| |
Closing The Loop on Solvent Cleaning Systems
Jose A. Berbel, Austin American Technology
Steve Stach, Austin American Technology
Dr. Hugh Cole, Cobar Americas
Abstract
The development of a closed-loop cleaning system with advanced process and statistical control for PCBs and other electronic assemblies has been recently made possible by a combination of separation techniques, electrochemical detection, and the special formulation of a CFC alternative nonhalogenated organic solvent with outstanding solubility and surface tension properties.
In the proposed cleaning system, the solvent is capable of undergoing a unique ion-exchange process in which the ions present in solution are transferred and retained in a macroreticular polymeric material, and therefore a "cleaned" solvent (free of ionic contamination) is regenerated repeatedly after each cycle.
The process is fully automated and controlled at all times based on ionic contamination levels by a conductivity controller and a pre-programmed PLC (programmable logic controller). By applying the Onsager theory as it relates to the conductance of ions in solution, the ionic contamination is directly measured by conductivity or resistivity detection. The complete cleaning cycle (wash, rinse, and dry) has been engineered to take place in one process chamber, in effect eliminating the need for additional materials handling or transferring (see process diagram).
Given the solvent drying characteristics (low boiling point and high vapor pressure), the drying cycle can be effectively achieved with compressed air at room temperature or with heated Nitrogen.
The solvent blend has been optimized in order to attain a high degree of sensitivity for quantitative ionic contamination detection while maintaining a Hanson number profile that is suitable for rosin fluxes. The solvent blend exhibits both polar and non-polar solvency.
INTRODUCTION
A new patent pending closed-loop solvent cleaning system has been developed at Austin American Technology as a result of the market necessity for a reliable, economical, and practical organic solvent based cleaning process for electronic assemblies and other precision cleaning applications. Given the confusion and distress created as a result of the current regulatory climate as it relates to environmental issues, this new closed-loop solvent cleaning process presents in some cases a direct alternative to CFC and HCFC solvents. The mentioned cleaning system is fully capable of using de-ionized water for high reliability/aqueous cleaning applications as well.
By developing and introducing this new cleaning system, Austin American Technology has taken a unique and combined approach to critical cleaning concerns: environmental responsibility, manufacturing costs, safety, quality, and reliability.
The cleaning process involves three major components:
1. System controller with dual process control capabilities (time mode / contamination level mode)
2. Solvent regeneration system
3. The cleaning organic solution (with characteristic chemical and physical characteristics).
The cleaning process can be summarized as follows: materials to be cleaned are placed in a process chamber, which is subsequently filled with clean solvent (free of ionic contamination). The solvent undergoes mechanical agitation as a result of a spray under immersion. The level of ionic contamination of the materials is concurrently monitored by conductivity detection of the solution as ions are removed from the material surfaces and dissolved in the solvent. Depending on the nature of the materials and the ions present, there will be a point in time when the conductivity of the solution reaches a stable level. At this point, the slope of the ionic concentration as a function of time is zero (dC/dt=0) which indicates that all of the freely ionizable contamination has been removed from the materials’ surfaces. The following step in the process involves the regeneration of the contaminated solution through an ion exchange column. The slope of the ionic concentration of the solvent as a function of time is negative (dC/dt<0) until the ionic concentration reaches the pre-determined desirable level of cleanliness. The process therefore provides an analytical real time assurance and reliability.
PROCESS ANALYSIS
1. Process control
The process control mechanism makes use of a resistivity/conductivity controller. This device basically consists of a Wheatson Bridge with a variable impedance. A high frequency alternating current is transmitted through a two opposing electrodes cell, which is immersed in solution. The impedance of the solution is counter balanced by the variable impedance until a null point is achieved. As the number of ions and their mobilities in solution increases, the resistivity decreases, and vice-versa (See reference 1 below). Sophisticated resistivity controllers are programmable and capable of transmitting real time data as well as controlling the process by electrical relays. For some cleaning applications, a time mode process control may be desirable, in which the materials are cleaned up to a preset time period regardless of resistivity/conductivity readings.
1. For Alpha 615 RMA flux (solute) in Megasolv™ JB (solvent):
|
Resistivity (MOhm-cm) |
Ionic Concentration (% vol.) |
38.01 |
0.004 |
36.17 |
0.008 |
34.06 |
0.016 |
31.09 |
0.028 |
29.27 |
0.04 |
26.98 |
0.06 |
25.22 |
0.08 |
23.97 |
0.1 |
22.96 |
0.12 |
21.66* |
0.16 |
From the previous table, a relationship of 1/R vs. Ö
c can be obtained. A least-square fitting provides the following equation:
1 / R = 0.0221 + 0.061
Ö
c.
In this case,
a
= 0.0221
and
l
= 0.0612
2. For KCI in Megasolv™
JB:
|
Resistivity (MOhm-cm) |
Ionic Concentration (mgr/L) |
37.55 |
0.0784 |
36.40 |
0.1568 |
35.24 |
0.2353 |
33.07 |
0.3921 |
29.78 |
0.7058 |
26.26 |
1.0979 |
24.32 |
1.3332 |
21.81 |
1.7253 |
14.40* |
3.2153 |
From the previous table, a relationship of 1/R vs. Ö
c can be obtained. A least-square fitting provides the following equation:
1 / R = 0.0226 + 0.0127
Ö
c
In this case,
a
= 0.0226
and
l
= 0.0127
*Resistivity values below these do not represent an accurate measurement of ionic contamination because of intrinsic ion-ion interactions. The Onsager theory only applies to dilute solutions (See reference 3 below).
where R is the resistivity expressed in MOhm-cm and Ionic concentration in mgr/L.
This graph illustrates the Mega™ Cleaner process capabilities for real time quantitative ionic contamination detection in equivalents of KCI. With a known solvent volume and a known surface area of the materials, resistivity readings can be directly expressed in mgr KCI/in².
2. Solvent regeneration / ion exchange
The solvent is recovered by an ion exchange mechanism taking place in special macroreticular polymeric resins. These resins are capable of undergoing an ion exchange process in non-aqueous media. The resins are chemically compatible with Megasolv™ JB and other organic agents. The ion exchange mechanism can be illustrated as follows (See reference 4 below):
1. Cation exchange:
A proton from the sulfonic acid group reacts with a cation in solution (analogous to a Sodium ion) in effect "trapping" the ion from the cleaning solution.
2. Anion exchange:
In this case, the hydroxide ion bonded to the tertiary amine reacts with an anion in solution. RZ represents the polymeric structure.
The macroreticular nature of the resins makes it possible to statistically utilize most available sites within the polymer. It is important to note that both ion exchange reactions are reversible. By applying Le Chatelier’s Principle, it can be deducted that the resins themselves are recoverable (the cation exchange resin can be treated with Sulfuric Acid and the anion exchange resin with Sodium Hydroxide).
This derivation takes into assumption a 100% efficiency (all ions in solution entering the ion exchange chamber are retained), steady state constant flow, and no concentration gradients throughout the solution.
The significance of this final mathematical expression is that the % efficiency of the ion exchange resins can be known by comparing the actual recovery time with the theoretical recovery time. When new resins are placed in the system, the approximate % efficiency is about 80%.
3. Mega™ Cleaner process illustration
The following graph represents a typical Mega™ Cleaner cleaning process. It is an actual cleaning cycle in which RMA flux was removed from printed circuit board. Line 1 indicates the start of the cleaning cycle. The solvent has a resistivity of about 33.5 MOhm-cm at that point. The time period between line 1 and line 2 (named C) is considered the actual cleaning/wash cycle. The solvent is agitated by the effect of immersion spray. A sharp drop in resistivity values is evident, indicating that the ionic contamination is being transferred from the boards’ surfaces into solution (dR/dT<0). At the intersection of line 2 and the curve, a near 0 slope is noticeable. This indicates that the cleaning/wash cycle is completed. At this point, the controller signals the start of the solvent regeneration/rinse cycle (SR). The slope of the curve is positive (dR/dT>0). This cycle continues until the preset resistivity value of the solvent is reached (in this case 40.0 MOhm-cm) indicated by line 3. The solvent at this point is automatically drained into the holding tank/reservoir, and the dryer is activated.
where R is the resistivity expressed in MOhm-cm and time in seconds.
4. Nature of the preferred solvent Megasolv™ JB
Megasolv™ JB offers an alternative to other cleaning solutions currently used for certain cleaning applications. Megasolv JB has been deliberately engineered to remove Rosin and Resin type contaminants from printed circuit boards. The chemical structures of the solvent blend constitute a thermodynamically stable solution with both polar and non-polar solubility capabilities.
The solvent exhibits the following physical properties.
|
Property |
Units/Conditions |
Values |
Average Molecular Weight |
u.m.a. |
128.0 |
Boiling Point |
ºC @ 760 mm |
147.1 |
Specific Gravity |
(Water = 1) |
0.944 |
Flash Point |
ºF |
116.0 |
Evaporation Rate |
(BuAc = 100) |
30.1 |
%Soluble in Water |
25ºC |
18% |
Refractive Index |
25ºC |
1.4 |
Surface Tension |
Dynes/cm |
25.0 – 27.0 |
Vapor Pressure |
mm Hg @ 20 ºC |
3.2 |
Viscosity |
cps @ 25 ºC |
1.5 |
Heat of Vaporization |
Cal / ºC |
89.0 |
Specific Heat |
Cal / gr ºC |
0.46 |
These parameters predict a wide range of solubilities for different substances commonly encountered in electronic assemblies, particularly rosin fluxes. Experimental results confirm that the given solvent blend effectively dissolves a wide variety of oils, soils, and salts.
The mentioned solvent blend can be modified to customize and optimize different cleaning applications while maintaining the key required characteristics for the applicability of the mentioned closed-loop cleaning process.
Other commercial solvent blends have been tested in the cleaning process with positive results. However, in most cases one or more characteristic features are lost such as the evaporation rate (related to the drying time), or the process control based on conductivity detection.
Other cleaning solutions such as water and 2-Propanol can be incorporated effectively into the present closed-loop cleaning process with minor equipment modifications.
Conclusion
Until recently, cleaning systems in electronic assembly operations simply consisted in closed-loop vapor degreasing systems in which halogenated solvents, CFCs and HCFCs in particular, where used given their desirable characteristics of effective cleaning, no flammability, and high vapor pressure. However, because of their negative environmental impact (ozone depleting and global warming potential) they have been phased out rather rapidly. The debate over feasible and practical replacement alternatives continues. Aqueous and semi-aqueous systems have been successfully implemented for some cleaning applications. However, a virtually waste free, inexpensive, reliable, and safe cleaning process has not been developed prior to the proposed Megacleaner™ process.
This innovating cleaning system not only fulfills the mentioned features, but in addition, presents capabilities of process control and real time quantitative analytical data, eliminating the need for additional quality control measuring devices. The typical cleaning cycle (including drying time) is between ten to fifteen minutes.
References
- D.P. Shoemaker, "Experiments in Physical Chemistry." 5th ed., chap. 8, McGraw-Hill, New York (1989).
- L. Onsager, Phys. Z. 28, 277 (1927).
- P.W. Atkins, "Physical Chemistry," 3rd ed., chap. 27, Freeman, New York (1986).
- C.E. Hardland, "Ion Exchange: Theory and Practice," 2nd ed., Royal Society of Chemistry (1994).
- C.M. Hansen, Journal of Paint Technology 39, (505) 104-117 (1967).
- A.F.M. Barton, Chemical Rev. 75 (6) 731-753 (1975).
Copyright © 1998 Austin American Technology Corporation, All Rights Reserved.
|