Optimizing processes and materials for flip chip reliability: careful material selection can yield FC assemblies with better reliabilities than some surface-mount components.The overall attractiveness of the flip chip A chip packaging technique in which the active area of the chip is "flipped over" facing downward. Instead of facing up and bonded to the package leads with wires from the outside edges of the chip, any surface area of the flip chip can be used for interconnection, which is typically done design has been enhanced in the past few years by continued improvements in achievable reliability and in production yield. As a result, flip chips are now commonly used in advanced high-volume products such as cell phones and global positioning system Global Positioning System: see navigation satellite. Global Positioning System (GPS) Precise satellite-based navigation and location system originally developed for U.S. military use. (GPS) devices and in critical-reliability devices such as cardiac pacemakers cardiac pacemaker A device that delivers a small electric shock to the heart to effect cardiac contraction at a pre-determined rate . To show the high potential of state-of-the-art underfill materials to achieve the highest yield and reliability in specific production environments, a study was undertaken at Fraunhofer IZM IZM (Fraunhofer) Institut für Zuverlaessigkeit und Mikrointegration IZM Izglitibas un zinatnes ministrija (Latvian Ministry of Education and Science) IZM Incredible Zombie Machine to investigate five commercially available, advanced underfill materials. The study used techniques previously developed or refined at Fraunhofer IZM, including the use of video equipment to permit the direct microscopic microscopic /mi·cro·scop·ic/ (mi?kro-skop´ik) 1. of extremely small size; visible only by the aid of the microscope. 2. pertaining or relating to a microscope or to microscopy. visualization Using the computer to convert data into picture form. The most basic visualization is that of turning transaction data and summary information into charts and graphs. Visualization is used in computer-aided design (CAD) to render screen images into 3D models that can be viewed from all of underfill flow rates and the use of an acoustic micro imaging system to nondestructively image material inhomogeneities. These techniques were used to detect voids and irregular filler fill·er 1 n. One that fills, as: a. Something added to augment weight or size or fill space. b. A composition, especially a semisolid that hardens on drying, used to fill pores, cracks, or holes in wood, plaster, particle distribution. Additionally, the influence of various substrate base materials, solder mask An insulating pattern applied to a printed circuit board that exposes only the areas to be soldered. types and geometries on the underfill materials was investigated via acoustic micro imaging. To correlate actual reliability with conclusions drawn about materials and processes, accelerated aging Accelerated aging is a testing method used to estimate the useful lifespan of a product when actual lifespan data is unavailable. This occurs with products that have not existed long enough to have gone through their useful lifespan: for example, a new type of car engine or a new tests were performed. Acoustic micro imaging was also used to nondestructively image the underfill features during this stage of the study. The overall flip chip geometry makes acoustic imaging especially useful because the silicon die itself is nearly transparent to ultrasound. Part of the impetus for this study was the continuous upgrading of underfill materials themselves by manufacturers. Chip sizes have become larger (15 mm+), while the underchip gap has become thinner (<40 microns). These changes have dictated improved resin/hardener systems having minimum reaction time and low viscosity, as well as increasingly smaller filler particles. The five materials investigated here appeared capable of meeting the multiple demands of profitable flip chip processing, including high yield, high throughput, low cost, suitability for large integrated circuits Integrated circuits Miniature electronic circuits produced within and upon a single semiconductor crystal, usually silicon. Integrated circuits range in complexity from simple logic circuits and amplifiers, about 1/20 in. (1. (ICs), fine pitch and insensitivity in·sen·si·tive adj. 1. Not physically sensitive; numb. 2. a. Lacking in sensitivity to the feelings or circumstances of others; unfeeling. b. to printed circuit board (PCB PCB: see polychlorinated biphenyl. PCB in full polychlorinated biphenyl Any of a class of highly stable organic compounds prepared by the reaction of chlorine with biphenyl, a two-ring compound. ) topography topography (təpŏg`rəfē), description or representation of the features and configuration of land surfaces. Topographic maps use symbols and coloring, with particular attention given to the shape and elevations of terrain. . These five materials were also selected because they appear to have low viscosity, short cure times, minimal air entrapment entrapment, in law, the instigation of a crime in the attempt to obtain cause for a criminal prosecution. Situations in which a government operative merely provides the occasion for the commission of a criminal act (e.g. , strong adhesion adhesion /ad·he·sion/ (ad-he´zhun) 1. the property of remaining in close proximity. 2. the stable joining of parts to one another, which may occur abnormally. 3. and post-cure thermo-mechanical properties that would promote long-term reliability. Previous experience and finite element See FEA. modeling (FEM FEM Female FEM Finite Element Method FEM Feminine FEM Finite Element Model FEM Fédération Européenne des Métallurgistes (European Metalworkers' Federation) FEM Faculdade de Engenharia Mecânica (Brasil) ) have suggested that underfill materials in this application should have a Young's modulus Young's modulus [for Thomas Young], number representing (in pounds per square inch or dynes per square centimeter) the ratio of stress to strain for a wire or bar of a given substance. between 2 and 12 MPa, a coefficient of thermal expansion coefficient of thermal expansion, n See expansion, thermal coefficient. (CTE (Coefficient of Thermal Expansion) The difference between the way two materials expand when heat is applied. This is very critical when chips are mounted to printed circuit boards, because the silicon chip expands at a different rate than the plastic board. ) between 20 ppm/[degrees]K and 40 ppm/[degrees]K, a glass transition (Tg) about 15[degrees]K above maximum use temperature and a viscosity below 1 Pa*s at an underfilling temperature between 70[degrees] and 90[degrees]C. The five materials selected met all of these criteria. To further test the behavior of these materials, three types of solder mask materials were used in the study. Material properties of the underfill materials are shown in Table 1, while those of the solder mask materials are shown in Table 2. After the material properties had been determined, the flow properties of the underfill materials were investigated using both flow module test vehicles and actual flip chip assemblies. The latter used an FR-4 substrate and held six 10 mm X 10 mm flip chips, peripherally bumped with 300[micro] pitches. Each flip chip die included a daisy chain Connected in series, one after the other. Transmitted signals go to the first device, then to the second and so on.  A SCSI Daisy Chain Both internal and external SCSI devices are daisy chained together. pattern for rapid resistance measurements. The solder solder (sŏd`ər), metal alloy used in the molten state as a metallic binder. The type of solder to be used is determined by the metals to be united. Soft solders are commonly composed of lead and tin and have low melting points. Hard solders (i. used was eutectic tin-lead (Sn-Pb) on an electroless nickel (Ni) underbump metallization Met`al`li`za´tion n. 1. The act or process of metallizing. . Bump diameter before assembly was about 100 microns and, after assembly, about 80 microns. The final gap varied with the thickness of the solder mask. A state-of-the-art no-clean flux was used, and reflow (1) The process of heating and melting the solder that has been screen printed onto a printed circuit board in order to bond chips and other components to the board. Surface mount chips (SMT) use the reflow method. Contrast with wave soldering. See also reflowable text. was carried out in an [N.sup.2] atmosphere with a maximum temperature of 230[degrees]C. The flow module test vehicles consisted of a glass slide adhesively fastened to a substrate. At one end, spacer particles between the glass and the substrate determined the height of the gap. The test vehicles were heated to the desired process temperature (typically 70[degrees]C to 90[degrees]C), and an underfill material was dispensed dis·pense v. dis·pensed, dis·pens·ing, dis·pens·es v.tr. 1. To deal out in parts or portions; distribute. See Synonyms at distribute. 2. To prepare and give out (medicines). 3. at the end of the test vehicle. Video equipment recorded underfill flow speed. Post-cure, an acoustic micro imaging system inspected the optically opaque underfill nondestructively for voids and agglomerations of filler particles. Where acoustic imaging showed a void or an agglomeration ag·glom·er·a·tion n. 1. The act or process of gathering into a mass. 2. A confused or jumbled mass: , the test vehicle was cross-sectioned for optical or scanning electron microscope scan·ning electron microscope n. Abbr. SEM An electron microscope that forms a three-dimensional image on a cathode-ray tube by moving a beam of focused electrons across an object and reading both the electrons scattered by the object and (SEM) imaging. Analysis of the Five Underfill Materials Differential scanning calorimetry Differential scanning calorimetry or DSC is a thermoanalytical technique in which the difference in the amount of heat required to increase the temperature of a sample and reference are measured as a function of temperature. , dynamic-mechanical analysis, thermo-mechanical analysis, thermo-gravimetrical analysis, rheology, tensiometry and atomic force microscopy microscopy /mi·cros·co·py/ (mi-kros´kah-pe) examination under or observation by means of the microscope. mi·cros·co·py n. 1. The study of microscopes. 2. were used to perform a thorough analysis of the five underfill materials. Material A had a mid-range viscosity (0.59 Pa*s at 70[degrees] C) and a mid-range surface tension (33 mN/m). The filler content was fairly high (70% by weight), and the 12 micron micron: see micrometer. One micrometer, which is one millionth of a meter or approximately 1/25,000 of an inch. The tiny elements that make up a transistor on a chip are measured in micrometers and nanometers. See process technology. particles were fairly large. Thermo-mechanically, this material was rather stiff (Young's modulus 10 GPa). It had a low CTE (26 ppm/[degrees]K) that was close to the CTE of the solder. Material B was similar in its viscosity (0.61 Pa*s at 70[degrees] C) and surface tension (39 mN/m) to Material A. Filler content was 62% by weight, and the particle size Particle size, also called grain size, refers to the diameter of individual grains of sediment, or the lithified particles in clastic rocks. The term may also be applied to other granular materials. of 15 microns was the largest of the five materials. The Young's modulus was 7.6 GPa, and the CTE was 28 ppm/[degrees]K. Material C had a low viscosity (0.19 Pa*s at 70[degrees] C), and its surface tension of 40 mN/m was close to that of Material B. Filler content was a low 40% of weight, and particle diameter was 10 microns. Thermo-mechanically, Material C was rather soft, having a Young's modulus of 2.8 GPa. The CTE was quite high at 40 ppm/[degrees]K. Material C had a humidity uptake (1.77 wt%) at least double that of the four other materials. Note that humidity uptake may or may not be related to filler content; Material E had a similar filler content but a humidity uptake of only 0.83 wt%. Material D had the highest viscosity (0.73 Pa*s at 70[degrees] C) and the lowest surface tension (32 mN/m). It also had the highest filler content (70 wt%) and fairly small particles (8 microns). It had a Young's modulus of 10 GPa like Material A. Because of the high filler content, it had a low CTE of 23 ppm/[degrees]K, which closely matched the CTE of eutectic SnPb solder. Material E had the lowest viscosity (0.04 Pa*s at 70[degrees] C) and the highest surface tension (44 mN/m). It also had the lowest filler content (30 wt%) and the smallest particles (5 microns). Relatively soft, it had a Young's modulus of 3.5 GPa. The CTE, like that of Material C, was 40 ppm/[degrees]K. Material E also had the highest coefficient of planar A technique developed by Fairchild Instruments that creates transistor sublayers by forcing chemicals under pressure into exposed areas. Planar superseded the mesa process and was a major step toward creating the chip. penetration (CPP cpp - C preprocessor. )--244 mm/s; the next highest CPP was 48 mm/s for Material C. The very high CPP was primarily a function of the very low viscosity of this material. Solder Mask Materials Three solder mask (SM) materials were used in the study. Atomic force microscopy showed little difference in surface roughness among the three materials. Surface energy of the solder mask materials was measured by the sessile sessile /ses·sile/ (ses´il) attached by a broad base, as opposed to being pedunculated or stalked. ses·sile adj. Permanently attached or fixed; not free-moving. drop method before and after reflow. In the initial state, strong variations in surface energy occurred, but these variations were greatly diminished after reflow. The geometry of the solder mask, then, appeared to be more important in influencing underfill flow. SM 1 was 2.5 to 3 times thicker than the other two materials; as a consequence, the underfill gap was about 65 microns for SM 2, about 60 microns for SM 3 and only 30 microns for SM 1. Flow Testing Empirical testing of flow behavior was carried out with the flow module test vehicles and the actual flip chip assemblies described earlier. A video camera recorded the flow process in the test vehicle; post-cure, both the test vehicles and the flip chip assemblies were inspected by acoustic micro imaging (Figure 1), followed by physical sectioning where significant defects were present. Material A, which had a CPP of 16 mm/s, flowed at a slow rate and was susceptible to particle clumping clumping /clump·ing/ (klump´ing) the aggregation of particles, such as bacteria, into irregular masses. clump·ing n. The massing together of bacteria or other cells suspended in a fluid. . In the flow modules, clumping of particles occurred after only 1.5 mm of travel; in the flip chip module, clumping occurred after 5 mm. Used with the relatively thick SM 1, Material A gave a gap of only 30 microns, and voids formed near the bumps. Cross sectioning showed no settling of the particles; particle distribution was uniform vertically across the gap. Overall, the processability of Material A was judged to be only poor to fair. The behavior of Material B was like that of Material A, primarily because the CPP (18 mm/s) was similar. Clumping of particles occurred after 1.5 mm of flow in the flow modules and after 5 mm in the flip chip module, although clumping was less pronounced than for Material A. Some voids occurred in the flow modules, but only after 30 mm of travel. Voids were also observed in the flip chip modules, but only around the bumps when SM 1 was used. Vertical distribution of the particles was uniform. Material B was judged to have fair processability. [FIGURE 1 OMITTED] [FIGURE 2 OMITTED] [FIGURE 3 OMITTED] Material C, having a much higher CPP of 48 mm/s, flowed rapidly for up to 10 mm in the flow modules; at this point the flow front became ragged rag·ged adj. 1. Tattered, frayed, or torn: ragged clothes. 2. Dressed in tattered or threadbare clothes: a ragged scarecrow. 3. . After 18 mm of travel, minor clumping of particles occurred but no air entrapment. In the flip chip modules, no particle clumping took place, and voids occurred only adjacent to the bumps in SM 1 modules. Cross sectioning showed even distribution of particles vertically. Overall, Material C was the second best material tested. With a CPP of 12 mm/s, Material D flowed slowly and unevenly. Particle clumping on a scale similar to Material B began after only 1.5 mm of travel in the flow modules, and air entrapment occurred after 20 mm. In the flip chip modules, particle clumping began after about 3 mm of travel, along with minor air entrapment. Voids also formed near the bumps in SM 1 modules. Particle distribution was uniform across the gap. Overall processability was poor. Material E had the lowest viscosity of the five materials and a CPP of 242 mm/sec. Clumping of particles occurred after 20 mm of travel, but no air was trapped by the flow front. In the flip chip modules no particle clumping and only very minor air entrapment near the bumps in the SM 1 modules occurred. Cross sectioning showed that particle distribution across the gap was uniform. This material had the highest overall processability. Reliability Testing The assemblies were tested for popcorn cracking according to according to prep. 1. As stated or indicated by; on the authority of: according to historians. 2. In keeping with: according to instructions. 3. JEDEC The division of the Electronic Industries Alliance (EIA) that deals with semiconductor standards (officially, the JEDEC Solid State Technology Association of EIA). JEDEC was formed in 1958 when the Joint Electron Tube Engineering Council (JETEC) split into two Joint Electron Device Std. 20, Levels 1 and 3. This testing typically caused delaminations (Figure 2), all of which resulted in electrical failures electrical failure n. Failure in which the cardiac inadequacy is secondary to disturbance of the electrical impulse. when the bump material moved into the gap created by the delamination delamination /de·lam·i·na·tion/ (de-lam?i-na´shun) separation into layers, as of the blastoderm. de·lam·i·na·tion n. 1. A splitting or separation into layers. 2. (Figure 3). Pressure cooker testing was carried out at 121[degrees] C and 100% relative humidity relative humidity n. The ratio of the amount of water vapor in the air at a specific temperature to the maximum amount that the air could hold at that temperature, expressed as a percentage. (RH). The typical failure was a delamination between the chip's passivation passivation the final stage in instrument manufacture, passing the finished instruments through a bath of nitric acid which removes foreign particles and promotes the formation of a protective coating of chromium oxide. layer and the underfill. Not all of these delaminations resulted in immediate electrical failures, probably because humidity testing puts little mechanical load on the assemblies. But delaminated assemblies subjected to thermal cycling tended to fail after only a few cycles (Figures 4, 5). Thermal cycling itself typically caused a delamination between the chip and the underfill, as well as resulting cracks in solder bumps and electrical failure (Figure 6). In addition, a few cracks were found at the interface between the substrate and the solder bumps. The cracks extended into the substrate. [FIGURE 4 OMITTED] The five underfill materials behaved as follows in reliability testing: Material A No delaminations were noted after Level 3 testing, but full delaminations and electrical failure occurred after Level 1. Peripheral delaminations were noted after only six hours in the pressure cooker; full delaminations and partial electrical failure occurred after 96 hours. In thermal cycling, delaminations occurred from SM 1 and SM 3 after 250 cycles and after 1,000 cycles from SM 2. Electrical failures occurred only after 1,500 cycles for SM 3. Material B No delaminations were noted after Level 3 testing, but full delaminations and electrical failure occurred after Level 1. Peripheral delaminations were noted after only six hours in the pressure cooker; full delaminations occurred with all three solder mask types; electrical failure happened after 24 hours for SM 3 and after 96 hours for SM 2. In thermal cycling, delaminations occurred from SM 1 after 500 cycles and after 1750 cycles from SM 2 and SM 3. Electrical failures happened only after 2000 cycles for SM 2. Material C No delaminations were noted after Level 3 testing, but full delaminations and electrical failure occurred after Level 1. Delaminations at the die corners were noted after only 24 hours in the pressure cooker; full delaminations and electrical failure happened with all three solder mask types after 96 hours. In thermal cycling, peripheral delaminations from SM 3 were noted after 250 cycles and after 1,000 cycles from SM 1 and SM 2. Electrical failures happened only after 1,500 cycles for SM 1. [FIGURE 5 OMITTED] [FIGURE 6 OMITTED] Material D No delaminations were noted after Level 3 testing, but full delaminations and electrical failure occurred after Level 1. Peripheral delaminations were noted after only six hours in the pressure cooker; strong delaminations and partial electrical failure happened with all three solder mask types after 96 hours. In thermal cycling, delaminations from SM 1 happened after 500 cycles and after 1,000 cycles from SM 2; no delaminations occurred from SM3. Electrical failures happened only after 1,500 cycles for SM 1 and SM 2. Material E No delaminations were noted after Level 3 testing, but strong to full delaminations and partial electrical failure occurred after Level 1. Slight delaminations were noted after six hours in the pressure cooker. No electrical failures happened after pressure cooker storage. In thermal cycling, first delaminations were noted from SM 1 after 500 cycles; no delaminations occurred from SM 2 and SM 3. Conclusions A test matrix of five capillary capillary (kăp`əlĕr'ē), microscopic blood vessel, smallest unit of the circulatory system. Capillaries form a network of tiny tubes throughout the body, connecting arterioles (smallest arteries) and venules (smallest veins). flow underfill materials and of three solder mask materials has been analyzed using analytical and empirical tools. The study showed that careful material selection for substrate, solder mask and underfiller can yield assemblies with reliabilities of at least 2,000 thermal cycles at -55[degrees]C/125[degrees]C--which is more than some commonly used surface-mount components can survive. Flip chip technology is well suited to current production and performance demands. On the horizon are even greater challenges, such as high temperature applications (at 150[degrees] and even 200[degrees] C) and underchip gaps in some radio frequency (RF) and sensor applications as small as 10 microns.
Underfiller
Unit A B
Liquid Properties
Viscosity @ 25[degrees]C [Pa*s] 12 10
Viscosity @ 70[degrees]C [Pa*s] 0.59 0.61
Viscosity @ 80[degrees]C [Pa*s] 0.39 0.51
Density [rho] [g/[cm.sup.3]] 1.8 1.7
Filler Content [wt%] 68 62
Max. Particle Size [[micro]m] 12 15
Surface Tension [gamma]
@ 25[degrees] C [mN/m] 33 39
Contact Angle [theta] @
25[degrees] C on Glass [[degrees]] 33 35
Cos [theta] 0.84 0.82
CPP [phi] @ 70[degrees]C
on Glass [mm/s] 16 18
Processing
Shelf Life @
-40[degrees] C [months] 9 9
Pot Life @ 25[degrees]C [days] 2 1/3
6@
Gelation Time [min] n.a. 121[degrees]C
Processing Temperature [[degrees]C] 70-100 70-90
30 min @ 10 min @
Recommended Cure Profile 165[degrees]C 160[degrees]C
Cured Properties
Thermal Conductivity [W/(mK)] 0.45 0.41
Glass Transition
Temperature DSC [[degrees]C] 140 n.a.
TMA [[degrees]C] 158 144
Coefficient of Thermal
Expansion a1 [[10.sup.-6]/K] 26 28
a2 [[10.sup.-6]/K] 87 104
Flexural Modulus [GPa] 10.3 7.6
Tensile Modulus [GPa] 5.5 3.2
Linear Shrinkage [%] 0.14 n.a.
Water Uptake [wt%] 0.47 0.70
Extractable Ionic
Content Cl- [ppm] 5 20
Na+ [ppm] 5 5
K+ [ppm] n.a. 5
Underfiller
C D E
Liquid Properties
Viscosity @ 25[degrees]C 8 20 1
Viscosity @ 70[degrees]C 0.19 0.73 0.04
Viscosity @ 80[degrees]C 0.17 0.47 0.09
Density [rho] 1.52 1.8 1.4
Filler Content 40 70 30
Max. Particle Size 10 8 5
Surface Tension [gamma]
@ 25[degrees]C 40 32 43
Contact Angle [theta]@
25[degrees] C on Glass 46 39 44
Cos [theta] 0.69 0.78 0.72
CPP [phi] @ 70[degrees]C
on Glass 48 12 242
Processing
Shelf Life @
-40[degrees] C 6 5 6
Pot Life @ 25[degrees]C >1/3 3 n.a.
2.7@
Gelation Time n.a. n.a. 150[degrees]C
Processing Temperature 70-100 70-100 70-80
5 min @ 15 min @ 1 h @
Recommended Cure Profile 150[degrees]C 150[degrees]C 150[degrees]C
Cured Properties
Thermal Conductivity 0.32 0.48 0.31
Glass Transition
Temperature DSC 135 n.a. 155
TMA 101 113 160
Coefficient of Thermal
Expansion a1 40 23 40
a2 133 84 147
Flexural Modulus 2.8 10 3.5
Tensile Modulus 2.8 n.a. n.a.
Linear Shrinkage 3 n.a. n.a.
Water Uptake 1.77 0.59 0.83
Extractable Ionic
Content Cl- 30 n.a. 15
Na+ 1 n.a. 2
K+ 1 n.a. 1
[??] Material Property determined at Fraunhofer IZM. [??] Calculated
Material Property.
TABLE 1: Material properties of flip chip underfillers studied.
Property Unit SM-1 SM-2 SM-3
Solder Mask
Thickness [[micro]m] 49 15 22
Solder Mask
Opening Diameter [[micro]m] 124 109 138
Solder Mask
Roughness [[micro]m] 0.3 0.2 0.3
Surface Energy
Initial State [mN/m] 31 38 29
Surface Energy
After Reflow [mN/m] 23 28 26
TABLE 2: Properties of solder mask materials.
Karl-Friedrich Becker is with the Fraunhofer Institute, Berlin, Germany; and Tom Adams “Tom Adams” redirects here. For other people known as Tom Adams, see Tom Adams (disambiguation). Tom Adams (born 1926) is an illustrator most famous for his Agatha Christie paperback cover designs. is a consultant with Sonoscan, Inc., Elk Grove Village Elk Grove Village, village (1990 pop. 33,429), Cook and Du Page counties, NE Ill., a suburb of Chicago; inc. 1956. With a population of c.100 at the time of its establishment on open farmland, the village has grown dramatically and steadily, largely because of its , IL; email: info@sonoscan.com. |
|
||||||||||||||||
Printer friendly
Cite/link
Email
Feedback
Reader Opinion