High Performance EMI Shielding Materials and Spraying Process Parameters for High Frequency FCBGA Application

Conventional electromagnetic interference (EMI) shielding solution being used to protect the devices has been shield-can type structures. However, in order to be in accordance with ultra-thin thickness of product newly EMI shielding methods have been intensively on the rise in these days. In order to meet these demands, recently, many researchers have been actively studied on electromagnetic shielding methods such as spray, electroplating, and sputtering methods. Among these, the spray methods are considered to have relative advantages in terms of the byproducts generated during the process, the cost for the initial investment, and the performance and capability of the shielding layer.

Electronic equipment generates undesirable electromagnetic energy that can interfere with the operation of proximately located electronic equipment due to EMI transmission by radiation and conduction. The electromagnetic energy can exist over a wide range of wavelengths and frequencies. In order to minimize problems associated with EMI, sources of undesirable electromagnetic energy are needed to be shielded and electrically grounded to reduce emissions into the surrounding environment. Shielding is designed to prevent both ingress and egress of electromagnetic energy relative to a barrier, a housing, or other enclosure in which the electronic equipment is disposed.

The key for EMI shielding is development of highly conductive materials and mature process technology. Although various coating technique such as Air spray, electro plating, Ultrasonic Spray, and sputtering has been intensively reported to define conductive layers, we employed air-spray method (Protec Co., LTD, SPR-100) which can achieve high UPH with low cost, high aspect ratio of top surface coating thickness to side wall coating thickness, and micron-scale thickness tolerance.

We demonstrated the package-level spraying solution with Ag epoxy composites (Ntrium, NSA-F3-R4) [1]. In the report, various types of Ag powder were tested to obtain optimized electrical resistivity and mechanical reliability. Additionally, we also studied the effect of spray machine's parameter such as pressure, speed, and droplet size on uniformity of sprayed conductive film.

In this study, we will show the investigation on relation between electrical resistivity and series-resistance of daisy chain of FCBGA (flip chip ball grid array) package. The change of the electrical properties will be demonstrated as a function of Ag contents change. Lastly, the brief result of AgCu epoxy composites will be introduced.

Ag paste was carefully formulated by combining mainly the components: resin, additives, Ag particles. Multifunctional resin is mixed to achieve appropriate properties of sprayed film. The leveling additives are the portion that control leafing speed of fillers sprayed on target substrate, which affects to the electrical resistivity and side coverage. In the filler part, spherical particle, flake-shaped particle, and mixture of spherical and flake-shaped Ag particles were blended and tested with other parts. All the contents of pastes are very similar to those of commercial paste which is based on epoxy resin matrix but different average particle shapes and sizes. The results containing spherical type only and mixture of spherical and flake-shaped Ag particles were not good as can be seen in Fig. 1. Each sample showed 7.9×10⁻⁶Ω·m and 8.5×10⁻⁷Ω·m, respectively. However, the result of flake-shaped Ag particle showed 2.9×10⁻⁷Ω·m.

From the electrical resistivity table in Fig. 2, we assumed as follows. As the content ratio of spherical particles increases, the probability of each Ag particles making point-contact increases as shown in Fig. 2(a). However, the higher the ratio of the content of the flake-shaped particles, the higher the probability that the Ag particles will come into surface-contact as shown in Fig. 2(b). Therefore, the electrical resistivity of a sample using only spherical particles is 10 times higher than that of flake-shaped particles alone.

In order to define conductive layer made by Ag paste, in which flake-type Ag powders were used, we employed air-spray method. (Protec Co., LTD, SPR-100) which can achieve high UPH with low cost, high aspect ratio of top surface coating thickness to side wall coating thickness, micron-scale thickness tolerance, and excellent convenient facilities for mass-production.

Spray EMI shielding technique is a relative newcomer to the industry when compared to the former processes such as electro plating and sputtering. Development has been underway using conductive materials for chip to substrate assembly, because the cost/benefit equation, enhanced workability and performance for package to be shielded from EMI have met the needs of the industry.

Fig. 3 shows the schematic illustration of our process for spraying which consists of simple 4 steps such as material mixing with diluent, sample loading to the tape, spraying for several min., curing at 190°C for 30 mins., and lastly sample unloading. This curing condition was obtained from the test as shown in Fig. 5.

A micro-ohm meter (HIOKI, RM3548) shown in Fig. 6 was used to measure the electrical properties such as the resistivity and resistance. All the samples were sprayed with the spray equipment and cured under atmosphere at 190°C for 30 minutes. The electrical properties ware measured in the shortest direction. The thickness of sprayed conductive film was measured with FE-SEM. Then, the electrical resistivity was calculated by measured resistance from micro-ohm meter and measured dimension from FESEM. By utilizing the result of resistance values measured by micro-ohm meter and thickness information measured by FESEM, we can calculate the electrical resistivity of sprayed shielding layer using ohm's law. (refer to Fig. 7)

Now, the series-resistance of FCBGA daisy chain will be abbreviated to “series-resistance”. Before the shielding materials were not sprayed on the FCBGA package, we cannot measure the series-resistance, because daisy chain was not connected to each one. Where, the daisy chain is the contact pad for the electrical ground. Therefore, after spaying the shielding material we can measure the series-resistance.

The micro-ohm meter was utilized to measure the series-resistance. For the measurement, as shown in the Fig. 8, one tip and the other tip of the micro-ohm meter were connected to the shielding layer and contact pad for electrical ground, respectively. This measured value is the series-resistance of FCBGA package. The important thing is that this series-resistance directly corresponds to the shield effectiveness (SE) level.

We investigated the tendency of the electrical resistivity and the series-resistance in compliance with curing condition change such as temperature and times. For each condition 8 pcs of FCBGA samples were prepared to perform the standard spray process as shown in Fig. 3.

After spraying process, each 8 pcs samples were cured at each curing condition. The series-resistance and the electrical resistivity were measured by using micro-ohm meter in compliance with the standard measurement process as shown in Fig. 7 and 8, respectively.

The electrical properties of samples cured at less than 130°C for 10, 20, 30min. showed very large dispersion, which means that the shielding material has not been cured enough. On the other hand, the samples cured at more than 190°C showed small dispersion. The degree of dispersion of all the samples is almost similar.

It was found that the series-resistance and the electrical resistivity values were decreased as the curing temperature increased. The curing cannot affect the electrical properties. For our Ag epoxy composite, we found that 190°C for 30min is the most reasonable point considering the electrical property and UPH for mass production.

As shown in Fig. 9 and 10, we have studied the effect of Ag filler content in the EMI shield layer of FCBGA to find out the optimum filler contents by measuring the resistivity of shielding layer and the series-resistance. The resistivity was decreased to 3.46 ×10⁻⁷Ω·m and 21.7 mΩ of the series-resistance could be obtained after 190°C curing condition with the 90~92% Ag filler contents. However, when the Ag contents in the solid components was below 90% or above 92%, the resistivity was significantly increased to 4.7×10⁻⁷Ω·m, which results in increase of series-resistance of daisy chain. As a result, we found optimum Ag contents to obtain much lower material resistivity was 90%~92% range.

Additionally, we have studied the mechanism for the increased resistivity in the region of 88% and 94%, respectively. we studied why the electrical properties were deteriorated in the region of and 88% and 94% of Ag contents. In order to describe this, we established a hypothesis to describe this phenomenon as shown in Fig. 11(a), (b), and (c). The cross-sectional analysis of those samples was conducted. As a result, the sample having 88% Ag contents had more gaps among Ag fillers than 90~ 92% of Ag contents as shown in Fig. 11(a), (b), and (c). Contrastively, 94% sample showed too much Ag fillers were there, which resulted in lack of binder. In this case, the surface of the Ag filler was not fully coated with the binder.

We have studied the dependency of the shielding materials resistivity and series-resistance of daisy chain in compliance with coating thickness change. For this thickness dependency study, we have coated the shielding layers with the thickness of 4um, 5um, 8um, 10um, 12um, and 16um by spray coating process as shown in Fig. 1 and shielding layers were defined on the surface of FCBGA for EMI shielding.

Cross-sectional analysis of conductive shielding film with FESEM was carried out to measure the thickness of shielding layer. We have found that the electrical resistivity of shielding materials was not enhanced beyond a specific thickness as shown Fig. 12, which indicated that there was no change of the material resistivity with the change of the coating layer thickness. Additionally, we found that at least 5um thickness of shielding materials were necessary to obtain the optimized value of series-resistance. For the last thing, we found that the tendency of series-resistance was the same to that of electrical resistivity as shown in Fig. 13.

For the 85 °C 85RH reliability test, we prepared 32 chips shown in Fig. 14 using spray procedure as shown in Fig. 3. All of the 32 chips were cured at 190 ° C for 30 minutes after spray to the EMC surface. Then, the resistance of each chip was measured before the test and recorded as shown in Fig. 15. Thereafter, the 85 °C85RH test was carried out for 196 hours.

The resistance change of the 32 chips was measured by micro-ohm meter. As a result, we found that the measured resistances of all chips have not been changed during the 85 °C 85RH test.

After performing 85 °C85RH test, color and resistance change were checked. Thereafter, according to ASTM D3359, cross-cut test was performed with 3M 610 tape. The lattice spacing was 2 mm × 2 mm. The result of cross-cut is shown in Fig. 16. It is possible to confirm the adhesion of 5B level on both the tape surface and the chip surface.

We prepared a general memory chip to evaluate the readability of laser-marking on the top surface of EMC as shown in Fig. 17 (a). Additional surface cleaning was not carried out. 10um-thick shielding layer was defined with spray method. Then, the coated package was cured at 190 °C for 30 minutes.

In order to measure shield effectiveness (SE), conductive thin films of 5um, 10um, and 15um thickness were prepared by utilizing obtained spray and material conditions. According to ASTM 4935, SE measurement was carried out. Agilent E5062A/EM2107A was used to measure SE. The measured frequency band is ranges from 30MHz ~ to 1.5GHz. The electromagnetic wave shielding effectiveness is calculated by checking the ratio of received power divided by transmitted power when the specimen is placed in the fixture. The electromagnetic wave shielding effect is calculated by the following equation. (1) SE (dB) = 10log P1 / P2 (2) where SE is the shielding effectiveness, P1 is the power when the specimen is present, and P2 is the power when there is no specimen. As shown in Figure 19, SE of resulting EMI shielding film made of Ag and matrix is as high as 60dB, 70dB, 75dB at 5um, 10um, 20um-thick film, respectively.

In these days, our group has intensively studied on cost-effective AgCu epoxy composite. The fillers of this composite consist of mixture of Ag and AgCu particles for EMI shielding conductive layer. We have tested the mixture ratio of AgCu and Ag filler contents in a wide range of 0%~100%. As a result, in an optimum point, we achieved to 5.50 ×10⁻⁷Ω·m of electrical resistivity and 25± 3mΩ of series-resistance, respectively. We expect that Far-field shielding effectiveness may be measured more than 60dB at 10um conductive shielding layer when measured in the range from 30MHz to 1.5GHz. Moreover, we already obtained spray parameters for uniform coating condition with this material as shown in Fig. 18.

In summary, we have investigated for the relationship between the resistivity of EMI shielding material and series-resistance of daisy chain in FCBGA with curing condition, mixture ratio of Ag filler contents, and shielding layer thickness. As a result, we found that the electrical resistivity of shielding material is the key to achieve high-quality SE level in FCBGA package. In order to achieve the best EMI shielding performance and low cost, the optimized EMI shielding thickness and filler contents for spray coating process were demonstrated. The package-level EMI shielding technology has been already researched in depth and has recently been approved for mass production in some of semiconductor package manufacturers.

At last, we demonstrated cost-effective AgCu epoxy composite as a new material solution of package-level EMI shielding briefly. The electrical resistivity is very similar to that of Ag epoxy composite. Therefore, we can achieve the similar series-resistance of FCBGA package. It is expected that this package-level EMI solution will be applied to various applications to replace the existing technology such as sputtering and plating in the near future such as WiFi module, Bluethooth module, LTE modules, flash memory, power amplifier and etc.