N2 Inerting for Reflow of Lead-free Solders
11/01/2003
Study demonstrates the benefit of N2.
BY C. CHRISTINE DONG, BRENDA F. ROSS AND ALEXANDER SCHWARZ
Due to environmental and health concerns about lead contamination, there is increasing pressure in legislation and a growing market demand to eliminate toxic materials. Lead-free soldering for the electronics industry is becoming a global trend, and several alloy systems alternatives have been recommended.
This change of solder material substantially impacts the electronics packaging industry, including the manufacturers of flux and solder paste, components, circuit boards, soldering equipment and industrial gas. The reason is that compared to traditionally used eutectic tin-lead solder, lead-free solders have different melting points, oxidation potentials, surface tensions and solubility
eactivity than the base metals to be soldered.
As an industrial gas supplier, we have supplied N2 to the packaging industry for reflow soldering of tin-lead solders for many years. Reflow soldering is the most widely used soldering method for surface mount technology. The key function of N2 on reflow soldering is to minimize oxidations on both solder and base metals to be soldered — ensuring soldering quality.
Focus of Study
In a production-scale reflow process, many processing variables interact with each other — increasing the complexity of a study. We relied on lab-scale experiments to get an initial understanding of the benefit of N2 for reflow of lead-free solders.
 |
null
 Figure 1. Self-alignment of printing offset and formation of solder balls in N2 for a solder paste with 25-45 µm particle size. |
null
There are two traditional lab-scale methods for evaluating soldering performance: a spreading test, and a wetting balance test. Generally, the spreading test is considered more reliable for investigating solder reflow — especially for lead-free solders. In a spreading test, a small volume of a solder to be tested is placed on the top of the metal surface to be soldered. Then, the extent of the solder coverage on the metal surface is measured under a simulated reflow condition. This soldering configuration is quite similar to those of real reflow processes.
In the wetting balance test, a metal coupon representing the substrate to be soldered is dipped into a molten solder bath under a controlled condition. Then, time-variant forces the coupon subjected (downward liquid surface tension force and upward buoyancy force) are measured. In this soldering arrangement, the volume of the metal coupon immersed in the solder bath is usually much smaller than that of the molten solder. This arrangement is quite different than the case in reflow soldering, and closer to that of wave soldering. For reflow of lead-free solders, it has been discovered that the solders can wet, but not spread on a copper surface with a common surface finish. The primary factor that restricts lead-free solders from spreading is the limited solubility of the base metal in the molten solder. For lead-free solders, the volume ratio between the molten solder and the base metal is extremely sensitive to the degree of wetting and spreading, so the result obtained by using the wetting balance test may not accurately represent the performance for reflow.
 |
null
 Figure 2. Self-alignment of printing offset and formation of solder balls in air for a solder paste with 25-45 µm particle size. |
null
To better understand the benefit of N2 on reflow of lead-free solders, we decided to focus on nonwetting/spreading-related soldering performances, such as formation of noncoalescent solder balls and self-alignments of surface mount components and printing offset.
Noncoalescent Solder Balls. In the case of insufficient fluxing or soldering oxidation during reflow, noncoalescent solder balls may form on the soldermask surrounding the solder pads — especially when fine-pitch solder paste is used. Formation of such balls results from a reduced surface tension of the molten solder in the presence of surface oxides, and leads to reduced cohesion energy of the molten solder.
The cohesion energy of a liquid is the energy of a liquid required to separate a liquid column with a unit cross section to create two unit surfaces. Therefore, the cohesion energy, Eco, equals two times the liquid surface tension, γlg, such as:
Eco=γlg
Self-Alignment. Misalignment between surface mount components and solder-printed traces or offsets between printed solder paste and corresponding metal pads on a circuit board are common problems — especially when automatic printing and placing methods are applied. During reflow, misalignments can be corrected if the surface tensions of the molten solder and base metal to be soldered are sufficiently high, or the oxidations on both solder and base metal are minimized.
The critical driving force for self-alignment is the adhesion energy, which is the energy required to pull apart a unit interface between a solid and a liquid into two separate surfaces. Based on this definition, the adhesion energy, Ead, can be expressed as:
Ead = γsg + γlg - γsl
In the case of self-alignment, γsg, represents the surface tension of the solid component lead (for component self-alignment) or the solid metal pad (self-alignment of printing offset). The surface tension of the liquid holder is again γlg, and γsl stands for the interfacial tension between the solid and liquid phases. The greater the adhesion energy, the stronger the attraction between the solid and liquid phases will be.
Requirements of Soldering Performance
A summary of necessary requirements for self-alignment, wetting and spreading is shown in Table 1. For self-alignment to spontaneously occur, the adhesion energy, Ead has to be greater than zero. This in turn leads to the requirement that the sum of the surface tensions of the lead component and the molten solder, γsg + γlg, be larger than the interfacial tension of the solid/liquid interface, γsl. Wetting, by definition, is contact angle, q, less than 90°. According to Young's equation, the contact angle, q, can be expressed as cosq = (γsg - γsl)/γlg. For a molten solder to wet on a metal substrate (q <90° and cosq > 0), the surface tension of the metal substrate, γsg, has to be larger than the interfacial tension of the solid/liquid interface, γsl. For solder spreading to be energetically favorable, the adhesion energy of the liquid solder on the metal surface, Ead, has to be greater than the cohesion energy of the liquid solder, Eco. The surface tension of the metal to be soldered, γsg, has to be larger than the sum of the surface tension of the liquid solder and the interfacial tension of the solid/liquid interface, γsg + γlg.
Among the three requirements of soldering performances, Table 1 indicates that the most difficult to satisfy is the one for solder spreading, γsg > γlg + γsl. Solder wetting, (γsg > γsl), is less challenging, and the requirement for self-alignment is the easiest to satisfy (γsg + γlg > γsl). It can be deduced that once a solder wets on a metal surface, the spreading may or may not occur. For a solder with relatively poor wetting, satisfactory self-alignment may be achieved. As long as soldering performance of the system is intrinsically good, the benefit of N2 inerting for that soldering performance can be demonstrated.
Experimental Results
Self-alignment of Surface Mount Component. To investigate the component's self-alignment ability in response to the reflow environment, a solder paste sample of 95.5 Sn/ 3.8 Ag/0.7 Cu with no-clean flux and a particle size range of 25-45 µm was printed on the trace area of a circuit board. Then, a component with pre-tinned leads was placed on the circuit board with a certain degree of misalignment to the solder printed traces. We controlled the degree of misalignment to be relatively constant from one sample to another. Prepared samples were reflowed in a heating stage under N2 or airflow. During each reflow, realignment was monitored in-situ under an optical microscope. The configuration of the samples was photographed before and after each reflow, demonstrating the percentage of rEcovery.
We compared the difference of component self-alignment in N2 vs. air. Table 2 is a summary of the measured results. It shows that for all the samples tested, the recoveries from the initial misalignment in N2 are all greater than 90 percent; however, only 50 to 60 percent of the recoveries can be achieved in air. The average recoveries in N2 and air are 93.2 percent and 53.6 percent, respectively. The higher self-alignment ability in N2 is believed to be caused by the higher surface tensions of the molten solder and component leads, as a result of minimized oxidation. As indicated in Equation 2, the high surface tensions lead to high adhesion energy. This beneficial effect of N2 on component self-alignment was also observed previously in the soldering system of eutectic tin-lead.
 |
null
 |
null
Self-alignment of printing offset. To study the significance of N2 inerting on the self-alignment of printing offset, samples were prepared by creating an offset between printed solder paste and the metal trace area of a circuit board. Prepared samples were also reflowed in the heating stage and in-situ observed under an optical microscope. Figures 1 and 2 are photographs showing self-alignment in N2 vs. that in air for ~50 percent printing offset in vertical direction. The solder was the same one previously used, of 95.5 Sn/ 3.8 Ag/0.7 Cu, with no-clean flux and a particle size range of 25-45 µm. We demonstrated that in N2 a complete recovery is achieved; in air, solder can withdraw from the soldermask area but cannot cover the entire metal pads. The high self-alignment ability in N2 is a result of the high adhesion energy maintained in the inert condition.
Formation of Noncoalescent Solder Balls. Samples with a printing offset can also be used to investigate the formation of noncoalescent solder balls. After the solder melts, it can automatically withdraw to the metal pads and may leave noncoalescent solder balls on the soldermask area, if the solder is oxidized to a certain degree. Comparing Figures 1 and 2 shows that there were few solder balls left on the soldermask area when reflow was conducted in air, and no solder balls for at all when reflowed in N2. It is understood that as the particle size in a solder paste decreases, the surface-to-volume ratio of each particle increases. This promotes solder oxidation and increases the tendency of forming noncoalescent solder balls. We also demonstrated the effect of the particle size of the solder on the formation of solder balls. In this case the solder used was 95.5 Sn/ 3.8 Ag/0.7 Cu, with no-clean flux and a particle size <31 µm. This confirmed that when the particle size of the solder is reduced, the formation of noncoalescent solder balls in ambient air is more severe — the benefit of N2 becomes more significant.
Conclusion
Our current study's results indicate that N2 plays an important role in correcting manufacturing defects produced before reflow. These manufacturing defects, such as component misalignment and printing offset, may remain in the final product if reflow is conducted in ambient air. In addition, such defects can also lead to the formation of noncoalescent solder balls in the soldermask area if reflow is conducted in air. The problem of the formation of solder balls will be more severe when fine-pitch solder paste is used.
References
For a complete list of references, please contact the authors.
C. CHRISTINE DONG, senior principal research scientist, and ALEXANDER SCHWARZ, senior research associate, may be contacted at Air Products and Chemicals Inc., 7201 Hamilton Blvd., Allentown, PA 18195-1501; (800) 654-4567. BRENDA F. ROSS, senior research technician, may be contacted at (610) 481-8078.