In 2006 we conducted two formal detailed experiments on the 01005-component assembly process. Two of the most significant findings from those experiments were that a 0.003” thick stencil is required to achieve an acceptable solder paste release percentage and that nitrogen is required for complete reflow of the very small volume of solder paste. For the most part,several other researchers had concluded the same for 01005 component assembly experiments.
Both the potential requirement to use nitrogen for the reflow soldering process and the use of a 0.003” thick stencil will be a major issue for high volume manufacturers. Introduction of Nitrogen will increase the cost and the use of a 0.003” stencil may cause insufficient solder paste for larger components on the same board.
The focus of this study is to discover methods to avoid the use of nitrogen in the reflow soldering process and to allow the use of a 0.004" thick stencil in the solder paste printing process. The questions that will be answered are:
- Can any available stencil technologies and solder paste product/type provide sufficient solder paste release to allow the use of a 0.004" thick stencil for 01005 components?
- Can we develop reflow profiles that will completely reflow 01005 components in an air atmosphere and/or are the solder paste suppliers developing new products that will allow the extremely small volumes of solder paste printed for 01005 components to be reflowed in an air atmosphere?
- If the use of nitrogen is required what is the optimum oxygen level to get complete reflow?

The mobile communication devices such as cell phones require high speed transmission of large volume data as well as reduction in size and weight. When signal is transmitted in high speed and frequency on PWB,signal integrity becomes a big problem. One of the most widely used methods to tackle the problem is to apply low Dk/Df materials. Even though various materials are available for that purpose,they tend to be quite expensive and require special care during the fabrication process. Therefore,PWB fabricators have been looking for more affordable and easily processable low Dk/Df material. To meet these demands,a novel PWB material has been developed. The special epoxy resin with rigid backbone was used,which does decrease Dk level but not significantly for the application. Therefore,hydrophobic hardener was adopted to compensate the difference in dielectric properties. The resulting material has Dk of 3.8 at 1 GHz,which is 10% lower than that of conventional FR-4. In addition,its Tg is higher than 170 C. It also shows an excellent thermal stability,withstanding longer than 30 minutes at T-288. The further development of material toward better dielectric properties was made possible with application of NE glass instead of conventional E glass. The resulting laminate has Dk of 3.5 as well as a Df of less than 0.01. The other properties of these materials,such as migration property,Td and reliability will be presented.

We demonstrate here a novel CCL (Copper Clad Laminate),which exhibits an extremely low transmission loss at mm-wave band. The CCL,which we developed,is based on a new fluoropolymer with adhesive characteristics. In contrast to conventional PTFE,the adhesive fluoropolymer allows us to apply a wholly dry process for CCL fabrication,which contributes to environmental load reduction.
It is well known that the factor of transmission loss mainly consists of conductive loss and dielectric loss. However,in conventional CCL data sheets,only the loss tangent data at specific frequencies are disclosed. Neither the dielectric loss nor the whole transmission loss at the other frequencies is known. In order to minimize the transmission loss at mm-wave band,the quantitative analysis for those factors is essential.
We proposed the evaluation method,which can clarify not only the dielectric loss but also the conductive loss of CCL up to 110GHz. Several transmission lines with different impedance were measured and analyzed; the three different losses were discriminated in straightforward manner. Besides the evaluation method,a highly accurate measurement technique for low loss transmission line was achieved.
Using several kinds of surface roughness of copper foil,we made CCL test samples and evaluated the transmission loss by the above-mentioned method. Since the results indicated that the surface roughness of copper foil remarkably influenced the transmission loss,profile free copper foil was used for developed CCL. Due to the adhesive characteristics of new fluoropolymer,enough peeling strength was obtained without extra surface treatment.
Finally we benchmarked our developed CCL to the Rogers RT/duroid 5880,with the world’s lowest loss characteristics,and the result showed improved loss characteristics compared to RT/duroid 5880.

Over the years,the EU RoHS restriction and lead-free capability is the hottest environmental protection subject. In technology trends,signal integrity performance gets more critical based upon today’s higher signal transmission speed demand in every field of applications such as computer CPU and GPU chipset levels,system operation frequency and a variety of communication bus and cables like PCI express,SATA II and AGP bus for computer systems. Signal communication speed will shift from 1-5 Gbps range up to 5-10Gbps depending on applications. In order to meet lead-free
requirements and severe processes conditions with good signal integrity performance,the laminate material will play a more and more critical and sensitive role in the system.
From consumer products to high-end applications,there is a need for certain electrical and thermal performance; it is essential to meet those requirements with cost effectiveness. As a base material supplier,we will hereby discuss material design and factors that influence signal integrity,including epoxy and hardener,resin chemical construction,laminate ply-up construction,amount of resin content,fabric weaving density,moisture pick-up and environment factors etc. for a massive mainstream application and low loss application under the hypothesis of lead-free capability.

For companies that choose to take the Pb-free exemption under the European Union’s RoHS Directive and continue to manufacture tin-lead (Sn-Pb) electronic products,there is a growing concern about the lack of Sn-Pb ball grid array (BGA) components. Many companies are compelled to use the Pb-free Sn-Ag-Cu (SAC) BGA components in a Sn-Pb process,for which the assembly process and solder joint reliability have not yet been fully characterized.
A careful experimental investigation was undertaken to evaluate the reliability of solder joints of SAC BGA components formed using Sn-Pb solder paste. This evaluation specifically looked at the impact of package size,solder ball volume,printed circuit board (PCB) surface finish,time above liquidus and peak temperature on reliability. Four different BGA package sizes (ranging from 8 to 45 mm2) were selected with ball-to-ball pitch size ranging from 0.5mm to 1.27mm. Two different PCB finishes were used: electroless nickel immersion gold (ENIG) and organic solderability preservative (OSP) on copper. Four different profiles were developed with the maximum peak temperatures of 210oC and 215oC and time above liquidus ranging from 60 to 120 seconds using Sn-Pb paste. One profile was generated for a lead-free control. A total of 60 boards were assembled. Some of the boards were subjected to an as-assembled analysis while others were subjected to an accelerated thermal cycling (ATC) test in the temperature range of -40oC to 125oC for a maximum of 3500 cycles in accordance with IPC 9701A standard. Weibull plots were created and failure analysis performed.
Analysis of as-assembled solder joints revealed that for a time above liquidus of 120 seconds and below,the degree of mixing between the BGA SAC ball alloy and the Sn-Pb solder paste was less than 100 percent for packages with a ball pitch of 0.8mm or greater. Depending on package size,the peak reflow temperature was observed to have a significant impact on the solder joint microstructural homogeneity. The influence of reflow process parameters on solder joint reliability was clearly manifested in the Weibull plots. This paper provides a discussion of the impact of various profiles’ characteristics on the extent of mixing between SAC and Sn-Pb solder alloys and the associated thermal cyclic fatigue performance.

Since the European 2002/95/EC RoHS directive enforcement on 1st July 2006,a dominating part of the electronic industry suppressed the use of Pb in electronic equipment. As one of the consequences,exempted industries like avionics,military and telecom servers are facing increasing SnPb component obsolescence issues. Before transitioning to the qualification and deployment of full lead-free for AHP electronic products,reliable backward solutions are urgently needed to overcome SnPb package procurement difficulties. The main concerns lie in Pb-free BGAs and SnBi leaded packages assembled with a conventional SnPb reflow soldering process which affects the solder joint microstructure and can compromise the 2nd-level reliability.
This paper deals with an extensive cooperation program between an EMS and an equipment maker engaged to evaluate the use of a specific “SnPb+” assembly process as an option to solve the backward compatibility issues inherent to these kinds of components. The SnPb+ process window as well as the incidence of reflow conditions on BGA mixed interconnects have been studied for a variety of generic components ranging from large plastic- and ceramic-BGAs down to 0.5mm pitch CSPs. The reliability has been assessed with statistics in comparison to full SnPb both under thermo-mechanical and mechanical stresses (vibrations,shocks) focusing on mission critical high-end applications in harsh environments. Numerous failure analyses have been conducted to evidence the failure modes and support the obtained reliability data.
This paper presents the –55°C/+125°C ATC reliability results of mixed BGAs assembled under various reflow conditions on dedicated test vehicles reflecting typical functional boards with the PCB finish as a variable (ENIG and Immersion Sn).
As a conclusion,results indicate that SnPb+ is a very promising industrial solution for complex assemblies with long term mission in severe environments. Completion of the overall program including mechanical testing will permit to fully validate the use of the SnPb+ process to secure the backward transition phase.

Immersion silver chemistry has been promoted as a final finish for solderability for several years now. There are different commercially available products that will deposit silver in a wide range of thicknesses. Some chemistries,because of their very aggressive nature can produce a thickness distribution within a circuit board of 20-30 millionths and higher on small pads versus large ground planes. A large difference in silver thickness will affect solder wetting times within the same board especially with SAC alloys. A large variation in silver thickness can be overcome in assembly. But it typically requires a more precise assembly process with a longer and or hotter soldering profile.
A study has been underway evaluating nitrate based immersion silver chemistry against other potential silver sources. The study evaluated several types of additives against PCB design. Data on thickness distribution,copper attack rate,and deposit porosity has resulted in new theories on how to control immersion silver deposits. Specific formulas have been found that provide a pore free deposit with minimal attack on electroless copper. A specific test vehicle was chosen that takes into account ground planes,isolated traces,interconnected pads and different metallization processes. Data on thickness distribution,copper attack rate,and deposit porosity has resulted in new theories on how to control immersion silver deposits. Specific formulas have been found that provide a pore free deposit with minimal attack on electroless copper.

Concerns
There are some areas that must be accepted if planning to run the newest SENIG process:
- Slower plating rate in nickel (to get 150µ“ it will take 20 minutes)
- It works better at higher temperatures (190°F is best operating temp)
- It has higher phosphorus in the deposit (as high as 11.7%) and some evidence suggests solderability may be reduced at such levels; we have an extensive solderability study on-going.
Advantages
It is possible to have a high phosphorus nickel bath last to 5 MTO’s while running SENIG. The obvious advantage of longer bath life is also bolstered by other benefits:
- No dummy panels
- No minimum loading
- No corrosion from stripper or OSP cycles
- Pass SO2 testing

It is known that pure tin will undergo an allotropic transformation below 13°C where it becomes a semiconductor with a 26% [1,2] volume increase,and in appearance turns from a bright shiny metallic material,white ß tin,to a dark blue/grey dust,a tin. Such a transformation for an electrical interconnect is disastrous,if it were to occur in any of the high tin lead-free alloys it would be a catastrophe. The elimination of lead,one of the best elements to arrest the transformation process,has resulted in a number of high tin content alloys about which the potential to transform is unknown. Environmental factors that may enhance or arrest the rate and the incubation period of the transformation processes are also unknown. Due to the optimal transformation temperature of approximately –35°C and the long time required for the transformation,a direct observation of the phenomenon has not been possible. This study proposes a new method for observing the ß/a transformation in situ using a time-lapse photographic technique. This study concentrates on pure tin,but the applicability of the method opens new possibilities for studying the phenomenon for other tin alloys,such as the two commonly encountered eutectics of SnCu and SnAgCu. The transformation progressed radially from the inoculation point,starting at the surface. Propagation into the bulk occurred by peeling; with the external layers tending to “roll out” due to the volume expansion of the internal layers. In the meantime,cracks parallel to the propagation direction formed. Typically a pure tin sample would completely transform in just over 24 hours.

This paper reports the results of an evaluation of tin mitigation processes for components,i.e. converting parts with pure tin (Sn) and other lead-free (Pb-free) finishes to tin-lead (SnPb) finishes.