Analysis and testing methods of reliability issues of lead-free solder joints

With the rapid development of the electronic information industry, fine pitch devices have developed, assembly density has become higher and higher, new SMT and MCM technologies have been born, and the solder joints in microelectronic devices have become smaller and smaller, while the mechanical, electrical and thermodynamic loads they bear are becoming heavier, and the reliability requirements are increasing. The SMT packaging technology widely used in electronic packaging and new chip size packaging (CSP), ball grid array (BGA) and other packaging technologies all require direct electrical and rigid mechanical connection between different materials (mainly bearing shear strain) through solder joints, and its quality and reliability determine the quality of electronic products.

　　The failure of a solder joint may cause the failure of the entire device, so how to ensure the quality of the solder joint is an important issue. Traditional lead-tin solder contains lead, and lead and lead compounds are highly toxic substances. Long-term use of lead-containing solder will cause serious harm to human health and living environment.

　　At present, the demand for lead-free soldering in the electronics industry is becoming more and more urgent, which has already had a huge impact on the entire industry. Lead-free solder has begun to gradually replace lead solder, but lead-free technology will inevitably bring new problems to the reliability of solder joints due to the differences in solder and the adjustment of welding process parameters. Therefore, the reliability of lead-free solder joints has also received more and more attention. This article describes the failure mode of solder joints and the factors that affect the reliability of lead-free solder joints, and also introduces the reliability test methods of lead-free solder joints.

　　Failure modes of solder joints

　　The reliability test of solder joints, including reliability test and analysis, aims to evaluate and identify the reliability level of integrated circuit devices and provide parameters for the reliability design of the whole machine on the one hand; on the other hand, it is to improve the reliability of solder joints. This requires necessary analysis of failed products, finding out the failure mode, and analyzing the cause of failure. Its purpose is to correct and improve the design process, structural parameters, welding process, etc. The failure mode of solder joints is very important for the prediction of cycle life and is the basis for establishing its mathematical model. The following introduces three failure modes.

　　1. Solder joint failure caused by welding process

　　Some unfavorable factors in the welding process and the subsequent inappropriate cleaning process may cause the failure of the solder joints. The reliability problem of SMT solder joints mainly comes from the production assembly process and the service process. In the production assembly process, due to the limitations of equipment conditions such as pre-welding preparation, welding process and post-welding inspection, as well as human errors in the selection of welding specifications, welding failures such as cold soldering, solder short circuits and Manhattan phenomenon are often caused.

　　On the other hand, in the process of use, the inevitable impact, vibration, etc. will also cause mechanical damage to the solder joints. For example, the rapid hot and cold changes during the wave soldering process will cause a temporary temperature difference on the components, causing the components to be subjected to thermal-mechanical stress. When the temperature difference is too large, stress cracks will occur in the ceramic and glass parts of the components. Stress cracks are an unfavorable factor that affects the long-term reliability of solder joints.

　　At the same time, in the assembly process of thick and thin film hybrid circuits (including chip capacitors), gold and silver corrosion often occurs. This is because the tin in the solder forms compounds with the gold and silver in the gold-plated or silver-plated pins, which reduces the reliability of the solder joints. Excessive ultrasonic cleaning may also affect the reliability of the solder joints.

　　2 Failure caused by aging

　　When molten solder contacts a clean substrate, intermetallic compounds (IMCs) are formed at the interface. During the aging process, the microstructure of the solder joint will coarsen and the IMC at the interface will continue to grow. The failure of the solder joint depends in part on the growth kinetics of the IMC layer. Although the intermetallic compounds at the interface are a sign of good welding, as their thickness increases during service, they will cause microcracks initiation and even fracture in the solder joint.

　　当其厚度超过某一临界值时，金属间化合物会表现出脆性，而由于组成焊点的多种材料间的热膨胀失配，使焊点在服役过程中会经历周期性的应变，形变量足够大时会导致失效。研究表明Sn60／Pb40软钎料合金中加入微量稀土元素镧，会减少金属化合物的厚度，进而使焊点的热疲劳寿命提高2倍，显著改善表面组装焊点的可靠性。

　　3 Failure caused by thermal cycling

　　When electronic devices are in service, the periodic on-off of the circuit and the periodic changes in ambient temperature will cause the solder joints to undergo a temperature cycle process. The thermal expansion mismatch between the packaging materials will produce stress and strain in the solder joints. For example, in SMT, the coefficient of thermal expansion (CTE) of the chip carrier material A1203 ceramic is 6×10-6℃-1, while the CTE of the epoxy resin/glass fiber substrate is 15×10-6℃-1. When the temperature changes, the solder joints will be subjected to certain stress and strain. Generally, the strain borne by the solder joints is 1% to 20%. In the THT process, the flexible pins of the device will absorb most of the strain caused by the thermal mismatch, and the actual strain borne by the solder joints is very small. In SMT, the strain is basically borne by the solder joints, which will lead to the initiation and expansion of cracks in the solder joints and eventually failure.

　　Since solder joints crack and fail due to thermal stress caused by mismatched thermal expansion coefficients, improving the thermal matching between leadless components and substrate materials is the most likely issue that people will pay attention to first. At present, new materials such as 42% Ni-Fe alloy (CTE=5×10-6℃-1), Cu-36% Ni-Fe alloy (indium tile alloy), Cu-Mo-Cu and quartz fiber composite materials have been studied and developed. Among them, the Cu-indium tile-Cu composite substrate changes the proportion of each component. The welded parts with lead soldering on this substrate have been subjected to 1500 thermal shock tests without solder joint failure. In addition, a technology of compounding a stress absorption layer with greater elasticity on the printed circuit board to absorb the stress caused by thermal mismatch has also been developed, and relatively good results have been achieved. However, the process of new substrate materials is complicated and the price is relatively expensive, and its practicality is subject to certain restrictions.

Factors Affecting Lead-Free Solder Joint Reliability

　　1. Performance requirements for lead-free solder

　　Traditional tin-lead solder is used as the standard material for connecting components and printed circuit boards because of its low price, easy welding, beautiful shape, and good physical, mechanical and metallurgical properties. It has formed a complete set of use processes and has long been favored by electronic manufacturers. However, due to the adverse effects of lead and lead compounds on human health and living environment, the call for restricting and prohibiting the use of lead-containing solder is growing. Governments of various countries have formulated corresponding regulations to restrict the use of materials and waste disposal of electronic products. The environmentally friendly requirements of electronic packaging have become a global trend. Therefore, the electronics industry is currently facing the requirements of lead-free, which has formed a huge impact on the entire industry. Lead-free solder has developed rapidly in recent years, and the most commonly used is the Sn-Ag-Cu series.

　　Solders used in the microelectronics field have very strict performance requirements, and lead-free solder is no exception. It not only includes electrical and mechanical properties, but also must have an ideal melting temperature. Considering both manufacturing process and reliability, Table 1 lists some important properties of solder alloys.

　　2 Factors affecting the reliability of lead-free solder joints

　　Compared with the traditional lead-containing process, lead-free soldering will inevitably have a certain impact on the reliability of solder joints due to the difference in solder and the adjustment of process parameters. First, the melting point of lead-free solder is relatively high, generally around 217°C, while the melting point of traditional Sn-Pb eutectic solder is 183°C. The increase in the temperature curve will bring about problems such as easy oxidation of solder and rapid growth of intermetallic compounds. Secondly, since the solder does not contain Pb, the wetting performance of the solder is poor, which easily leads to the self-calibration ability, tensile strength, shear strength, etc. of the product solder joints not meeting the requirements. Taking a certain manufacturer as an example, the unqualified rate of solder joints in the original lead-containing process is generally around 50×10-6 (0.05%) on average, while the unqualified rate of lead-free process rises to 200×10-6~500×10-6 (0.2~0.5%) due to the poor wettability of solder.

　　In view of the fact that there are still many problems in the reliability of lead-free solder joints, it is necessary to analyze this. The reliability problems of lead-free solder joints mainly come from: shear fatigue and creep cracks of solder joints [7, 8, 9], electromigration [8, 10], cracks formed by intermetallic compounds at the interface between solder and substrate [7, 8, 11, 12], short circuit caused by Sn whisker growth [7, 8], electrical corrosion and chemical corrosion problems, etc. In the following, we mainly introduce some factors that affect the reliability of lead-free solder joints from the perspectives of design, materials and processes.

　　(1) Design: Reasonable design issues of PCB. Such as unreasonable pad design, dense distribution of components with high heat generation, "high-rise effect" caused by adjacent tall components during reflow soldering, and hot air shock.

　　(2) Materials: The choice of solder is extremely important. At present, most of them use tin-silver-copper alloy series, and the liquidus temperature is 217℃-221℃, which requires the reflow soldering to have a higher peak temperature. As mentioned above, this will bring about problems such as high-temperature oxidation of solder and conductor materials (such as Cu foil) and rapid growth of intermetallic compounds. Because during the soldering process, when the molten solder contacts the soldering substrate, a layer of intermetallic compound (IMc) will be formed at the interface due to the high temperature. Its formation is not only controlled by the reflow soldering temperature and time, but also its thickness will increase with time during the later use.

　　Studies have shown that intermetallic compounds on the interface are a key factor affecting the reliability of solder joints. The presence of an excessively thick intermetallic compound layer can lead to fracture of the solder joint, reduced toughness and resistance to low-cycle fatigue, and thus reduced reliability of the solder joint. Taking the most mature Sn-Ag lead-free solder as an example, due to its higher melting point, the corresponding reflow temperature will also increase. In addition, the Sn content in lead-free solder is higher than that in Sn-Pb solder. Both of these increase the rate of formation of intermetallic compounds on the interface between the solder joint and the substrate, leading to premature failure of the solder joint.

　　In addition, due to the different components of lead-free solder and traditional Sn-Pb solder, their reaction rates and reaction products with pad materials such as Cu, Ni, AgPd, etc. may be different, and the solder joints will also show different reliability. At the same time, the compatibility of solder and flux will also have a great impact on the reliability of solder joints. Studies have shown that the incompatibility between the components of solder and flux will lead to reduced adhesion. In addition, due to the mismatch of thermal expansion coefficients, the periodic fatigue failure of solder will be accelerated. Therefore, special attention should be paid to selecting solder and flux with excellent compatibility to withstand the high temperature impact during lead-free reflow.

　　In addition, each interconnected soldering component comes from different manufacturers, so the quality of the components is inevitably uneven, such as poor solderability of device pins, which has a great impact on the reliability of lead-free process solder joints. A typical example is the quality problem of PCB pads. Due to some shortcomings of the previous hot air leveling (HASL) pad coating process, the organic solderability protective layer (OSP) and Ni/Au coating processes are currently widely used by OEM manufacturers.

　　Among them, Ni/Au coating has two methods: immersion gold method and gold plating method. The immersion gold method is more favored by domestic manufacturers due to its simple process, but it is difficult to control the thickness of the Au layer. The insufficient thickness of the Au layer often leads to oxidation of the Ni layer underneath, affecting the performance of the solder joint during reflow soldering. In this case, manufacturers can generally use Auger electron spectrometer (AES) to accurately measure whether the thickness of the Au layer of the PCB pad meets the specifications.

　　(3) Process: In the process of SMT and MCM manufacturing, problems such as improper solder storage temperature, insufficient solder on pads, and improper reflow temperature curve setting are often encountered. For lead-free soldering, the optimization of the reflow process temperature curve is very important. Excellent process can ensure the formation of high-reliability soldering while keeping the peak temperature as low as possible.

　　Therefore, except for Japan, consumer electronics companies in other countries seem to have accepted the tin-silver-copper alloy series, in which silver accounts for 3.0% to 4.7% and copper accounts for 0.5% to 3.0%. The melting points of alloys with different components are not much different, basically between 217℃ and 221℃, while the liquidus temperature of tin-lead alloy (63% tin and 37% lead) is 183℃, a difference of 34℃.

　　Therefore, key variables in the reflow process, such as peak temperature, time above liquidus temperature, immersion time, immersion temperature, and ramp rate caused by the selection of flux and solder paste, are closely monitored to ensure that the reflow process maintains a Cpk of 1.33 or higher. Another point to note is the use of Bi-containing lead-free solder. Studies have found that when Bi-containing solder comes into contact with Sn-Pb-coated devices, Sn-Pb-Bi eutectic alloy will be generated after reflow soldering, with a melting point of only 99.6°C, which can easily lead to cracking of the soldering site. Therefore, when using Bi-containing lead-free solder, it is necessary to pay attention to whether the device coating is Sn-Pb coating.

　　In addition, there are void issues in lead-free soldering processes [14, 15]. Voids are a common defect in interconnect solder joints during reflow soldering, and are particularly prominent in devices such as BGA/CSP. Due to the large differences in the size, location, proportion, and measurement of voids, there is still no unified safety assessment of void levels. Experienced engineers are accustomed to classifying voids that are not large (the sum of the volume of small-sized voids does not exceed 0.5% of the volume of the solder joint), have a void ratio of less than 15% to 20%, and are not concentrated at the connection as a common defect in reflow soldering, and consider them acceptable;

　　On the other hand, according to Motorola's research results, voids with a diameter of 3μm to 5μm can actually improve the long-term reliability of solder joints because they can prevent the expansion of cracks in solder joints to a certain extent. However, it is generally believed that large voids, or voids reaching a certain proportion, will have an adverse effect on reliability.

　　Therefore, voids are still an issue that must be paid attention to in lead-free soldering. In the molten state, the surface tension of Sn/Ag/Cu alloy is greater than that of Sn-Pb alloy. The increase in surface tension will inevitably make it more difficult for the gas to overflow during the cooling stage, resulting in an increase in the void ratio. This has been confirmed in the development of lead-free solder paste. The results show that the number of voids in solder joints using lead-free solder paste is greater than that using tin-lead solder paste.

　　Large voids and some small spherical voids are caused by the volatilization of flux. The ratio of flux in solder paste is the most direct factor affecting the voids in solder joints, so lead-free solder paste still has a lot of room for improvement. As a new generation of lead-free solder paste products, Multicore (96SCLF32OAGS88) has achieved a great leap in technology by increasing the activity of flux at high temperatures, which can reduce the void level of lead-free solder joints to about 7.5%. In the past two years, with the progress in material research, the second-generation universal lead-free solder paste has been developed. In addition to having a wider process window, easier application, and better appearance, the most important thing is to solve the void problem.