said first submount and bonding said integrated optics chip to said first submount.

31. A method according toclaim 28, wherein said first submount includes longitudinally-oriented recesses, and wherein said step of attaching said optical fiber array to said first submount comprises the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

32. A method according toclaim 27, wherein said step of attaching an optical fiber array comprises the steps of:

providing a second submount having a second standoff structure protruding from a first surface thereof;

juxtaposing an overhang portion of said integrated optics chip overhanging said first submount against said second standoff structure such that said integrated optics chip contacts said second standoff structure in a third plurality of contact portions of said second standoff structure;

bonding said overhang portion of said integrated optics chip to said second submount; and

attaching said optical fiber array to said second submount.

33. A method according toclaim 32, wherein said second submount is spaced longitudinally from said first submount.

34. A method according toclaim 32, wherein said second submount includes longitudinal recesses, and wherein said step of attaching said optical fiber array to said second submount comprises the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said second submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said second submount.

35. A method according toclaim 27, further comprising the step of connecting an output end of one of said optical fibers to provide pump optical energy to an optical amplifier.

36. A method according toclaim 27, further comprising the step of connecting the output end of each of said optical fibers to provide pump optical energy to a respective optical amplifier.

37. A method according toclaim 1, wherein said integrated optics chip has first and second optical outputs, both having a common plane of optical polarization, further comprising the steps of:

providing a polarization coupler having first and second polarization maintaining input fibers and an output fiber;

attaching said first input fiber so as to receive optical energy from said first optical output, said first input fiber being attached with a first rotation relative to said output fiber; and attaching said second input fiber so as to receive optical energy from said second optical output, said second input fiber being attached with a second rotation 90° different from said first input fiber.

38. A product produced using the method of any of claims1,9,10,27 and35.

39. Optical apparatus comprising:

a first submount having a standoff structure protruding from a first surface thereof;

an optical emitter chip bonded to said first submount, said optical emitter chip having a first optical emitter and contacting said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said emitter chip, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line; and

an integrated optics chip having a first optical receiver, said integrated optics chip being bonded to said first submount in such a way that said first optical receiver can receive optical energy emitted by said first emitter.

40. Apparatus according toclaim 39, wherein said integrated optics chip contacts said standoff structure in a second plurality of contact portions of said standoff structure, said second plurality of contact portions including all points on said standoff structure which contact said integrated optics chip, at least three consecutive ones of said contact portions in said second plurality of contact portions along a given straight line being mutually isolated from each other along said given straight line.

41. Apparatus according toclaim 39, further comprising an epoxy bonding said emitter chip to said first submount, said epoxy being located between two of said contact portions on said first submount and not on any of said contact portions.

42. Apparatus according toclaim 39, further comprising solder bonding said emitter chip to said first submount, said solder being located between two of said contact portions on said first submount and not on any of said contact portions.

43. Apparatus according toclaim 39, further comprising:

an electrical trace disposed on said first submount between said contact portions;

and an electrical connection pad on said emitter chip; and

an electrically conductive material bonding said emitter chip to said first submount and making electrical contact with both said electrical trace on said first submount and said electrical connection pad on said emitter chip.

44. Apparatus according toclaim 39, further comprising:

an electrical trace disposed on said first submount between said contact portions;

and an electrical connection pad on said integrated optics chip; and

an electrically conductive material bonding said integrated optics chip to said first submount and making electrical contact with both said electrical trace on said first submount and said electrical connection pad on said integrated optics chip.

45. Apparatus according toclaim 44, wherein said electrically conductive material is a solder.

46. Apparatus according toclaim 39, wherein said emitter chip includes a plurality of optical emitters arranged along a first edge of said emitter chip, and wherein at least three consecutive ones of said contact portions along a straight line parallel to said first edge of said emitter chip are mutually isolated from each other along said straight line parallel to said first edge.

47. Apparatus according toclaim 39, wherein said standoff structure comprises a plurality of ribs arranged such that each rib includes a respective first segment which is in said first plurality of contact portions and a respective second segment which extends under said integrated optics chip.

48. Apparatus according toclaim 39, wherein said integrated optics chip includes a first optical output, further comprising a first optical fiber having a receiving end disposed such that it can receive optical energy output from said first optical output of said integrated optics chip.

49. Apparatus according toclaim 48, further comprising an optical amplifier having an optical pump input connected to receive optical energy from said first optical fiber.

50. Apparatus according to claim48, wherein said integrated optics chip comprises a mode converter in an optical path from said first optical receiver to said first optical output.

51. Apparatus according toclaim 48, wherein said integrated optics chip comprises:

an input waveguide segment receiving optical energy from said first optical receiver and supporting an input optical mode;

an output waveguide segment downstream of said input waveguide segment and providing optical energy to said first optical output and supporting an output optical mode; and

a mode converter disposed in an optical path from said input optical waveguide segment to said output waveguide segment and converting said input optical mode to said output optical mode.

52. Apparatus according toclaim 51, wherein said input optical mode optimizes acceptance of optical energy from said first emitter into said input waveguide segment.

53. Apparatus according toclaim 51, wherein said output optical mode optimizes delivery of optical energy from said first output of said integrated optics chip into said first optical fiber.

54. Apparatus according toclaim 51, wherein said input optical mode optimizes acceptance of optical energy from said first emitter into said input waveguide segment, and wherein said output optical mode matches an optical mode supported by said first optical fiber.

55. Apparatus according toclaim 51, wherein said mode converter comprises a waveguide taper.

56. Apparatus according toclaim 39, wherein said integrated optics chip comprises:

a first waveguide segment in a first optical path from said first optical receiver to a first optical output of said integrated optics chip;

a first feedback grating superimposed on said first waveguide segment; and

a first optical monitor disposed to receive optical energy scattered out of said waveguide by said feedback grating.

57. Apparatus according toclaim 39, wherein said integrated optics chip comprises:

a first waveguide segment in a first optical path from said first optical receiver to a first optical output; and

a first wavelength stabilizer disposed in said first waveguide segment.

58. Apparatus according toclaim 57, wherein said wavelength stabilizer comprises a first feedback grating superimposed on said first waveguide segment.

59. Apparatus according toclaim 57, wherein said integrated optics chip further comprises a first waveguide taper in said first path upstream of said first waveguide segment.

60. Apparatus according toclaim 57, wherein said integrated optics chip further comprises a second waveguide taper in said first path downstream of said first waveguide segment.

61. Apparatus according toclaim 57, wherein said emitter chip further has a second optical emitter and said integrated optics chip has a second optical receiver, said integrated optics chip being disposed further so that said second optical receiver can receive optical energy emitted by said second emitter,

and wherein said integrated optics chip further comprises a second wavelength stabilizer in a second waveguide segment in a second optical path downstream of said second optical receiver.

62. Apparatus according toclaim 61, wherein said first and second wavelength stabilizers are tuned to different wavelengths,

and wherein said integrated optics chip further comprises a directional coupler in said first path downstream of said first stabilizer and coupling optical energy from said second path downstream of said second stabilizer into said first path toward said first optical output.

63. Apparatus according toclaim 61, wherein said first and second wavelength stabilizers are tuned to different wavelengths,

wherein said second optical path in said integrated optics chip terminates with a second optical output of said integrated optics chip,

further comprising a fused fiber coupler having first and second receiving optical fibers, said first receiving fiber being disposed such that it can receive optical energy output from said first optical output of said integrated optics chip, and said second receiving fiber being disposed such that it can receive optical energy output from said second optical output of said integrated optics chip.

64. Apparatus according toclaim 39, wherein said emitter chip further has a second optical emitter and said integrated optics chip has a second optical receiver, said integrated optics chip being disposed further so that said second optical receiver can receive optical energy emitted by said second emitter,

and wherein said integrated optics chip further comprises a directional coupler having a first input in a first optical path downstream of said first optical receiver, a second input in a second optical path downstream of said second optical receiver, and an output in an optical path upstream of said first optical output.

65. Apparatus according toclaim 39, wherein said integrated optics chip comprises a polarization rotator disposed in a first optical path from said first optical receiver to a first optical output of said integrated optics chip.

66. Apparatus according toclaim 65, wherein said emitter chip further has a second optical emitter and said integrated optics chip has a second optical receiver, said integrated optics chip being disposed further so that said second optical receiver can receive optical energy emitted by said second emitter,

and wherein said integrated optics chip further comprises a directional coupler in said first path downstream of said polarization rotator and coupling optical energy into said first path toward said first optical output from said second optical receiver.

67. Apparatus according toclaim 39, wherein said integrated optics chip comprises a power monitor disposed to monitor optical power received by said first optical receiver.

68. Apparatus according toclaim 39, wherein said integrated optics chip comprises a controllable optical switch.

69. Apparatus according toclaim 39, wherein said emitter chip further has a second and third optical emitters and said integrated optics chip has second and third optical receivers, said integrated optics chip being disposed further such that said second optical receiver of said integrated optics chip can receive optical energy emitted by said second emitter and such that said third optical receiver of said integrated optics chip can receive optical energy emitted by said third emitter,

and wherein said integrated optics chip further comprises:

a first waveguide in a first path from said first optical receiver to a first optical output of said integrated optical chip;

a second waveguide in a second path from said second optical receiver to a second optical output of said integrated optical chip; and

a cross-connect switching structure which switches optical energy from said third optical receiver selectably into (a) said first path toward said first optical output, (b) or said second path toward said second optical output, or (c) neither.

70. Apparatus according toclaim 39, wherein said integrated optics chip has a plurality of optical outputs, further comprising an optical fiber array having a plurality of optical fibers, each of said fibers having a respective receiving end disposed to receive optical energy output from a respective one of the optical outputs of said integrated optics chip.

71. Apparatus according toclaim 70, wherein said optical fiber array is attached to said first submount.

72. Apparatus according toclaim 71, wherein each of said optical fibers is affixed in a respective longitudinally-oriented v-groove in said first submount downstream of said integrated optics chip.

73. Apparatus according toclaim 71, wherein said first submount includes longitudinally-oriented recesses, further comprising:

a fiber holder having a plurality of longitudinally-oriented v-groove in the undersurface thereof, each of said fibers being affixed in a respective one of said v-grooves, and said fiber holder being attached to said first submount with the undersurface of said fiber holder facing said first surface of said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

74. Apparatus according toclaim 70, wherein said integrated optics chip has an overhang portion overhanging said first submount, further comprising:

a second submount having a second standoff structure protruding from a first surface thereof, said overhang portion of said integrated optics chip being attached to said second standoff structure such that said integrated optics chip contacts said second standoff structure in a third plurality of contact portions of said second standoff structure,

said optical fiber array being attached to said second submount.

75. Apparatus according toclaim 74, wherein said second submount is spaced longitudinally from said first submount.

76. Apparatus according toclaim 74, wherein said second submount includes longitudinally-oriented recesses, further comprising:

a fiber holder having a plurality of longitudinally-oriented v-groove in the undersurface thereof, each of said fibers being affixed in a respective one of said v-grooves, and said fiber holder being attached to said second submount with the undersurface of said fiber holder facing said first surface of said second submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said second submount.

77. Apparatus according toclaim 70, further comprising an optical amplifier having an optical pump input connected to receive optical energy from one of said optical fibers.

78. Apparatus according toclaim 70, further comprising a plurality of optical amplifiers each having an optical pump input connected to receive optical energy from a respective one of said optical fibers.

79. Apparatus according toclaim 39, wherein said integrated optics chip has first and second optical outputs, both having a common plane of optical polarization, further comprising:

a polarization coupler having first and second polarization maintaining input fibers and an output fiber, said first input fiber being attached so as to receive optical energy from said first optical output, said first input fiber being attached with a first rotation relative to said output fiber, and said second input fiber being attached so as to receive optical energy from said second optical output, said second input fiber being attached with a second rotation 90° different from said first input fiber.

80. Optical apparatus comprising:

a submount having a standoff structure protruding from a first surface thereof;

an integrated optics chip bonded to said submount, said integrated optics chip having a first optical receiver and contacting said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said integrated optics chip, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line; and

an optical emitter chip having a first optical emitter, said optical emitter chip being bonded to said submount in such a way that said first optical receiver can receive optical energy emitted by said first emitter.

81. A mounting method for optical components, comprising the steps of:

providing a submount having a standoff structure protruding from a first surface thereof;

juxtaposing an optical emitter chip against the standoff structure such that said emitter chip contacts said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said emitter chip after said step of juxtaposing, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line;

bonding said emitter chip to said submount with a bonding agent which contacts said submount only in regions thereof other than on said first plurality of contact portions; and

juxtaposing an integrated optics chip against said standoff structure such that a first optical receiver of said integrated optics chip can receive optical energy emitted by said first emitter.

82. A method according toclaim 81, wherein said step of bonding said emitter chip to said submount comprises the steps of:

applying an epoxy between two of said contact portions on said submount; and

after said step of juxtaposing, curing said epoxy.

83. A method according toclaim 81, wherein said step of bonding said emitter chip to said submount comprises the steps of:

flowing solder between two of said contact portions on said submount; and

after said step of juxtaposing, cooling said solder.

84. A method according toclaim 81, further comprising the step of forming solder bumps between said contact portions on said submount,

wherein said step of bonding said emitter chip to said submount is performed as part of said step of juxtaposing said emitter chip against said standoff structure.

85. A method according toclaim 81, wherein said step of bonding said emitter chip to said first submount comprises the step of applying an electrically conductive bonding agent to an electrical trace disposed on said first submount between said contact portions, said bonding agent being in electrical contact with an electrical connection pad on said emitter chip after said step of bonding said emitter chip to said first submount.

86. A method according toclaim 81, wherein said step of bonding said integrated optics chip to said first submount comprises the step of applying an electrically conductive bonding agent to an electrical trace disposed on said first submount between said contact portions, said bonding agent being in electrical contact with an electrical connection pad on said integrated optics chip after said step of bonding said integrated optics chip to said first submount.

87. A method according toclaim 81, wherein said emitter chip includes a plurality of optical ports arranged along a first edge of said emitter chip,

and wherein said straight line is a straight line parallel to said first edge of said emitter chip.

88. A method according toclaim 87, wherein first and second consecutive mutually isolated contact portions along said straight line are spaced from each other by a first inter-standoff spacing,

and wherein said straight line is a straight line that is closer to said first edge than said first inter-standoff spacing.

89. A method according toclaim 81, wherein said step of juxtaposing an integrated optics chip against said standoff structure comprises the step of juxtaposing said integrated optics chip against the standoff structure such that said integrated optics chip contacts said standoff structure in a second plurality of contact portions of said standoff structure, said second plurality of contact portions including all points on said standoff structure which contact said integrated optics chip after said step of juxtaposing said integrated optics chip against the standoff structure, at least three consecutive ones of said contact portions in said second plurality of contact portions along a given straight line being mutually isolated from each other along said given straight line

90. A method according toclaim 89, further comprising the step of:

bonding said second optical component to said submount with a bonding agent which contacts said submount only in regions thereof other than on the contact portions in said second plurality of contact portions.

91. A method according toclaim 90, wherein said standoff structure comprises a plurality of ribs arranged such that each rib includes a respective first segment which is in said first plurality of contact portions and a respective second segment which is in said second plurality of contact portions.

92. A method according toclaim 81, further comprising the step of providing a first optical fiber disposed such that a receiving end of said first optical fiber can receive optical energy output from a first optical output of said integrated optics chip.

93. A method according toclaim 92, further comprising the step of connecting an output end of said first optical fiber to provide pump optical energy to an optical amplifier.

94. A method according toclaim 81, wherein said integrated optics chip comprises a mode converter in an optical path from said first optical receiver to a first optical output of said integrated optics chip.

95. A method according toclaim 94, wherein said mode converter comprises a waveguide taper.

96. A method according toclaim 81, wherein said integrated optics chip comprises:

a first waveguide segment in a first optical path from said first optical receiver to a first optical output of said integrated optics chip, and

a first wavelength stabilizer disposed in said first waveguide segment.

97. A method according toclaim 96, wherein said wavelength stabilizer comprises a first feedback grating superimposed on said first waveguide segment.

98. A method according toclaim 81, wherein said emitter chip further has a second optical emitter,

wherein said step of juxtaposing includes the step of juxtaposing said integrated optics chip against said standoff structure such that a second optical receiver of said integrated optics chip can receive optical energy emitted by said second emitter,

and wherein said integrated optics chip further comprises a directional coupler having a first input in a first optical path downstream of said first optical receiver, a second input in a second optical path downstream of said second optical receiver, and an output in an optical path upstream of said first optical output.

99. A method according toclaim 81, wherein said integrated optics chip comprises a polarization rotator disposed in a first optical path from said first optical receiver to a first optical output of said integrated optics chip.

100. A method according toclaim 81, wherein said integrated optics chip comprises a power monitor disposed to monitor optical power received by said first optical receiver.

101. A method according toclaim 81, wherein said integrated optics chip comprises a controllable optical switch.

102. A method according toclaim 81, wherein said integrated optics chip has a plurality of optical outputs, further comprising the step of attaching an optical fiber array having a plurality of optical fibers, such that receiving ends of each of said fibers can receive optical energy output from a respective one of the optical outputs of said integrated optics chip.

103. A method according toclaim 102, wherein said step of attaching an optical fiber array comprises the step of attaching said optical fiber array to said first submount.

104. A method according toclaim 103, further comprising the step of forming a plurality of longitudinally-oriented v-grooves in said first submount downstream of said integrated optics chip,

and wherein said step of attaching said optical fiber array to said first submount comprises the step of affixing each of said fibers in a respective one of said v-grooves.

105. A method according toclaim 104, wherein said step of affixing occurs after said steps of bonding said emitter chip to said first submount and bonding said integrated optics chip to said first submount.

106. A method according toclaim 103, wherein said first submount includes longitudinally-oriented recesses, and wherein said step of attaching said optical fiber array to said first submount comprises the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

107. A method according toclaim 102, wherein said step of attaching an optical fiber array comprises the steps of:

providing a second submount having a second standoff structure protruding from a first surface thereof;

juxtaposing an overhang portion of said integrated optics chip overhanging said first submount against said second standoff structure such that said integrated optics chip contacts said second standoff structure in a third plurality of contact portions of said second standoff structure;

bonding said overhang portion of said integrated optics chip to said second submount; and

attaching said optical fiber array to said second submount.

108. A method according toclaim 107, wherein said second submount is spaced longitudinally from said first submount.

109. A method according toclaim 107, wherein said second submount includes longitudinal recesses, and wherein said step of attaching said optical fiber array to said second submount comprises the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said second submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said second submount.

110. A method according toclaim 102, further comprising the step of connecting an output end of one of said optical fibers to provide pump optical energy to an optical amplifier.

111. A method according toclaim 102, further comprising the step of connecting the output end of each of said optical fibers to provide pump optical energy to a respective optical amplifier.

112. A method according toclaim 81, wherein said integrated optics chip has first and second optical outputs, both having a common plane of optical polarization, further comprising the steps of:

providing a polarization coupler having first and second polarization maintaining input fibers and an output fiber;

attaching said first input fiber so as to receive optical energy from said first optical output, said first input fiber being attached with a first rotation relative to said output fiber; and

attaching said second input fiber so as to receive optical energy from said second optical output, said second input fiber being attached with a second rotation different from said first input fiber.

113. A product produced using the method of any of claims81,83,90,102 and111.

114. A mounting method for optical components, comprising the steps of:

providing a submount having a standoff structure protruding from a first surface thereof;

juxtaposing an integrated optics chip against the standoff structure such that said integrated optics chip contacts said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said integrated optics chip after said step of juxtaposing, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line;

bonding said integrated optics chip to said submount with a bonding agent which contacts submount only in regions thereof other than on said first plurality of contact portions; and

juxtaposing an optical emitter chip against said standoff structure such that a first optical receiver of said integrated optics chip can receive optical energy emitted by said first emitter.

115. A method for making an optical assembly, comprising the steps of:

providing a first submount having a standoff structure protruding from a first surface thereof;

pressing an optical array emitter chip against the standoff structure such that at least said emitter chip deforms and said emitter chip contacts said standoff structure in a first plurality of contact portions of said standoff structure, said emitter chip having a plurality of optical emitters;

bonding said emitter chip to said first submount;

juxtaposing against said standoff structure an optical fiber array having a plurality of optical fibers, such that a receiving end of each of said fibers can receive optical energy emitted by a respective one of said optical emitters, and

bonding said integrated optics chip to said first submount.

116. A method according toclaim 115, wherein said optical emitters are disposed on a subject edge of said emitter chip,

and wherein said standoff structure comprises a plurality of at least three segments which are mutually isolated from each other at least under said subject edge of said emitter chip, and which are arranged such that each segment includes a respective first portion which contacts said emitter chip and a respective second portion which contacts said optical fiber array.

117. A method according toclaim 115, wherein said step of bonding said emitter chip to said first submount comprises the step of applying an electrically conductive bonding agent to an electrical trace disposed on said first submount between said contact portions, said bonding agent being in electrical contact with an electrical connection pad on said emitter chip after said step of bonding said emitter chip to said first submount.

118. A method according toclaim 115, wherein said first plurality of contact portions includes all points on said standoff structure which contact said emitter chip after said step of pressing,

wherein said plurality of optical emitters are arranged along a first edge of said emitter chip,

and wherein at least three consecutive ones of said contact portions along said first edge of said emitter chip are mutually isolated from each other along said first edge.

119. A method according toclaim 115, further comprising the step of pressing said optical fiber array against said standoff structure after said step of juxtaposing, such that at least said optical fiber array deforms and said optical fiber array contacts said standoff structure in a second plurality of contact portions of said standoff structure.

120. A method according toclaim 115, wherein said step of juxtaposing includes the steps of:

activating a first one of said optical emitters to emit optical energy and monitoring optical energy captured by one of said optical fibers; and

repositioning said optical fiber array laterally relative to said first submount in response to said step of monitoring.

121. A method according toclaim 120, wherein said step of juxtaposing further includes the step of activating a second one of said optical emitters to emit optical energy and monitoring optical energy captured by a second one of said optical fibers,

and wherein said step of repositioning is performed further in response to said step of monitoring optical energy captured by a second one of said optical fibers.

122. A method according toclaim 115, wherein said emitter chip includes first and second opposite major surfaces, and wherein said step of pressing comprises the steps of:

affixing said first surface of said emitter chip to a compliant surface of a chuck;

moving said chuck such that said second surface of said optical component contacts said standoff structure; and

pressing said chuck toward said first submount until said emitter chip deforms to contact said standoff structure in said first plurality of contact portions.

123. A method according toclaim 115, wherein said first submount includes longitudinally-oriented recesses, and wherein said steps of juxtaposing and bonding said integrated optics chip to said first submount collectively comprise the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

124. A method according toclaim 115, further comprising the step of connecting an output end of one of said optical fibers to provide pump optical energy to an optical amplifier.

125. A method according toclaim 115, further comprising the step of connecting the output end of each of said optical fibers to provide pump optical energy to a respective optical amplifier.

126. A product produced using the method of any of claims115,124 and125.

127. Optical apparatus comprising:

a first submount having a standoff structure protruding from a first surface thereof;

an optical array emitter chip bonded to said first submount, said optical emitter chip having a plurality of optical emitters and contacting said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said emitter chip, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line; and

an optical fiber array having a plurality of optical fibers, said optical fiber array being bonded to said first submount in such a way that said a receiving end of each of said fibers can receive optical energy emitted by a respective one of said optical emitters.

128. Apparatus according toclaim 127, wherein said optical fiber array contacts said standoff structure in a second plurality of contact portions of said standoff structure, said second plurality of contact portions including all points on said standoff structure which contact said optical fiber array, at least three consecutive ones of said contact portions in said second plurality of contact portions along a given straight line being mutually isolated from each other along said given straight line.

129. Apparatus according toclaim 127, further comprising:

an electrical trace disposed on said first submount between said contact portions;

and an electrical connection pad on said emitter chip; and

an electrically conductive material bonding said emitter chip to said first submount and making electrical contact with both said electrical trace on said first submount and said electrical connection pad on said emitter chip.

130. Apparatus according toclaim 127, wherein said optical emitters are arranged along a first edge of said emitter chip,

and wherein at least three consecutive ones of said contact portions along a straight line parallel to said first edge of said emitter chip are mutually isolated from each other along said straight line parallel to said first edge.

131. Apparatus according toclaim 127, wherein said standoff structure comprises a plurality of ribs arranged such that each rib includes a respective first segment which is in said first plurality of contact portions and a respective second segment which extends under said optical fiber array.

132. Apparatus according toclaim 127, wherein said first submount includes longitudinally-oriented recesses,

wherein said optical fiber array comprises a fiber holder having a plurality of longitudinally-oriented v-groove in the undersurface thereof, each of said fibers being affixed in a respective one of said v-grooves, and

and wherein said fiber holder is attached to said first submount with the undersurface of said fiber holder facing said first surface of said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

133. Apparatus according toclaim 127, further comprising an optical amplifier having an optical pump input connected to receive optical energy from one of said optical fibers.

134. Apparatus according toclaim 127, further comprising a plurality of optical amplifiers each having an optical pump input connected to receive optical energy from a respective one of said optical fibers.

135. Optical apparatus comprising:

a submount having a standoff structure protruding from a first surface thereof;

an integrated optics chip bonded to said submount, said integrated optics chip having a first optical receiver and contacting said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said integrated optics chip, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line; and

an optical emitter chip having a first optical emitter, said optical emitter chip being bonded to said submount in such a way that said first optical receiver can receive optical energy emitted by said first emitter.

136. A mounting method for optical components, comprising the steps of:

providing a submount having a standoff structure protruding from a first surface thereof;

juxtaposing an optical array emitter chip against the standoff structure such that said emitter chip contacts said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said emitter chip after said step of juxtaposing, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line;

bonding said emitter chip to said submount with a bonding agent which contacts said submount only in regions thereof other than on said first plurality of contact portions; and

juxtaposing against said standoff structure an optical fiber array having a plurality of optical fibers, said optical fiber array being arranged such that receiving ends of each of said fibers can receive optical energy emitted by a respective optical emitter on said optical array emitter chip.

137. A method according toclaim 136, wherein said step of bonding said emitter chip to said first submount comprises the step of applying an electrically conductive material to an electrical trace disposed on said first submount between said contact portions, said material being in electrical contact with an electrical connection pad on said emitter chip after said step of bonding said emitter chip to said first submount.

138. A method according toclaim 136, wherein said optical emitters on said emitter chip are arranged along a first edge of said emitter chip,

and wherein said straight line is a straight line parallel to said first edge of said emitter chip.

139. A method according toclaim 138, wherein first and second consecutive mutually isolated contact portions along said straight line are spaced from each other by a first inter-standoff spacing,

and wherein said straight line is a straight line that is closer to said first edge than said first inter-standoff spacing.

140. A method according toclaim 136, wherein said step of juxtaposing an optical fiber array against said standoff structure comprises the step of juxtaposing said optical fiber array against the standoff structure such that said optical fiber array contacts said standoff structure in a second plurality of contact portions of said standoff structure, said second plurality of contact portions including all points on said standoff structure which contact said optical fiber array after said step of juxtaposing said optical fiber array against the standoff structure, at least three consecutive ones of said contact portions in said second plurality of contact portions along a given straight line being mutually isolated from each other along said given straight line.

141. A method according toclaim 140, further comprising the step of:

bonding said optical fiber array to said submount with a bonding agent which contacts said submount only in regions thereof other than on the contact portions in said second plurality of contact portions.

142. A method according toclaim 141, wherein said standoff structure comprises a plurality of ribs arranged such that each rib includes a respective first segment which is in said first plurality of contact portions and a respective second segment which is in said second plurality of contact portions.

143. A method according toclaim 136, wherein said first submount includes longitudinally-oriented recesses, and wherein said step of juxtaposing an optical fiber array against said standoff structure comprises the steps of:

attaching each of said fibers in a longitudinally-oriented v-groove in the undersurface of a fiber holder; and

attaching said fiber holder to said first submount, said fibers depending below the undersurface of said fiber holder and into said recesses in said first submount.

144. A method according toclaim 136, further comprising the step of connecting the output end of each of said optical fibers to provide pump optical energy to a respective optical amplifier.

145. A product produced using the method of either of claims136 and144.

146. A mounting method for optical components, comprising the steps of:

providing a submount having a standoff structure protruding from a first surface thereof;

juxtaposing against the standoff structure an optical fiber array having a plurality of optical fibers, such that said optical fiber array contacts said standoff structure in a first plurality of contact portions of said standoff structure, said first plurality of contact portions including all points on said standoff structure which contact said optical fiber array after said step of juxtaposing, at least three consecutive ones of said contact portions along a straight line being mutually isolated from each other along said straight line;

bonding said optical fiber array to said submount with a bonding agent which contacts said submount only in regions thereof other than on said first plurality of contact portions, and

juxtaposing an optical array emitter chip against said standoff structure such that receiving ends of each of said fibers can receive optical energy emitted by a respective optical emitter on said optical array emitter chip.

Images