Abstract
A multi-pin wafer level chip scale package is achieved. One or more solder pillars and one or more solder blocks are formed on a silicon wafer wherein the one or more solder pillars and the one or more solder blocks all have a top surface in a same horizontal plane. A pillar metal layer underlies the one or more solder pillars and electrically contacts the one or more solder pillars with the silicon wafer through an opening in a polymer layer over a passivation layer. A block metal layer underlies the one or more solder blocks and electrically contacts the one or more solder pillars with the silicon wafer through a plurality of via openings through the polymer layer over the passivation layer wherein the block metal layer is thicker than the pillar metal layer.
Claims
1. A multi-pin wafer level chip scale package comprising: one or more solder pillars and one or more solder blocks on a silicon wafer wherein said one or more solder pillars and said one or more solder blocks all have a top surface in a same horizontal plane and wherein at least one side of said solder blocks in a two-dimensional plane is greater than 600 μm and wherein said solder pillars are thinner and narrower than said solder blocks; a pillar metal layer directly underlying each of said one or more solder pillars and electrically connecting said one or more solder pillars with said silicon wafer through an opening in a polymer layer over a passivation layer on said silicon wafer, wherein said pillar metal layer comprises: a redistribution layer (RDL) trace contacting said silicon wafer at a silicon pad exposed by said opening in said polymer layer over said passivation layer; and a block metal layer directly underlying each of said one or more solder blocks and electrically connecting said one or more solder blocks with said silicon wafer through a plurality of via openings through said polymer layer over said passivation layer on said silicon wafer, wherein said block metal layer comprises: a redistribution layer via (RDL_VIA) contacting said silicon wafer through said plurality of via openings in said polymer layer and said passivation layer to said silicon wafer; and an under block metal (UBM) layer covering said RDL_VIA layer in said plurality of via openings wherein said block metal layer is thicker and wider than said pillar metal layer wherein said RDL trace has a thickness of at least 4 μm, said RDL_VIA layer has a thickness of at least 25 μm, said UBM layer has a thickness of at least 25 μm and a width of at least 600 μm in at least one direction, said solder pillar has a height of at least 120 μm and said solder block has a height of at least 100 μm and a width of at least 600 μm.
2. A multi-pin wafer level chip scale package comprising: one or more solder pillars and one or more solder blocks on a silicon wafer wherein said one or more solder pillars and said one or more solder blocks all have a top surface in a same horizontal plane and wherein at least one side of said solder blocks in a two-dimensional plane is greater than 600 μm and wherein said solder pillars are thinner and narrower than said solder blocks; a pillar metal layer directly underlying each of said one or more solder pillars and electrically connecting said one or more solder pillars with said silicon wafer through an opening in a polymer layer over a passivation layer on said silicon wafer, wherein said pillar metal layer comprises: an under pillar metal (UPM) trace contacting said silicon wafer at a silicon pad exposed by said opening in said polymer layer over said passivation layer; and a block metal layer directly underlying each of said one or more solder blocks and electrically connecting said one or more solder blocks with said silicon wafer through a plurality of via openings through said polymer layer over said passivation layer on said silicon wafer, wherein said block metal layer comprises: an under block metal layer (UBM) contacting said silicon wafer through said plurality of via openings in said polymer layer and said passivation layer to said silicon wafer and having a thickness overlying said plurality of via openings wherein said UPM trace has a thickness of at least 8 μm and at most 25 μm, said UBM layer has a thickness of at least 25 μm and a width of at least 600 μm in at least one direction, said solder pillar has a height of at least 120 μm and said solder block has a height of at least 100 μm and a width of at least 600 μm.
3. The multi-pin wafer level chip scale package according to claim 2 wherein said pillar metal layer further comprises: a redistribution layer (RDL) trace contacting said silicon wafer at said silicon pad exposed by said opening in said polymer layer over said passivation layer; and said under pillar metal (UPM) trace on said RDL trace.
4. The multi-pin wafer level chip scale package according to claim 2 wherein said block metal layer further comprises: a redistribution layer via (RDL_VIA) contacting said silicon wafer through said plurality of via openings in said polymer layer and said passivation layer to said silicon wafer; and said under block metal (UBM) layer covering said RDL_VIA layer in said plurality of via openings.
5. The multi-pin wafer level chip scale package according to claim 3 wherein said RDL trace has a thickness of at least 4 μm.
6. The multi-pin wafer level chip scale package according to claim 4 wherein said RDL_VIA layer has a thickness of at least 25 μm.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) In the accompanying drawings forming a material part of this description, there is shown:
(2) FIG. 1 is a cross-sectional representation of a wafer level chip scale package (WLCSP) of the prior art.
(3) FIG. 2A is a cross-sectional representation of a MP-WLCSP in a first preferred embodiment of the present disclosure.
(4) FIG. 2B is an alternative cross-sectional representation of a MP-WLCSP in a second preferred embodiment of the present disclosure.
(5) FIG. 3A is an enlarged cross-sectional representation of a portion of the MP-WLCSP of FIG. 2A in a first preferred embodiment of the present disclosure.
(6) FIG. 3B is an enlarged cross-sectional representation of a portion of the MP-WLCSP of FIG. 2B in a second preferred embodiment of the present disclosure.
(7) FIG. 4A is an isometric view of solder balls in a WLCSP of the prior art.
(8) FIG. 4B is an isometric view of a solder block in the MP-WLCSP of the present disclosure.
(9) FIGS. 5A-5V are cross-sectional representations of steps in a process of a first preferred embodiment of the present disclosure.
(10) FIGS. 6A, 6C-6J, 6L, and 6N are cross-sectional representations of steps in a process of a second preferred embodiment of the present disclosure.
(11) FIG. 6B is a wafer level representation of the RDL_VIA strip placement for the second preferred embodiment of the present disclosure.
(12) FIG. 6K is a wafer level representation of the UBM strip placement for the second preferred embodiment of the present disclosure.
(13) FIG. 6M is a wafer level representation after a sawing and trimming step for the second preferred embodiment of the present disclosure.
(14) FIGS. 7A-7L are cross-sectional representations of steps in a process of a third preferred embodiment of the present disclosure.
(15) FIG. 8 is a top view representation of a sample layout of solder blocks and solder pillars in the present disclosure.
(16) FIG. 9 is a top view representation of various UBM shapes in the present disclosure.
(17) FIG. 10 is a top view representation of various solder block structure shapes in the present invention.
(18) FIG. 11 is a cross-sectional representation of alternative pillar top surfaces in the present disclosure.
(19) FIG. 12 is a cross-sectional representation of a first alternative embodiment of the present disclosure.
(20) FIG. 13 is a cross-sectional representation of a second alternative embodiment of the present disclosure.
(21) FIG. 14 is a cross-sectional representation of a third alternative embodiment of the present disclosure.
(22) FIG. 15 is a cross-sectional representation of a fourth alternative embodiment of the present disclosure.
(23) FIG. 16 is a cross-sectional representation of a fifth alternative embodiment of the present disclosure.
DETAILED DESCRIPTION
(24) The present disclosure presents chip scale packaging technology (WLCSP) having different material stackup and different package pins to support the future demands of the semiconductor industry in a far better way than the existing chip scale packaging technology. FIG. 2A illustrates the cross section of the present disclosure, Multi Pin WLCSP, showing the different material layers and types of solder contacts. FIG. 2B illustrates the same chip scale package fabricated using an alternative assembly processing method.
(25) The wafer level packaging metal layer 22a and 22b (Under Pillar and Under Block Metal) have differing thicknesses and/or widths correlated with expected current density. That is, pins 24a (solder pillars) for low current units will have thinner and narrower underlying metal layer 22a (Under Pillar Metal) while pins 24b (solder blocks) for high current units will have thicker and wider underlying metal layer 22b (Under Block Metal). Low current is considered to be less than or equal to about 2 amperes (A) while high current is greater than about 2 A.
(26) The wafer level packaging metal layer 18a and 18b (RDL traces and RDL_VIAs) have differing thicknesses and/or widths correlated with expected current density. That is, metal layer carrying low currents, 22a (Under Pillar Metal) is connected to silicon pads through thinner and narrower underlying metal distribution layer 18a (RDL traces), while high current carrying metal layer 22b (Under Block Metal) is connected to silicon pads through thicker and wider underlying metal layer 18b (RDL_VIA).
(27) The thicker/wider Under Block Metal (UBM) 22b connected to solder block (SB) 24b and the relatively thinner/narrower Under Pillar Metal (UPM) 22a connected to Solder Pillars (SP) 24a allows for multiple solder inter-connects for packaging. There are no point contacts as would be the case with a solder ball, but the flat cylindrical or pillar shaped solder contact 24b provides a large contact area. The solder blocks 24b are plated, not bumped. The remaining solder interconnects 24a are also non-spheres. They are cylindrical or pillar shaped and plated, not bumped.
(28) The multi-pin WLCSP of the present disclosure provides better thermal, electrical, mechanical, and board level performance than the existing WLSCP packaging. It supports a higher current and provides a better package life and quality when supporting high DC current than the existing packages. Standard industry packaging design rules are followed by the disclosed multi-pin WLCSP.
(29) FIG. 3A illustrates a high current carry packaging pin and underlying structures in enlarged cross-section. In this embodiment, solder block (SB) 24b sits on top of a thick and wide under block metal (UBM) 22b. The UBM is connected to the silicon pads 12 by means of RDL_VIA 18b. The solder block and UBM are both wider than about 600 μm. FIG. 3B illustrates a similar high current carry packaging fabricated using an alternative assembly processing method.
(30) In standard WLCSP (Prior Art FIG. 1), electromigration occurs because of high current crowding at the Under Ball Metal (UBM) (22)/solder ball (24) interface since the cross-sectional area of the solder ball is larger than the UBM cross sectional area. In the MP-WLCSP of the present disclosure, as seen for example in FIG. 3A, the cross sectional area of the Under Block Metal (UBM) 22b is almost the same as that of the solder block 24b. The surface contact between the UBM 22b and the solder block 24b is better than UBM 22-Solder ball 24 contact in the WLCSP. This also helps in achieving higher currents by reducing the current crowding.
(31) Table 1 (below) compares the maximum DC current values per packaging pin at three different temperatures for the typical WLCSP (column 2) and the MP-WLCSP of the present disclosure (column 3). Maximum current values for WLCSP are taken from the EM study done by an assembly supplier on a package having a 250 μm solder ball (24), 245 μm UBM (22) diameter, and 8.3 μm UBM (22) thickness. The maximum DC current values for MP-WLSCP are calculated values, given a solder block (24b) width of 1200 μm, UBM (22b) thickness of 50 μm and UBM (22b) width of about 1200 μm.
(32) TABLE-US-00001 TABLE 1 Temperature WLCSP Current/ MP-WLCSP Current/ ° C. solder ball (A) solder block (A) 100 1.28 ~11.396 125 0.92 ~7.87 140 0.77 ~6.14
(33) As shown in Table 2, an electrical simulation was run on high power blocks designed using WLCSP (column 2) and MP-WLCSP (column 3) stackup schemes. In WLCSP (FIG. 4A), four solder balls 24 were used to connect the package/chip to the PCB. In MP-WLCSP (FIG. 4B), one solder block 24b was used to connect the package/chip to the PCB. The parasitic extractions were done under the same AC and DC conditions. From the results shown in Table 2 (below), one can see that the parasitic resistance and inductance of MP-WLCSP are less than those of the WLCSP. The parasitic self capacitance is slightly higher for MP-WLCSP, due to the fact that the surface area of the Under Block Metal (22b) copper is greater in MP-WLCSP compared to standard WLCSP.
(34) TABLE-US-00002 TABLE 2 Parasitic comparison WLCSP MP-WLCSP DC resistance (mOhm) 0.337 0.0866 AC resistance (mOhm) 0.410 0.132 DC inductance (nH) 0.017 0.016 AC inductance (nH) 0.014 0.013 Self Capacitance (fF) 34.063 35.22
(35) In some of the WLCSP packages, some manufacturing rules are compromised (violated) in order to achieve lower parasitic values, to improve the current density, and to reduce the chip size. Such violations might result in reducing the board level reliability of the chip scale packages. In MP-WLCSP, the new stack up dimensions and thicknesses help us achieve very good electrical performance without having to compromise or violate the critical chip scale packaging rules, hence making it more robust with respect to board level reliability and mechanical performance. Some of the important manufacturing rules that are not violated by the MP-WLCSP of the present disclosure include no silicon pads (12) under the Under Block Metal (22b) and Under Pillar Metal (22a), a minimum silicon pad size (12) of 42 μm, and an Under Block Metal (22b) and Under Pillar Metal (22a) density of greater than 25% for a chip size greater than or equal to 5×5 mm.sup.2.
(36) Referring now more particularly to FIGS. 5A-5V, a first preferred embodiment process of the present disclosure will be described. FIG. 5A illustrates a wafer incoming from the wafer fabrication area. Silicon 10 has been coated with a passivation layer 14 through which silicon connection pads 12 have been formed. After inspection of the silicon wafer, a first polymer layer 16 is coated onto the wafer and patterned as shown in FIG. 5B. For example, the polymer may be a polyimide or a polybenzoxazole or any other polymer. The thickness of the polymer can be >=7.5 μm.
(37) Now, referring to FIGS. 5C-5V, a departure from the typical WLCSP process can be seen. In FIG. 5C, RDL trace (low current carrying trace) 18a seed layer is sputtered uniformly on the top of Polymer1 (16). The material used can be Ti/Cu or any other adhesive or seed metal. The preferred thickness of the seed layer 18a is <=0.5 μm.
(38) In FIG. 5D, photo resist PR1 (25) is coated and developed. The development of the PR1 will be based on the mask designed for RDL traces (18a). Then, a RDL trace metal is plated over the exposed RDL trace seed layer as illustrated in FIG. 5E. The preferred material used for plating is Copper (Cu), but any other plating metal layer can be used. The combination of plated metal layer and seed layer forms the RDL trace (18a). After removing the photo resist PR1 (25) and etching the unwanted RDL trace seed layer, the desired RDL trace 18a is left over. This is illustrated in FIG. 5F. The preferred thickness of the RDL trace 18a is >=4 μm and <=25 μm.
(39) Next, as shown in FIG. 5G, photo resist PR2 (27) is coated and developed. The development of the PR2 will be based on the mask designed for RDL_VIA (18b). Then, as illustrated in FIG. 5H, a RDL_VIA seed layer is sputtered uniformly on the top of PR2 (27). The material used can be Ti/Cu or any other adhesive or seed metal. The preferred thickness of the seed layer 18b is <=2 μm.
(40) Now, photo resist PR3 (29) is coated and developed on top of the seed layer, as shown in FIG. 5I, based on the mask designed for the RDL_VIA (18b) plating. FIG. 5J illustrates plating of the RDL_VIA metal over the exposed RDL_VIA seed layer. The preferred material used for plating is Copper (Cu), but any other plating metal can be used. The combination of plated metal layer and seed layer forms the RDL_VIA (18b).
(41) After removing the photo resist PR2 (27) and PR3 (29) and etching the unwanted RDL_VIA seed layer, the desired RDL_VIA 18b is left over. This is illustrated in FIG. 5K. The preferred thickness of RDL_VIA 18b is >=25 μm and <50 μm.
(42) Next, as shown in FIG. 5L, a second polymer layer 20 is deposited to a thickness of 7.5 μm or more and patterned to provide openings for placing of pins. For example, the polymer may be a polyimide or a polybenzoxazole, or any other polymer.
(43) Referring now to FIG. 5M, the Under Pillar Metal (UPM) seed layer 22a for the low current carrying UPM is sputtered. This layer may be Titanium/copper or any other adhesive or seed metal. The preferred thickness of the seed layer 22a is <=0.5 μm.
(44) As illustrated in FIG. 5N, photo resist PR4 (31) is coated and developed. The development of the PR4 will be based on the mask designed for UPM (22a). Then, a UPM metal is plated over the exposed UPM seed layer as illustrated in FIG. 5O. The preferred material used for plating is Copper (Cu), but any other plating metal could be used. The combination of plated metal layer and seed layer forms the UPM 22a having a preferred thickness >=8 μm and <=25 μm.
(45) Now, as shown in FIG. 5P, the solder pillar (SP) 24a is plated over the UPM 22a. The plating material may be SAC405, SAC305, or SAC Q or any other solder alloy.
(46) As shown in FIG. 5Q, the PR4 31 is stripped and the UPM seed layer not covered by the solder pillar 24a is etched away.
(47) Next, the high current area pins are to be formed. As shown in FIG. 5R, a photo resist PR5 (33) is coated and developed, based on the mask designed for the Under Block Metal (UBM) 22b. A UBM seed layer 22b for the high current carrying signals is sputtered over the PR5 and within the openings. This layer may be Titanium/copper or any other adhesive or seed metal. The preferred thickness of the seed layer 22b is <=2 μm.
(48) Now, another photo resist layer PR6 (35) is coated and developed as shown in FIG. 5S. Next, as shown in FIG. 5T, a UBM metal is plated over the exposed UBM seed layer. The preferred material used for plating is Copper (Cu), but any other plating metal could be used. The combination of plated metal layer and seed layer forms the UBM 22b having a preferred thickness of >=25 μm and <=50 μm.
(49) Now, as shown in FIG. 5U, the solder block (SB) 24b is plated over the UBM 22b. The plating material may be SAC405, SAC305, or SAC Q or any other solder alloy.
(50) As shown in FIG. 5V, the PR5 33 and PR6 35 are stripped and the UBM seed layer not covered by the solder block 24b is etched away.
(51) It is important that both the solder pillar 24a and solder block 24b are plated to the same final level so that both end up in the same horizontal plane. The final thickness of the solder pillars will be greater than about 120 μm and the final thickness of the solder blocks will be greater than about 100 μm.
(52) FIG. 8 illustrates a top view of a sample layout of solder pillars 24a and solder blocks 24b. For better understanding, the solder pillars 24a are shown with hatching and solder blocks 24b are not hatched. The assembly processing cost of MP-WLCSP is almost the same as the conventional WLCSP process cost.
(53) FIG. 9 illustrates various UBM shapes. FIG. 10 illustrates various solder block structure shapes. Any shape is possible, but at least one side of the UBM (FIG. 9) or the solder block (FIG. 10) in a two-dimensional plane should be greater than 600 μm. In both figures, X is the maximum edge to edge distance in one direction and Y is the maximum edge to edge distance in the other direction. In order to support high currents, the UBM and the solder block can have X>Y (if X is greater than 600 μm) or Y>X (if Y is greater than 600 μm) or X=Y (if X and Y are greater than 600 μm).
(54) The solder pillars 24a and blocks 24b may have a flat or a curved top surface. FIG. 11 shows two exemplary pillar curved top surfaces 81 and 82 and a pillar flat top surface 83. Solder blocks also may have either flat or curved top surfaces, not shown. Curved surfaces might be better than flat surfaces from the point of view of soldering the device onto PCB's.
(55) Referring now to FIGS. 5A-5H and 6A-6N, a second preferred embodiment of the present disclosure is described. In this second alternative process, the steps are identical to the first embodiment through FIG. 5H. As shown in FIG. 5H, RDL trace 18a has been formed and the seed layer 18b in the high current area has been deposited.
(56) Now, as illustrated in FIG. 6A, RDL_VIA strip 18b is placed in the vias. FIG. 6B shows a top view of a portion of a simplified wafer 100 having a plurality of chips 102. The strip 18b is placed on all of the chips on the wafer simultaneously. This is a simple way to provide the thick, wide RDL_VIA for the high current carrying package pins. The RDL_VIA strip can be copper or any other metal. After the RDL_VIA strip 18b is placed, the seed layer is etched away where it is not covered by the strip and photoresist 27 is removed, as shown in FIG. 6A.
(57) Next, as shown in FIG. 6C, a second polymer layer 20 is deposited to a thickness of >=7.5 μm and patterned to provide openings for placing of pins. For example, the polymer may be a polyimide or a polybenzoxazole, or any other polymer.
(58) Referring now to FIG. 6D, the UPM seed layer 22a is sputtered to a thickness of <=0.5 μm. This layer may be Titanium/copper or any other adhesive or seed metal.
(59) Referring now to FIG. 6E, a photo resist PR3 (29) is coated and developed. The development of the PR3 will be based on the mask designed for UPM (22a). Then, in FIG. 6F, a UPM metal is plated over the exposed UPM seed layer. The preferred material used for plating is Copper (Cu), but any other plating metal could be used. The combination of plated metal layer and seed layer forms the UPM 22a having a preferred thickness >=8 μm and <=25 μm.
(60) Now, as shown in FIG. 6G, the solder pillar (SP) 24a is plated over the UPM 22a. The plating material may be SAC405, SAC305, or SAC Q or any other solder alloy.
(61) The PR3 29 is stripped and the UPM seed layer not covered by the solder pillar 24a is etched away, as shown in FIG. 6H.
(62) Next, the high current area pins are to be formed. As shown in FIG. 6I, a photo resist PR4 (31) is coated and developed, based on the mask designed for the Under Block Metal (UBM) 22. A UBM seed layer 22 for the high current carrying signals is sputtered over the PR4 and within the openings. This layer may be Titanium/copper or any other adhesive or seed metal. The preferred thickness of the seed layer 22 is <=2 μm.
(63) Now, a UBM strip 22b is placed within the PR4 openings, as shown in FIG. 6J. FIG. 6K is a wafer level view of the placing of the UBM strip 22b on the chips of the wafer. UPM 22a are shown. The solder pillars and photo resist are not shown in this figure.
(64) Returning to FIG. 6J, another photo resist layer PR5 (33) is coated and developed. Now as shown in FIG. 6L, the solder block (SB) 24b is plated over the UBM strip 22b. The plating material may be SAC405, SAC305, or SAC Q or any other solder alloy.
(65) It is important that both the solder pillar 24a and solder block 24b are plated to the same final level so that both end up in the same horizontal plane. The final thickness of the solder pillars will be greater than about 120 μm and the final thickness of the solder blocks will be greater than about 100 μm.
(66) FIG. 6M shows the wafer level view after photoresist stripping and the final step of sawing and trimming away the excess strip material, leaving the RDL_VIA and UBM as shown. FIG. 6N shows the final cross-sectional view after PR4 31 and PR5 33 stripping, etching away the UBM seed layer not covered by the solder block 24b, sawing, and trimming.
(67) A third preferred embodiment of the present disclosure is described with reference to FIGS. 5A-5B and 7A-7L. In this process, the first two steps are identical to FIGS. 5A-5B. Next, as shown in FIG. 7A, RDL layer 18 is sputter deposited over the patterned polymer layer 16 and contacting the silicon pads 12 using a thick uniform sputtering process. The RDL layer may comprise Titanium/Copper or any other seed metal material sputtered to a thickness of <=2 μm.
(68) A first photo resist layer PR1 (25) is coated and developed according to the pattern for the RDL trace, shown in FIG. 7B. The RDL seed layer is etched to reduce its thickness in the low current area exposed by the mask 25. The etching should remove the seed layer up to a thickness of <=1.5 μm so that a final thickness of <=0.5 μm seed layer thickness remains in the low current area.
(69) The PR1 25 is stripped and a new photo resist PR2 27 is formed leaving openings in both the low and high current areas. Copper, or other metal, is plated on the seed layer exposed by mask 27 in a thick uniform plating process, as shown in FIG. 7C, to form plated layer 18b.
(70) The photoresist mask 27 is stripped and another mask PR3 29 is formed to expose only the low current area. The RDL plated layer is etched back to reduce the Cu thickness in the low current area 18a, as shown in FIG. 7D.
(71) Next, as shown in FIG. 7E, the PR3 29 is stripped and the sputtered RDL layer 18 is etched away. Because it is much thicker, the plated RDL layer remains mostly intact.
(72) The thickness of the final RDL_trace 18a is >=4 μm and <=25 μm. The thickness of the final RDL_VIA 18b is >=25 μm and <=50 μm.
(73) Referring now to FIG. 7F, a second polymer layer 20 is deposited to a thickness of >=7.5 μm and patterned to provide openings for placing of pins. For example, the polymer may be a polyimide or a polybenzoxazole or any other polymer.
(74) Referring now to FIG. 7G, the seed layer 22 for the UBM for high current signals and the UPM for low current signals is sputtered to a thickness of about <=2 μm. This layer may be Titanium/copper or any other seed metal. This is a thick, uniform sputtering.
(75) A photo resist PR4 (31) is coated and developed over the UBM layer 22 having openings where the low current carrying pins will be formed, as shown in FIG. 7H. The seed layer 22 is etched to reduce its thickness in the low current areas. The final thickness is <=0.5 μm.
(76) PR4 31 is stripped and a new mask PR5 33 is formed having openings where both high and low current carrying pins are to be formed. Thick uniform metal for UPM and UBM 22b is plated in a thick uniform process in the openings, as shown in FIG. 7I. The plating material may be Copper or any other plating metal.
(77) The PR5 33 is stripped and another mask PR6 35 is formed with an opening only over the low current areas. The UPM layer 22b is etched back, as shown in FIG. 7J to reduce the UPM for the low current pin for a total seed plus plating thickness 22a of >=8 μm and <=25 μm.
(78) PR6 35 is stripped and new mask PR7 37 is formed on seed layer 22 with openings where the pins are to be formed. As shown in FIG. 7K, solder is plated on the UPM and UBM material exposed by the mask 37. It is important that both the solder pillar 24a and solder block 24b are plated to the same final level so that both end up in the same horizontal plane.
(79) Finally, the PR7 37 is removed and the metal 22 not covered by the solder is etched away to complete the package, as shown in FIG. 7L.
(80) The multi-pin WLCSP of the present disclosure provides better thermal, electrical, mechanical, and board level performance than the traditional wafer level chip scale packaging. This package supports higher current coming from analog blocks and has better quality package life when supporting high DC current for a longer time as compared to the WLCSP. The MP-WLCSP exhibits better electrical performance without having to violate any of the package design rules. Process flow, form factor, and assembly processing cost are similar to the traditional WLCSP.
(81) Several alternatives or modifications may be made to the MP-WLCSP of the disclosure.
(82) FIG. 12 illustrates a fan out wafer level chip scale package. The die is further encapsulated in a mold 71. The fan out WLCSP of the present disclosure uses the solder blocks and pilllars as in the standard MP-WLCSP of the present disclosure.
(83) FIG. 13 illustrates a second modification where no RDL vias are used for the high current pins. In this alternative, the UBM 22b is formed within the vias directly on the first polymer layer 16. Advantages of this modification are fewer process steps, hence cost reduction and shorter manufacture cycle time.
(84) FIG. 14 illustrates a third modification where no UPM is used under the solder pillars 24a. The pillars are plated directly on the RDL traces 18a. Advantages of this modification are fewer process steps, hence cost reduction and shorter manufacture cycle time.
(85) FIG. 15 illustrates a fourth modification where no RDL is used under the solder pillars 24a. The UPM traces 22a directly contact the silicon pads 12. Advantages of this modification are fewer process steps, hence cost reduction and shorter manufacture cycle time.
(86) FIG. 16 illustrates a fifth modification where only one polymer layer 20 is used. Since there is only one polymer layer, there is also no RDL/RDL_VIA layer. Advantages of this modification are fewer process steps, hence cost reduction and shorter manufacture cycle time.
(87) Although the preferred embodiment of the present disclosure has been illustrated, and that form has been described in detail, it will be readily understood by those skilled in the art that various modifications may be made therein without departing from the spirit of the disclosure or from the scope of the appended claims.
