- Program Areas
- Installation Energy and Water
- Environmental Restoration
- Munitions Response
- Resource Conservation and Resiliency
- Weapons Systems and Platforms
Development of Azeotropic Blends to Replace TCE and nPB in Vapor Degreasing Operations
Dr. Darren Williams | Sam Houston State University
Azeotropic blends that have similar properties as trichloroethylene (TCE) and n-propyl bromide (nPB, 1-Bromopropane, 1-BP) but without undesirable environmental, occupational, safety, and health properties were tested as sustainable drop-in replacements for TCE and nPB in vapor phase cleaning operations.
Some facilities are moving from nPB and TCE to blends with trans-1,2-dichloroethylene (tDCE), but this is seen as an interim solution since tDCE is regulated as a VOC. Therefore, the statement of need (SON) limited the scope of this project to a search for azeotropic blends that do not contain tDCE. Fully-enclosed vapor degreasing and aqueous degreasing systems were also ruled out as potential technologies of study by the SON.
The Hansen solubility parameters and ESH data on over 22,000 substances were analyzed to find blend components that were likely to form non-flammable blends. The most promising blend recipes were fractionally distilled to determine if a low-boiling azeotrope formed. The azeotropes were characterized with low-resolution Raman spectroscopy. The flash point behavior was determined according to ASTM D56 Tag closed cup analysis. Those azeotropes that passed the flash point screen were tested in a vial-based grease solvency test. The most promising azeotropes at this point were blended in larger quantities so that vapor degreasing trials against nPB and TCE could be held in a Branson 125 7.6-L (2-gal) capacity vapor degreaser. The initial plan to deflux reflow-soldered printed circuit boards was not realized due to extreme variation in the amount of baseline ionic contamination on freshly soldered boards. This led to the development of a vial-based solder defluxing test, that is very repeatable and useful for comparative testing of solvents. The soils used in these tests were a heavy marine grease and a general purpose solder paste. The fluid properties of the azeotropes were also characterized to yield the Hansen solubility parameters, density, volumetric expansion coefficient, viscosity, surface tension, and wettability index.
No fluids in this study cleaned grease or solder flux residue as well as TCE or nPB.
Seven new azeotropic blends were discovered during this project. Four azeotropes failed the flash point test. One failed the vial-based screening test against marine-grade grease. Two of the azeotropic blends survived the flash point and solvency screening tests. They are designated AZ6 (80% Suprion, 20% p-chlorobenzotrifluoride, Tb = 110 °C) and AZ7 (65% HFE-7500, 35% p-chlorobenzotrifluoride, Tb = 125 °C).
These two blends were evaluated in a Branson B125 vapor degreaser in comparison to the performance of nPB and TCE. All were required to clean marine-grade grease from SS316, Al7075, and brass parts with a 10-min vapor step, a 5-min 40-kHz ultrasonic step, a 5-min vapor rinse, and a 10-minute sub-zero drying step. The nPB and TCE removed 100% of the soil on all three part types as measured gravimetrically. The blend AZ6 removed less than 20% of the soil (16(±11)% on Al7075, 13(±2)% on SS316, and 4(±1)% on brass), and AZ7 removed less than 60% of the soil (55(±7)% on Al7075, 36(±3)% on SS316, and 15(±2)% on brass).
This is understandable in light of their Hansen solubility parameters. Each blend has a substantial amount of non-flammable hydrofluoroether solvent, which dramatically lowers the dispersion intermolecular attractive forces in the blend. Therefore, these two blends are not feasible candidates for replacing nPB and TCE in vapor degreasing operations.
In the vial-based defluxing study, the nPB was able to remove 81(±11)% of the ionic contamination, and TCE was able to remove 94(±6)% of the ionic contamination. The performance of the azeotropic blends was poor with AZ6 removing 3.0(±0.8)% and AZ7 removing 13(±6)%. These tests utilized 1 hour of 40 kHz ultrasonic extraction by the solvent in a capped vial at ambient temperature. The poor performance of our new azeotropic blends in this ambient temperature test does not warrant confidence that cleaning would improve in vapor defluxing. Therefore, these two blends are not feasible candidates for replacing nPB and TCE in vapor defluxing operations.
Both blends AZ6 and AZ7 are based on the VOC-exempt solvent CBTF, which is a drop-in replacement for toluene in automotive paint formulations. Specialty Materials Company suggests a CEL of 25 ppm. This is low. However, the volatility of CBTF is < 0.1 of nPB and < 0.05 of TCE. This lower volatility means that CBTF blends would be much easier to control with cooling coils.
These azeotropes are not suitable for replacing nPB or TCE. However, AZ6 and AZ7 do have HSPs that suggest they might be useful against silicone grease, Krytox, pump oil, and jet oil. This prediction would need to be tested at locations that deal with these types of soils.
If there is a subset of DoD cleaning operations at installations that require the strong solvency of nPB or TCE, then a process modification must be considered. The installation of air-tight degreasers with engineered controls would be worth the investment if these particular cleaning operations are critically important to the war fighter (i.e. liquid oxygen components, gyroscopes, optics, etc). These systems are designed with very efficient solvent handling and vacuum drying of parts. This greatly reduces worker exposure and solvent losses. They are typically large, automated, and enclosed, but the efficiencies in solvent recovery are substantial offsetting the purchase and installation expense. Because the solvent vapor zone is fully contained and oxygen free, inexpensive and effective flammable solvents may be used. Working toward this type of process change would maintain confidence in mission-critical cleaning while greatly reducing the solvent related ESH and regulatory risks.
Whereas the results for our blends were not particularly impressive, there are other positive outcomes of this work. The full list of 188 individuals and 107 companies and agencies is included in Appendix A. The project produced ten standard operating procedures included in Appendix B that may be adopted at other facilities.