Sulfurous vapors emitted from EPDM rubber can damage electronic instruments.

Sulfurous emissions can cause big problems with electronic components that depend upon the chemical inertness of silver. EPDM rubber compounds emit sulfurous compounds when heated moderately if the EPDM is vulcanized with sulfur-containing compounds. The emissions may be carbon disulfide ([CS.sub.2]) and carbon oxysulfide (also known as carbonyl sulfide, formula COS).

This article discusses identification of the problem and a simple resolution of the problem. It hypothesizes but does not confirm the possible sources of sulfur vapors from sulfur-vulcanized EPDM.

Problems with sulfurous emissions

Silver is often the metal of choice for electrical contacts in switches and relays. It is often the material of choice for thin-film or thick-film deposits of conductive materials used for electronic filters or tuners. Unfortunately, silver is susceptible to reaction with sulfurous compounds as noted in the free energy discussion later. A silver sulfide film is not electrically conductive and a switch contact may fail because of this black surface tarnish. In extreme cases, the silver may actually be etched away, causing a change of resonant frequency of an electronic tuner.

Copper is even more reactive to sulfurous vapors, as are both lead and tin in solder. The nature of these materials and their uses influence whether the sulfide tarnish might be critical in performance of the electronic instrument.

Disk drive manufacturers have worked through the problems of carbon disulfide emissions from EPDM, particularly where the seals of bearings were made with this elastomer. Sulfide emissions have wreaked havoc with lubricants and the bearing materials of disk drives.

A NASA web site lists outgassing characteristics of materials that might be considered for use in spacecraft and satellites - http://epims.gsfc.nasa.gov/og-cgi/uncgi. The six listings there for EPDM rubber artifacts show 3.4 to 26.4 weight percent of the rubber outgassing during 24 hours at 125 [degrees] C, and show from 1 to 14.6% condensable matter from these materials - much more than is permissible for these applications. The basic listing does not identify the nature of the outgassing chemicals from the EPDM.

Chemical reactions and their free energies

Published data, e.g., Handbook of Chemistry and Physics, provide data on the free energy of formation of many different chemical compounds. The free energy of reaction at 20 [degrees] C can be calculated directly from these numbers by subtracting the formation energies of the reactants from the formation energies of the products of reaction. This requires a stoichiometrically balanced equation, of course. The free energy of an element is considered to be zero.

If there is a net decrease of free energy, the reaction has a probability of going in the direction as written. This is comparable to the potential energy condition of a rock on a hillside - it can spontaneously roll downhill to a lower energy level - it can't spontaneously roll uphill. The calculation does not necessarily predict whether the reaction will take place, because the rock may be embedded in soil and may need to be kicked upward quite a bit before it will roll downward. For example, the reaction of cellulose with oxygen is a downhill reaction, but fortunately papers must be brought up to a higher energy state (ignition temperature) before they ignite and burn to a lower energy state.

A number of reactions are is in table 1 showing probability of these events. The two first of these reactions show how carbon disulfide may be converted into the carbon oxysulfide. The next four show how silver and copper may tarnish in contact with each of these vapors.

Table 1 - reactions and probabilities

1. 2 [CS.sub.2] + [30.sub.2] [right arrow]
   2 COS + [S0.sub.2] [Delta]G =
   -537 kJ/gmol of [CS.sub.2]
2. 2 [CS.sub.2] + [H.sub.2]0 [right arrow] COS + [H.sub.2]S
   [Delta]G = -108 kJ/gmol of [CS.sub.2]
3. [CS.sub.2] + 4 Ag + [O.sub.2] [right arrow] 2 [Ag.sub.2]S
   + 2 [CO.sub.2] [Delta]G = -413 kJ/gmol of [CS.sub.2]
4. 2 COS + 4 Ag + [O.sub.2] [right arrow] 2 [Ag.sub.2]S
   + 2 [CO.sub.2] [Delta]G = - 284 kJ/gmol of COS
5. 2 [CS.sub.2] + 4 Ag + [O.sub.2] [right arrow] 2 [Ag.sub.2]S
   + 2COS [Delta]G = -277 kJ/gmol of COS
6. 2 [CS.sub.2] + 2 CuO + [O.sub.2] [right arrow] 2 CuS +
   [CO.sub.2] [Delta]G = -569 kJ/gmol of [CS.sub.2]
7. [CS.sub.2] + 2 Zn[CO.sub.3] [right arrow] 2 ZnS + 3
   [C0.sub.2] [Delta]G = -187 kJ/gmol of [CS.sub.2]
8. [CS.sub.2] + Zn + [O.sub.2] [right arrow] 2 ZnS + [CO.sub.2]
   [Delta]G = -663 kJ/gmol of [CS.sub.2]

The zinc carbonate reaction (#7) with carbon disulfide shows how the zinc compound can be used as a safe "getter" or reactant to remove carbon disulfide from an environment. However, this reaction is less energetic than the reaction of carbon disulfide with silver, so it would need to be located in such a manner to be a shield or fence between the source and the silver to be protected.

Reaction #8 shows how a galvanized steel screen or a brass screen could be used as a formed shield for such protection.

Identification and quantitation of sulfurous compounds An electronics manufacturer received customer complaints of nonfunctioning "black boxes," after about five years of use. He saw the visible evidence of black silver tarnish on sensitive surfaces of returned instruments. This prompted an examination of the silver by scanning electron microscopy with energy dispersive spectroscopy, which confirmed that sulfur was present on the silver surface. Figure 1 shows a photo of a silver film resonator with its connection to circuits on the PC board, and figure 2 shows the damage done to a resonator, with electrical disconnect in the lower right corner and two black spots of silver sulfide on the central surface.

[ILLUSTRATIONS OMITTED]

The manufacturer then learned that the EPDM gaskets used to hermetically seal the electronic package contained sulfur. With this clue, some boxes and some gaskets were taken to Hauser Laboratories for identification and quantitation of the vapors within the box. A box was swept with helium gas which was then analyzed by gas chromatography with mass spectroscopy (GC/MS) and both carbon disulfide and carbon oxysulfide were confirmed to be present. The box had a free volume of approximately 1.85 liters and the EPDM gasket weighed approximately 10.8 grams. The box material was a glass-filled polycarbonate, and the printed circuit board appeared to be a customary epoxy G- 10 laminate.

These electronic packages can be exposed to temperatures up to 150 [degrees] F during shipping and repeatedly during use. Most of the customer complaints came from "sunbelt" locations in the U.S. For this reason, an experimental program was undertaken using GC/MS to learn:

* Accelerated sulfurous vapor generation with three EPDM gaskets free in a box, heated to 65 [degrees] C for two hours.

* Residual sulfurous vapors upon flushing with helium after removal of the sealing gasket from a "used" box to learn whether the case and circuit board may have absorbed the sulfur compounds. The box was heated to 65 [degrees] C for two hours with a silicone rubber gasket in place.

* Decay characteristics of sulfurous vapors in the box while heated to 65 [degrees] C and flushed with helium.

* Similar tests to confirm effectiveness of treatments for removal of the sulfurous vapors.

Figure 3 shows the electronic box with its circuit board and with three of the EPDM gaskets after the first outgassing test. This box had an overwhelming amount (approximately 3.7 micrograms) of noxious vapors upon removal from the oven, which was mostly carbon disulfide.

[ILLUSTRATION OMITTED]

The second test showed that both of the sulfurous vapors were still present in the box after removal of the EPDM gasket. Carbon oxysulfide was particularly prevalent in the electronic components, and carbon disulfide was particularly prevalent in the box material.

Figure 4 shows the characteristics of carbon disulfide concentration with flushing time in a box that had been Pleated two hours at 65 [degrees] C. These are concentrations at room temperature for three days after removal from the oven, and three days of flushing extrapolated in order to fully remove the noxious vapors at about five days.

[ILLUSTRATION OMITTED]

Treatments to remove sulfurous vapors

Since activated carbon is an effective adsorber of these vapors in the thermal desorption cartridge used for GC/MS, a packet of activated carbon should be an effective absorber if placed within the sealed box. Such a packet was prepared using 12 grams of a chemical grade charcoal adsorbent, and it was found to adsorb the offending gases effectively.

A commercial supplier of adsorbent packets was contacted and a packet containing activated carbon and potassium permanganate was suggested as having proven effectiveness in hydrogen sulfide environments. The oxidizing reaction for the permanganate looked probable by the following reaction:

[CS.sub.2] + 4KMn[O.sub.4] [right arrow] 4Mn[O.sub.2] + 2[K.sub.2][SO.sub.4] + 2[K.sub.2][SO.sub.4] + C [Delta]G = -1853 kJ/gmol

However, tests showed that the sulfur concentration was not decreased in a period of 16 hours at 65 [degrees] C.

A packet was then made up of this supplier's activated carbon with no additives in a home-made Tyvek envelope (portions of a mailing envelope). With 15 grams of carbon in place, the carbon disulfide concentration in a used box was now knocked down by 99.8%. Only 0.05 nanograms of [CS.sub.2] remained in the box. The carbon supplier expected the carbon to absorb up to 0.6% its weight of [CS.sub.2] and to 0.1% its weight of COS. Although the total capacity of charcoal could not be quantified over its lifetime, this material was considered to be very effective for adsorbing the microgram quantities of sulfurous compounds desorbing from the package.

A packet of zinc carbonate may be equally effective as a "getter" for the sulfur compounds by reaction #7 in table 1 if judiciously located, and the reaction shows 0.3 grams of carbon disulfide reacting with one gram of zinc carbonate. Copper and lead compounds also react with carbon disulfide, but zinc is more benign environmentally.

Avoidance

It is better to avoid the problem than to have to fix the problem. If EPDM is the elastomer of choice for an electronics package, it should be specified as having a peroxide cure. And qualification samples of the rubber should be chemically analyzed for sulfur content before the product is sent out for beta tests.

The list of vulcanizing or accelerating chemicals in table 2 is not uncommon in EPDM elastomers. All have carbon atoms attached to two sulfur atoms in their structures, and these might be sources of the carbon disulfide outgassing. Free sulfur is often used at 2 phr concentration in EPDM elastomers.

Table 2 - common vulcanizers, accelarators for EPDM

Acronym   Chemical Name
MBT       2-mercaptobenzothiazole
TMTD      tetramethylthiouram monosulfide
DPTT      dipentamethylene thiuram tetrasulfide
ZDBC      zinc dibutyldithiocarbamate

Industry sources have mentioned that the EPDM industry presently vulcanizes with sulfur in about 75% of the products and with peroxide in about 25%.

A peroxide-cured EPDM should be free from sulfur. Silicone rubber is a more expensive alternative, but a very effective alternative for non-sulfurous elastomers.

Conclusion

The vulcanizing agents listed above are used in many other elastomers, and we have not heard of any carbon disulfide problems in other than EPDM products. Most elastomers have a large number of reactive butadiene groups and thus are very reactive to any sulfur that may be present. EPDM is unique in that it has only a few crosslinking sites (vinyl radicals) and thus has a greater probability of residual sulfurous compounds from decomposition of the vulcanizing and accelerating agents.

Ray Hauser began his rubber industry career in 1950 at Connecticut Hard Rubber where he was co-inventor of the first U.S. patented process for bonding Teflon. His present consulting projects include problem solving, patent litigation and products liability litigation.