JPS6252501A - Manufacture of microlens - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention]

TECHNICAL FIELD OF THE INVENTION This invention relates to the manufacture of microlenses, and more particularly to a method for integrally manufacturing microlenses as fully integrated components such as optoelectronic devices. More particularly, the present invention applies current semiconductor manufacturing techniques to the fabrication of microlenses and microlens arrays, thereby facilitating the fabrication of integrated optical devices. More particularly, the present invention provides a method for manufacturing self-aligning microlens arrays that minimizes the time and expense associated with establishing and maintaining desired image and/or object plane alignment to the microlens array. Regarding. The microlens manufacturing method of the present invention was assigned to the applicant [Pedestal type microlens manufacturing method ('D-8
3068) U.S. application entitled J (filing date
This is a modification of the method described in , Application No. ). Accordingly, the method will be described in the context of and with reference to the same exemplary surroundings so as to clarify its distinguishing features. & ■ Integrated optical system
cs ) is extremely useful in the rapidly growing field of optoelectronics. Solid-state diode lasers use beam shaping optics.
tics), and solid-state photodetectors are commonly available. I use collection optics. These and other optoelectronic devices are small, short focal length lenses (referred to herein as "microlenses").
Substantial effort and expense has therefore been devoted to the development of microlenses and microlens arrays. Unfortunately, however, the technology for manufacturing such lenses is limited to LSI (Large Scale Integration) and VLSI, which are currently used in the manufacture of optoelectronic devices.
(Extremely large-scale integration) The development of semiconductor manufacturing technology was delayed. As a general rule, microlenses and microlens arrays are molded parts manufactured by replicating the contours of precisely machined dies. For example, I. N. Ozerov et al., “Internal shape of the contour of a die for manufacturing a lens array with spherical elements,” Soviet Journal of Optical Technology (USA), Vol.
See Vol. 8, No. 1, January 1981, pp. 49-50. In fact, semiconductor manufacturing technology is based on stepped index zone plate lenses.
It has been proposed for the production of molds for the production of replicas of latc'1ens). See El Bilia et al., "Photolithographic Production of Thin Film Lenses," Obtakes Communication, Vol. 5, No. 4, June 1972, pp. 232-5. Although the copying method is suitable for manufacturing high-quality, uniform microlenses and microlens arrays at low cost, the lenses must be attached and optically aligned after being removed from the mold. For this reason, great care must be taken to ensure that the lens is properly aligned and securely fixed or held in place when used in an optoelectronic device. Even then, the required optical alignment is difficult to maintain over long periods of time. This is because environmental factors such as large temperature fluctuations result in differences in the thermal expansion of the lens and the optoelectronic device to which it is attached. It was also reported that planar arrays of graded index microlenses have been fabricated. M. Oikawa et al., “Array of distributed index planar microlenses prepared by ion exchange technology,” Japanese Journal of Applied Physics, Vol. 20, No. 4, April 1981, no. See pages 296-8. Additionally, criteria and methods for evaluating microlenses and microlens arrays have been proposed. G.B. Shuttlesky et al.'s ``effective depth of field''

[Effects of Manufacturing Errors in Pull Lens Arrays] Volume 41, No. 11, 1974
“Evaluation of the integrated quality of microstructured lens arrays” by G. Et. Glukhosky et al., November 2013, pp. 507-9, Soviet Journal of Optical Technology, Vol. 40, No. 7. , July 1973, pp. 413-5. However, these foreign publications generally relate to independent (i.e., non-integrated) manufacturing and testing techniques for microlenses and microlens arrays, and the coupling and alignment encountered when conventional microlenses are used. It is not intended to substantively alleviate the problem. The structure and manufacturing method of the integrated microlens are described in Wai Ishihara et al. [High light sensitivity IL-CCD image sensor with integrated resin lens array], International Electron Device Meeting, 1.
983, pp. 497-500. They can achieve wide-area COD using roughly standard semiconductor manufacturing technology.
(Charge-Coupled Device) Reported in the development of a method for manufacturing strip lenses for image sensor arrays. According to their description of the method, (1) depositing a second layer of resin on top of the annealed, thermoset, smooth base resin layer;
(2) photolithographically patterning the second resin layer 11ifll to form a strip-like pattern; and (3) heating the second resin layer so that it flows, thereby forming the strip-like pattern. The pattern was changed in shape to form a series of semi-cylindrical rolls, or convex lenses. Therefore, this strip lens was integrated with the COD array, thereby avoiding the coupling and alignment problems of conventional microlenses. However, there is no control over the manufacturing process of such integrated lenses that is sufficient to produce reproducible results for manufacturing microlenses and microlens arrays to fairly precise optical specifications. There is another problem. In particular, one of the fundamental drawbacks in the aforementioned method is that the flow of the lens-forming resin by the second layer is sufficient to form microlenses with a defined geometry, i.e. the spacing The formation of geometrically uniform, dense arrays of individually addressable microlenses distributed over a large area has not been properly controlled. Instead, this method description suggests that the thermal flow of the resin forming the lenses is not substantially coupled by anything other than heat dissipation/dissipation properties in the equipment system, and that the lenses flow into each other and mix. This appears to make it extremely difficult, if not impossible, to predict optical coupling properties with reasonable accuracy. Although precise process control is not required in the manufacture of various types of microlenses, the optical specifications of microlenses and microlens arrays are usually based on the fact that the optical properties of each microlens are independently determined. The process is so rigorous that it requires proper process controls to ensure that. According to the present invention, microlenses and microlens arrays are fabricated on a substrate, such as a coated opto-electronic device, using a relatively cool metal layer in which an aperture is formed. Formed on such a substrate by free flowing molten lens material within a defined area, which is a confining area bounded by a non-wettable encapsulant such as be. This non-wettable containment prevents the lens material from spreading beyond the aperture edges. Furthermore, the area of each aperture is made small so that the internal pressure due to surface tension acting on the molten lens material is much greater than the pressure due to gravity, which allows the limited lens material to expand its volume. and if1, it is made to essentially form a semicircular shape with a contact radius as determined by the mouth shape. Advantageously, standard surface coating and photolithographic patterning techniques can be used to prepare the substrate for the manufacture of microlenses and microlens arrays, including the deposition of lens materials. Preferential radiant heating of the lens material allows for melting of the lens material while maintaining the encapsulant layer at a substantially lower temperature. DETAILED DESCRIPTION OF THE ILLUSTRATED EMBODIMENTS The present invention will hereinafter be described in detail with respect to one illustrated embodiment, but it must be understood that there is no intention to limit it to this embodiment. On the contrary, the intention is to cover all modifications, variations, and equivalents included within the spirit and scope of the invention as defined in the appended claims. Referring to the drawings, and with particular reference now to FIG. 1, a solid state optical memory 11 is shown to provide an example of an opto-electronic device, and a method of manufacturing an integrated microlens of the present invention is illustrated. This is particularly important in this regard. This memory 11 is related to this and to the common assignee 1! No. 678,145 [Dense Block Oriented Solid State Optical Memory J (Dense Block Oriented Solid State Optical Memory J (
D/82066) is described in considerable detail,
A simple explanation will suffice. "Blocks" of multi-bit data are selectively restored from memory 11 in response to instructions. For this purpose, the memory 11 includes an array of microlenses 12a to 121, and a photodetector 148 is inserted through the data mask 15.
~Photo emitter 13a~ in parallel on the array of 14i~
13j array, and the data mask has a data-dependent transfer profile. All of the photoemitters 13a to 13j form an image on each of the photodetectors 148 to 14i, and the photoemitter 1
The transmission profile of data mask 15 determines whether a given one of photodetectors 148-141 is illuminated or not illuminated when one and only one of photodetectors 3a-13j are selectively energized. It is done like this. Correct optical alignment of the memory 11 is such that (if all bits of the selected data block are optically addressed in parallel by one of the selectively activated photoemitters 13a-13j, fiil The bits of the optically addressed data block are then read out in parallel by photodetectors 148-141, respectively, thereby spatially detecting the high "1" and low "O" logic level digital electrical signals. A collimating lens 16 is advantageously included to collimate the light emitted by the photoemitters 13a-13j, thereby ensuring that the light emitted by the photoemitters 13a-13j is converted into a uniform sequence. The light received through all the apertures will be at substantially the same specific angle. Therefore, the images formed by the microlenses from photoemitter arrays 138-13j will be at substantially the same angle as those of photodetector arrays 148-14i. The data mask 15 not only has uniform width and depth magnification, but also essentially centered proper optical alignment on the photoeye arrays 148-14i, respectively. is an optically opaque film or metal that is suitably programmed to have a spatial distribution pattern of pores or apertures while data is placed thereon (using equipment not shown). ), and these apertures are formed at precisely predetermined locations.In this way, a logic level bit, e.g., ``1'', is provided by these apertures, while the opposite human logic The level "0" is provided by the portion left as opaque.The data mask 15 requires an intermediate layer (not shown) to form its planar substrate, but the photodetector 1
48-14i can be coated. Bits are densely packed on the data mask 15,
The memory 11 can be configured with a memory capacity of up to 16 megabits, or can be configured as a memory 11 with external dimensions of slightly more than 1 in. Accordingly, the microlenses 12a-121 have essentially the same focal length, providing virtually diffraction-limited imaging performance, and without significant lateral scattering of reflected light from the data mask 15. is important. Such requirements are such that the microlenses 12a to 12i are compatible with the data mask 15' and the photodetectors 148 to
The attendant requirement that microlenses 12a to 14i must be carefully optically aligned with each other is
2i to Data Mask 15/Photo Detector 148~
This is very easily achieved by an integrated manufacturing method that provides a complete integral coating on the 14i assembly. According to the present invention, microlenses and microlenses can be fabricated to exact specifications for optoelectronic devices such as data mask 15/photodetector 14a-14i assemblies of memory 11, or other suitable substrates such as quartz wafers. Array bodies 12a to 12
A method for integrally manufacturing a microlens array such as i is provided. Existing semiconductor manufacturing techniques are used in performing the lens manufacturing process, so that the risk of damaging the opto-electronic device while one or more microlenses are being formed is substantially reduced. Referring to FIGS. 2a-2b, in accordance with the present invention, the substrate for the microlenses 12a-12i (e.g., the data mask 15/photodetector 14a-14i assembly) is first made of polyimide 120 of a predetermined thickness. (partially shown) or by a similar optically transparent material of approximately the same refractive index and thickness, thereby establishing the desired spatial relationship between the focal planes of the microlenses 12a-12i and their substrate. be done. For example, the thickness of the coating 20 is selected such that the focal plane of the microlenses 12a-121 relative to the substrate coincides with the plane of the data mask 15 (FIG. 1). A thin optically opaque metal, such as an aluminum film 21, is then deposited over the coating 20, such as by vacuum evaporation, and this metal 21 is then photolithographically patterned as shown in Figure 2a. Then, apertures 21a to 211 (only a few are visible) with accurate dimensions are penetrated at predetermined positions. These openings 21a to 21i are microlenses 12a to 12, respectively.
Form a stop for i. In this way, for example, these are centered on the photodetectors 14a to 141 (
1), forming a circular profile approximately 15 microns in diameter. To perform the next step, a non-volatile (nonVo 18C) such as Shipley 1400-27 Photoresist is used as shown in Figure 2b.
) An optical resin-based positive photoresist 22 is coated from the apertured metal E, typically by spin coating at 4000 rpm for about 30 seconds to a coating thickness of approximately 1 micron. . Resin coating 2
In order to protect the material from undesirable thermal distortion, it is baked hard at this point and also deep U, V hardening is performed. Earl Allen et al., “Deep U, V, Curing of Patterns in Positive Photoresists,” Journal of
The Electrochemical Society, Volume 129, 1
See June 982, pp. 1379-81. Thereafter, a highly reflective coating, such as another vacuum-deposited aluminum film or similar metal, is deposited over coating 22 and photolithographically patterned to match each opening 21a-21i to form a precise pattern. Dimension opening 23
a to 23i are formed. Hemispherical lenses 12a to 121
For example, if the dimensions are given along with the manufacturing of the openings 23a to 2
It is noted that 3i has a diameter of approximately 30 microns, or in other words approximately twice the diameter of the lens apertures 21a-21i. Next, a positive photoresist based on a nonvolac optical resin (for example, Sibley (
A relatively thick coating of polymeric lens material, such as another coating of TF-20 photoresist (Shipley), is applied over the apertured metal 23, e.g.
The coating was deposited by spin coating for 40 seconds at 0 Orpm to form a coating approximately 15 microns thick. Because of the relatively thick layer of lens material desired, this coating is preferably deposited in two steps, with the first coating of lens material being softened at about 90° C. for about 10 minutes before forming the second coating. , which is then preferably allowed to soften for about 30 seconds. Following the coating of the lens material, it is photolithographically patterned as shown in FIG. As will be apparent, the centering of the lens material masses 24a-24i is not particularly important, as is the formation of flow paths by non-wetting expansion of the lens material. Suitable tolerances are provided between their circumferences and the edges of the openings 23a-231 to prevent this. Similarly, the cylindrical shape of the masses 24a-241 ensures that they are all made of the same amount of material. It will be appreciated that the lens material located within the apertures 23a-23i, as described below, may be of any suitable geometry to ensure that the lenses 128-12i, respectively.
, and therefore the equivalent material is the lens 12a~
12i are essentially the same. Now, in order to form the microlenses 12a to 12i,
The lens material cells 24a to 24i are marked @25 in Fig. 2d.
Ultraviolet rays of appropriate intensity (
1, R,). A carbon dioxide laser or a silicon carbide source is used as the source of this radiation. Is the melted lens material 24a-241 a resin liquid? Wet I22 and open the opening 23.
a to 23i to freely flow laterally. The metal 23 is also exposed to ultraviolet radiation, but has been empirically found to heat only slightly due to its reflectance, and the relationship between the metal 23 on the one hand and the molten lens material 24a-24i on the other hand is There is a considerable temperature difference between them. As a result, metal 23 becomes a containment body for molten lens material 248-24i. This is because the edges are relatively cold (i.e., have poor wettability or do not wet), which is why the openings 23a-23
This is because the lens material is prevented from flowing beyond the boundary i. Furthermore, in order to prevent any undesired flow of lens material beyond the boundaries of the apertures 23a-23:
The wettability of the metal 23 can be reduced, for example by etching or coating with a surfactant, to increase the interfacial energy gear between the gold 1ii 23 and the molten lens material 24a-241. Typically, any one of the microlenses 12a to 12
The volume of lens material provided to form a hemispherical microlens is limited to approximately no more than 2πr3/3, where r is the radius of the apertures 23a-231, which is suitable for forming hemispherical microlenses. It is a large amount. A small amount of material is used to form a partially hemispherical shaped microlens, and is useful in more generally determining the amount of lens material provided per microlens as a "balance amount". This means that the material is sufficiently low in acid to reach an equilibrium state when completely confined by the open edges. The inherent surface tension of the molten lens material 24a-24i is such that it imparts a substantially constant radius to the microlenses 12a-121, and the gravitational pull of the microlenses 12a-121 while they cool and resolidify. No significant deformations occur. As already explained, the opening 23
The non-wetting boundaries or edges of a to 231 are lens material 24a
~24j, thereby limiting the flow of lenses 12a~1
21 from mixing with each other. In this way, a non-hemispherical microlens (not shown) is inserted into the aperture 23.
By changing the shapes of a to 23i, they can be manufactured using the technology of the present invention. For example, (1) an elongated strip-like aperture is formed for the manufacture of cylindrical lenses, and (2) an elliptical pedestal is formed for the manufacture of elliptical lenses. Gravity tends to deform the microlenses 128-121 if it is large, but if the internal pressure of the molten photoresist is significantly greater than the pressure due to gravity, no significant gravitational deformation occurs. In other words, 2t
If /r>>gρr, gravity can be ignored, where t and ρ are the molten lens materials 24a to 24i, respectively.
where r is the radius of the apertures 23a-23i, and q is the gravitational force acting on the molten lens material. In view of the foregoing, the present invention provides a fully controlled method for fabricating microlenses, which integrates and fabricates microlenses and microlens arrays onto opto-electronic devices or other substrates. It will now be understood that this is what is used for. Furthermore, it will be apparent that the method of manufacturing microlenses according to the present invention can be accomplished using existing semiconductor manufacturing techniques.

[Brief explanation of drawings]

FIG. 1 is a step-by-step illustration of a method for manufacturing block-oriented solid-state optical memory cells with integrated microlens arrays manufactured in accordance with the present invention. FIG. 3 is a cutaway plan view of a microlens array manufactured according to the present invention. 11 Optical memory 12a-12
i Microlens array 13a to 13j Photo emitters 14a to 141 Photodetector 15 Data mask 16 Collimating lens 20
Polyimide layer 21.23 Aluminum film 21a-21
i, 23a-231 opening 22 photoresist 24a-241 lens mass

Claims (1)

[Claims] A method of manufacturing a microlens on a substrate, comprising: selecting a lens material; and coating the substrate with a layer of a material that is easily wetted by the lens material, thereby making the lens material wettable. forming a surface, depositing on the wettable surface an encapsulant that is not easily wettable by the lens material, and forming at least one predetermined shaped aperture through the encapsulant. to expose a preselected area of the wettable surface, and to create a predetermined equilibrium laterally on the wettable surface across the area exposed by the aperture. allowing a volume of molten lens material to flow freely, said volume being selected such that the flow of said lens material is restrained laterally circumferentially of said area by said containment; so that the molten lens material substantially conforms to the shape of the aperture and occupies a semicircular arc-shaped profile of a substantially constant radius determined by the volume thereof, and the lens-forming material A method for manufacturing a microlens, comprising the step of: curing the microlens to form a microlens having the shape and outline described above.