Methods of manufacturing an integrated circuit having stress tuning layer

Abstract

Warpage and breakage of integrated circuit substrates is reduced by compensating for the stress imposed on the substrate by thin films formed on a surface of the substrate. Particularly advantageous for substrates having a thickness substantially less than about 150 ?m, a stress-tuning layer is formed on a surface of the substrate to substantially offset or balance stress in the substrate which would otherwise cause the substrate to bend. The substrate includes a plurality of bonding pads on a first surface for electrical connection to other component.

Claims

1. A method of manufacturing a semiconductor device, the method comprising: forming a stress layer on a back side of a semiconductor wafer; and tuning the stress layer by forming a series of concentric annular regions.

2. The method of claim 1, wherein the stress layer comprises glass.

3. The method of claim 1, wherein the stress layer comprises ceramic.

4. The method of claim 1, wherein the stress layer comprises molding compound.

5. The method of claim 1, wherein the stress layer comprises fluorine-doped silicon oxide.

6. The method of claim 1, wherein the stress layer comprises silicon carbide.

7. The method of claim 1, further comprising electrically connecting external connections to wire bonds, the external connections being located on an opposite side of the semiconductor wafer from the stress layer.

8. A method of manufacturing a semiconductor device, the method comprising: depositing one or more layers on a first side of a semiconductor substrate; placing a protective film over the one or more layers on the first side of the semiconductor substrate; grinding the semiconductor substrate while the protective film is present; depositing a stress tuning layer on an opposite side of the semiconductor substrate from the protective film; determining a pattern for the stress tuning layer; patterning the pattern into the stress tuning layer; and removing the protective film.

9. The method of claim 8, wherein the determining the pattern is performed at least in part using empirical data.

10. The method of claim 8, wherein the removing the protective film exposes contact pads.

11. The method of claim 10, further comprising placing a solder ball in physical connection with one of the contact pads.

12. The method of claim 10, further comprising placing a wire in physical connection with one of the contact pads.

13. The method of claim 8, wherein the depositing the stress tuning layer is performed at a temperature of no greater than 400? C.

14. The method of claim 8, wherein the depositing the stress tuning layer deposits the stress tuning layer to a thickness of less than about 20 ?m.

15. A method of manufacturing a semiconductor device, the method comprising: depositing a plurality of layers over a semiconductor substrate to form a plurality of chips, individual ones of the plurality of chips having a width of about 20 mm; grinding the semiconductor substrate, wherein during the grinding the plurality of chips is structurally supported with a first film; before removing the first film, structurally supporting the semiconductor substrate with a second film, the second film being located on an opposite side of the semiconductor substrate from the first film; modifying the structural support provided by the second film by patterning a trench, the trench having at least one dimension larger than 20 mm; and removing the first film.

16. The method of claim 15, wherein the trench is annular in shape.

17. The method of claim 15, wherein the trench is straight.

18. The method of claim 15, wherein the trench is aligned with one or more crystalline planes or orientations of the semiconductor substrate.

19. The method of claim 15, wherein the depositing the plurality of layers is performed at least in part at a temperature of 400? C. or more and the structurally supporting the semiconductor substrate with the second film is performed at a temperature of less than 400? C.

20. The method of claim 15, wherein the grinding the semiconductor substrate reduces a thickness of the semiconductor substrate to less than 25 ?m.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) For a more complete understanding of the present invention, and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:

(2) FIGS. 1a and 1b illustrate warping of a semiconductor substrate and an integrated circuit chip, respectively, arising from stress imposed by overlying thin films;

(3) FIGS. 2a through 2c illustrate in cross section steps in the manufacture of a first illustrative embodiment of the present invention;

(4) FIG. 3a through 3c illustrate in plan view the bottom side of an exemplary wafer having a stress tuning layer on the bottom side;

(5) FIG. 3d illustrates in plan view the top side of an exemplary wafer having a stress tuning layer on the bottom side;

(6) FIGS. 4a and 4b schematically illustrate in cross-sectional view steps in the manufacture of an alternative embodiment of the present invention;

(7) FIGS. 5a and 5b schematically illustrate alternative flip chip packaging embodiments of the present invention; and

(8) FIGS. 6a and 6b schematically illustrate wire bond packaging embodiments of the present invention; and

(9) FIG. 7 graphically illustrates the relationship between wafer diameter and warpage imposed by thin films formed on the wafer.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

(10) The making and using of the presently preferred embodiments are discussed in detail below. It should be appreciated, however, that the present invention provides many applicable inventive concepts that can be embodied in a wide variety of specific contexts. The specific embodiments discussed are merely illustrative of specific ways to make and use the invention, and do not limit the scope of the invention

(11) FIGS. 2a through 2c illustrate in cross-sectional detail an illustrative embodiment of the present invention. Semiconductor wafer 12 has formed thereon a plurality of layers 9, 11, 13, which layers are collectively referred to herein as film or films 14. Although only three layers are shown for illustration, one skilled in the art will recognize that any number of layers may be formed on wafer 12 (sometimes alternatively referred to herein as substrate 12) during the process of manufacturing an integrated circuit. In the illustrative embodiments, wafer 12 is a bulk silicon wafer as is common in the art of, perhaps 8 inches, 12 inches, or even 16 inches in diameter. Typically bulk silicon wafer 12 is about 620 ?m (about 31 mils) thick during the integrated circuit manufacturing process. In other embodiments, wafer 12 may be formed of a thin semiconductor layer formed on a buried oxide (box) layer, which in turn is formed on a bulk substrate such as silicon. This so-called silicon on insulator (SOI) configuration is well known in the art and so the details are not illustrated in the figures.

(12) Film 14 is composed of numerous layers, sometimes referred to as thin films, that are formed as part of a typical manufacturing process for forming an integrated circuit. As is known in the art, integrated circuits are typically formed of doped regions (not shown) formed at least partially in an upper surface of the substrate 12. Various layers are then formed atop of substrate 12 and patterned to complete the integrated circuit manufacturing process. These thin films include, e.g., one or more doped polysilicon layers that may be used to form gate electrodes, a contact etch stop layer, a inter-layer dielectric layer (ILD), inter-metal dielectric layers (IMDs), metallic interconnect layers, etch stop layers, and the like. Commonly a passivation layer is formed atop the structure to protect the integrated circuit from contamination, moisture, and the like. In a typical integrated circuit having eight metallization layers and the concomitant IMD and etch stop layers, more than twenty different thin films may be formed on the wafer surface during manufacturing.

(13) The films that comprise collective film 14 are deposited in a variety of manners, including chemical vapor deposition (CVD), plasma enhanced vapor deposition (PEVD), atomic layer deposition (ALD), sputtering, electro-plating, electro-less plating, and the like. The films are deposited at elevated temperatures, typically at 400 C or more. During deposition, the respective films generally do not impose a stress on the underlying layer. After the deposition process, however, as the device returns to room temperature, the different coefficients of thermal expansion between the wafer and the respective thin films formed on the wafer come into play. Due to the different coefficients of thermal expansion and the modulus of the respective thin films, stress arises in the films and a complimentary stress is imposed on the underlying wafer. It is this stress that causes the wafer and subsequently formed die to warp, as shown in FIGS. 1a and 1b, respectively.

(14) One skilled in the art will recognize that certain component films of composite film 14 will impose an inherent compressive stress on underlying wafer 12. Other component films may impose an inherent tensile stress on wafer 12. Hence, the stress caused by one film might tend to counter-balance or negate the effects of the stress caused by another film. Empirical evidence suggests, however, that with conventional integrated circuit processes, particularly for MOS processing, the collective stresses of the composite films will cause film 14 to impart an overall tensile stress on underlying silicon wafer 12. The magnitude of this stress will depend upon the composition and deposition parameters for the individual layers of film 14 as well as upon the composition of wafer 12. In some embodiments, however, film 14 may impart a compressive stress on underlying wafer 12.

(15) As illustrated in FIG. 2a, wafer 12 has a thickness, typically around 620 ?m (31 mils). Traditionally, integrated circuits provided satisfactory performance with a wafer having such a thickness. The wafer would be sawed into individual chips, and these individual chips would typically be mounted onto a supporting lead frame before being encapsulated. In emerging technologies, however, it is desirable to considerably thin the wafer thickness prior to packaging.

(16) FIG. 2b illustrates the wafer of FIG. 2a after a protective film 16, often called blue tape or UV tape, has been formed over the top surface, sometimes referred to as the front service, of the semiconductor device. Note that this occurs after the integrated circuit has been fully manufactured, including the formation of metallization layers and a passivation layer, as is well known in the art. FIG. 2b also schematically illustrates exemplary bond pads 20 formed in a top film(s) of the device. As is known in the art, these bond pads 20 are electrical contact points for electrically connecting the integrated circuit device to the outside world. Bond pads 20, which are formed before applying protective film 16, may also impose additional tensile or compressive stress on wafer 12. Protective film 16 is applied to the device to protect the integrated circuit from mechanical stress and from contamination during the subsequent back grinding process. Next, as is known in the art, wafer 12 is back-ground, or polished, to decrease the wafer's thickness. This back grinding can be accomplished in a number of well-known ways and the specifics are not necessary for an understanding of the present invention.

(17) The result of the back grinding or polishing step is illustrated in FIG. 2c, wherein wafer 12 has been reduced to a thickness of no more than perhaps 75 ?m (3 mils) to perhaps less than 50 ?m (2 mils) or even 25 ?m (1 mil). Note that protective film 16 remains on the device even after the back grinding process. Protective film 16 provides mechanical support to the wafer and prevents the warping that would otherwise occur due to the stress of film 14. Next, and as further illustrated in FIG. 2c, stress tuning layer 18 is formed on the back side, or bottom side, of wafer 12. In this embodiment, stress tuning layer 18 is formed on the surface of wafer 12 opposite from the surface on which is formed layer 14.

(18) In an illustrative embodiment, layer 14 imposes an inherent tensile stress on wafer 12. If left unopposed, this tensile stress would tend to bow or warp wafer 12 in the manner illustrated in FIG. 1a, as illustrated by deflection arrows 17 in FIG. 2c. As discussed above, protective film 16 presents this bowing from occurring, but protective film 16 has to be removed in order to complete the packaging of the integrated device. Therefore, stress tuning layer 18 is preferably formed on the backside of wafer 12 prior to removal of protective film 16. Stress tuning layer is selected and deposited to that it also imparts an inherent tensile stress on wafer 12. This tensile stress is imparted on the opposite surface as the stress imparted by film 14, however, as shown by deflection arrows 19. As is schematically illustrated by the deflection arrows 17, 19, the stress imparted by stress tuning layer 18 tends to negate or counter-balance the stress caused by films 14. The result is that wafer 12, and the subsequently formed chips when wafer 12 is sawed apart, is much less prone to warping. After stress tuning layer 18 is formed, protective film 16 can be removed and packaging can continue.

(19) One skilled in the art will recognize that a variety of materials and deposition techniques can be employed for forming stress tuning layer 18 on the backside of wafer 12. Recall that one of the driving motivations behind the present invention is the desire to decrease the overall thickness of the resulting integrated circuit. For this reason, it is preferable that stress tuning layer be as thin as possible, while still offsetting the effects of film 14. In the illustrative embodiments, stress tuning layer 18 has a thickness of less than about 20 ?m. Dielectric materials such as silicon nitride, silicon oxide, silicon oxynitride, and the like are good candidates for stress tuning layer 18, as these materials and methods for depositing them are well known and common in the industry. Silicon nitride, in particular, has stress properties that can be relatively well controlled through the deposition techniques employed. Alternatively, other dielectrics such as low-k dielectric, polyimide, glass, plastic, ceramic, molding compound, and the like could be employed. Exemplary low-k dielectrics include carbon-doped silicon oxide, fluorine-doped silicon oxide, silicon carbide.

(20) In still other embodiments, a conductive material such as nickel, chromium, or the like could be employed for stress tuning layer 18. Such materials may provide added benefits such as better thermal conductivity and enhanced grounding capacity for the wafer. Generally, it is desirable that whatever material is selected, stress tuning layer be deposited at a temperature of below about 400 C. This is particularly significant because of the need to stay within a pre-defined thermal budget, as is known in the art, in order to, e.g., avoid excessive migration of doped impurity regions.

(21) Various techniques can be employed for depositing stress tuning layer 18, including CVD, PECVD, spin-on coating, and the like. After stress tuning layer 18 is formed, protective film 16 can be removed without concern for warpage of wafer 12. Once protective film 16 is removed, and after a subsequent dicing step, electrical contact can be made to bond pads 20.

(22) In some embodiments, electrical contact made is by wire bonding to the bond pads. In other embodiments, electrical contact can be made by placing the device over a substrate upon which solder bumps have been formed, aligned with the placement of contact pads 20 in the so-called flip chip configuration.

(23) Stress tuning layer can remain blanket deposited onto wafer 12, or stress tuning layer can be formed and subsequently patterned. Patterning stress tuning layer 18 can be employed to further tune or adjust the stress imposed upon wafer 12. For instance, it may be desirable to pattern slots in stress tuning layer in order to concentrate or reduce the amount of stress being applied to select regions of wafer 12. In other embodiments, stress tuning layer may be removed from the portions of wafer 12 corresponding to the scribe lines (being the lines along which wafer 12 will ultimately be sliced to form the individual chips). This approach may be particularly advantageous in order to reduce the possibility that stress tuning layer 18 will peal, crack, or delaminate as a result of the mechanical stresses imposed during the sawing process.

(24) FIG. 3a illustrates in plan view the back side or bottom side of wafer 12 after stress tuning layer 18 has been blanket deposited. FIGS. 3b and 3c illustrates in plan view the back side of wafer 12 after stress tuning layer has been patterned as described above. In FIG. 3b, trenches 22 in the form of concentric rings are formed in stress tuning layer, thus dividing stress tuning layer into concentric annular regions. A variation of FIG. 3b might involve removing stress turning layer entirely from the central portion of wafer 12 and leaving one (or more) large annular region around the periphery of wafer 12, where the effects of warpage are likely to be greatest. FIG. 3c illustrates yet another variation. In this example, trenches 22 are formed in parallel lines across wafer 12, thus forming stress tuning layer 19 into a series of parallel strips running across wafer 12. Other examples include checker board type patterns formed in stress tuning layer 18 and removing stress tuning layer from the periphery of the wafer or the periphery of individual chips. Due to the crystalline nature of wafer 12, it may be that wafer 12 is more prone to warping or strain in one direction as opposed to another direction. This can be accommodated that patterning stress tuning layer 18 in so that the stress tuning layer patterns are substantially aligned with one or more crystalline planes or orientations of the wafer. Numerous variations to the illustrated patterns will become apparent to one of skill in the art with routine experimentation and are within the contemplated scope of the present invention. Stress tuning effects can be adjusted by controlling the thickness of the stress tuning layer in combination with patterning or removing portions of the stress tuning layer. While illustrative patterns have been discussed, many other patterns are possible, including specialized patterns, which may be empirically derived, to match localized stress requirements.

(25) FIG. 3d schematically illustrates the top side or front side of wafer 12, the side whereon the integrated circuit devices are formed, in plan view. Note that this is the opposite side of the wafer illustrated in FIGS. 3a through 3c. FIG. 3d illustrates scribe lines 24 running across wafer 12. These scribe lines 24, define the dimensions of the individual chips 26 into which wafer 12 will be sawed, as is known in the art. It is believed that embodiments of the present invention will have particularly beneficial impact on relatively large chips 26. For instance, chips having a width t.sub.1 of about 20 mm or greater and a width t.sub.1 of about 20 mm or greater is particularly prone to warpage resulting from the stress imposed by films 14. While the present invention is not necessarily limited to chips of such dimensions, it is believed that the present invention would be particularly advantageous for chips having dimensions of about 20?20 mm, or greater, and having a thickness of about 150 ?m or less.

(26) FIGS. 4a and 4b illustrate an alternate embodiment in which stress tuning layer 18 is formed on a top surface of wafer 12, i.e., on top of film 14. As shown in FIG. 3a, stress tuning layer 18 is formed directly atop film 14 after formation of the integrated circuit is otherwise substantially complete. It should be noted that in this illustrated embodiment, stress tuning layer 18 is other than a layer normally associated with fabrication of an integrated circuit. In other words, stress tuning layer 18 should not be confused with a passivation layer or other component layer of film 14 typically formed in the manufacture of an integrated circuit. An advantageous feature of a separate stress tuning layer is to allow stress balancing without sacrificing other properties of the passivation layer. It is, hence, contemplated that the integrated circuit device will have a passivation layer, etc., and will further have stress tuning layer in addition to those other layers. Because stress tuning layer is formed atop film 14 prior to further processing, it may not be necessary to form protective film 16 on the device prior to backgrinding wafer 12. Wafer 12 can then be background or polished to a desired thickness as illustrated by dotted line 24.

(27) In the alternative, after formation of stress tuning layer 18 atop film 14, protective film 16 can be formed directly atop stress tuning layer 18, as illustrated in FIG. 4b. Processing can then continue with wafer 12 being thinned down to a desired thickness, as also shown in FIG. 4b by dotted line 24.

(28) FIG. 4b also shows schematically the stress imposed by film 14 (deflection arrows 17) and stress tuning layer 18 (deflection arrows 19). Note that in this embodimentwhere stress tuning layer 18 and film 14 are both formed on the same surface of the substrate, stress tuning layer 18 must impose an opposite stress as that imposed by film 14. In other words, assuming film 14 imparts a tensile stress on wafer 12, stress tuning layer must in this embodiment be selected and deposited so as to impart a compressive stress on wafer 12 in order to off-set or counter balance film 14. Conversely, if film 14 provides an overall compressive stress, stress tuning layer 18 must be selected and deposited to provide a counter-balancing tensile stress.

(29) FIGS. 5a and 5b schematically illustrate in cross section an exemplary chip in an integrated circuit package, in this case a so-called flip-chip package. FIG. 5a illustrates a flip-chip package employing a chip manufactured with the stress tuning layer on the back side of wafer 12, such as was illustrated in FIGS. 2a through 2c. Specifically, wafer 12, which has already been thinned as described above, is flipped over so that its top surface faces an underlying supporting substrate 30. Solder bumps or balls 32 formed on substrate 30 are placed so as to align with contact pads 20 of the integrated circuit device. When the device is placed upon these bumps 32, electrical contact is established between the device and the package. Typically, an underfill material (not shown) is injected or otherwise formed in the space between wafer 12 and substrate 30. This underfill material is typically formed in a fluid state and subsequently cured to harden. This underfill may also have a coefficient of thermal expansion and a modulus that differs from wafer 12, thus further contributing to stress applied on wafer 12. Stress tuning layer 18 can be selected and deposited to also counteract the effects of underfill, if appropriate.

(30) FIG. 5b illustrates a flip chip package employing a stress tuning layer 18 on the front surface of wafer 12. In such an embodiment, appropriate openings must be formed in stress tuning layer 18 (as well as in any underlying passivation layer) in order to allow for good mechanical and electrical contact between contact pads 20 and solder bumps 32. Again, an underfill material between wafer 12 and supporting substrate 30 might also contribute to the overall stress imposed on wafer 12which can be at least partially counteracted by stress tuning layer 18, if appropriate.

(31) FIGS. 6a and 6b illustrate yet another embodiment in which wafer 12 is placed in a wire bonding package. FIG. 6a illustrates an embodiment wherein stress tuning layer is formed on the back side of wafer 12. In this embodiment, stress tuning layer 18 faces an underlying support substrate 30, such as a lead frame pad. Bond wires 38 contact and electrically connect bond pads 20 to lead fingers 36. Glue or underfill 34 is formed between stress tuning layer 18 and lead pad substrate 30. As discussed above, stress tuning layer may, in some embodiments, be selected and formed so as to offset at least some of the stress otherwise caused by underfill 34. As is known in the art, the entire assembly illustrated in FIGS. 6a (and 6b as will be described below), is encapsulated in a plastic or ceramic package, which package seals and protects the chip from environmental and mechanical factors.

(32) FIG. 6b illustrates a wire bond package with a chip that has had stress tuning layer 18 formed on the front or top surface of wafer 12. Note in this instance that appropriate holes must be formed in stress tuning layer 18 in order that bond wires 38 can contact and electrically connect bond pads 20 to lead fingers 36.

(33) FIG. 7 graphically illustrates the relationship between wafer diameter and the amount of warpage displacement arising from thin films formed on the wafer. As shown by the legend, different thin films, including a nitride film (NIT), aluminum film (Al), and polyimide film (PI) were modeled for wafers ranging from three inches in diameter to sixteen inches in diameter and the amount of resulting warpage was simulated. Simulation was based upon the coefficient of thermal expansion (CTE) mismatch between the thin film and the underlying silicon wafer and the modulus of the film. As is clear, warpage increases by orders of magnitude for relatively thin wafers of, say, 25 ?m (see lines 101 and 103) compared to the warpage for relatively thick wafers of 725 ?m (see lines 105 and 107). It is this warpage that may be substantially ameliorated by embodiments of the present invention, in the manners described above.

(34) Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims. Moreover, the scope of the present application is not intended to be limited to the particular embodiments of the process, machine, manufacture, and composition of matter, means, methods and steps described in the specification. As one of ordinary skill in the art will readily appreciate from the disclosure of the present invention, processes, machines, manufacture, compositions of matter, means, methods, or steps, presently existing or later to be developed, that perform substantially the same function or achieve substantially the same result as the corresponding embodiments described herein may be utilized according to the present invention. Accordingly, the appended claims are intended to include within their scope such processes, machines, manufacture, compositions of matter, means, methods, or steps.