INEMI Team Develops Test to Expedite Evaluation of Conformal Coatings

The INEMI Conformal Coating Evaluation for Improved Environmental Protection project team has developed and demonstrated a time-saving approach for evaluating conformal coatings.

Conformal coatings are used in electronics assemblies to protect printed circuit boards and components mounted on them from the deleterious effects of corrosive environments that have high concentrations of gases such as sulfur dioxide, hydrogen sulfide, free sulfur, chlorine, oxides of nitrogen and ozone. Particulate matter with low deliquescent relative humidity (DRH) can electrically short circuit features with potential differences across them by forming low resistance bridges (electrical short circuits) when the relative humidity in the air is above the DRH of the particulate matter.

As data centers and electronic devices proliferate worldwide into geographies with high levels of pollution and high relative humidity, the use of conformal coatings becomes necessary and is no longer restricted to mission-critical and/or military hardware. The decreasing feature size of components is also a factor. With decreasing feature gaps that dust particles and corrosion product particles can more readily bridge, conformal coatings are increasingly needed in today's designs.

Commercially available conformal coatings are available in a wide range of price points, application methods, and effectiveness in protecting the underlying metal from corrosion. The conventional method of testing the effectiveness of conformal coatings is to expose the conformally coated hardware to a corrosive environment for extended periods of time lasting many months and determine the mean time to failure. This means of testing is both inconvenient and slow. Even where the corrosion of the coated components can be monitored, such as in the case of surface-mounted resistors, it can take more than a year to evaluate a coating and the testing is done under very limited conditions of temperature, humidity and environmental corrosivity.

The INEMI Conformal Coating Evaluation for Improved Environmental Protection project team has developed and demonstrated a time-saving approach for evaluating conformal coatings. It involves coating thin films of copper and silver and monitoring the corrosion rates of the coated thin films while subjected to corrosive and humid environments. Effective conformal coatings protect the underlying metal thin films well.

Using a modified version of the INEMI flowers of sulfur (FoS) chamber the team exposed conformally coated thin films of copper and silver to a sulfur gas environment. Performances of acrylic, silicone and atomic layer deposited (ALD) conformal coatings were studied as a function of temperature and relative humidity. 

Testing Approach

The thin-film test vehicle (Figure 1) used for evaluating conformal coatings consisted of a serpentine metal (copper or silver) thin film (800nm thick) sputtered on oxidized silicon die 15x15mm. The FoS chamber was modified so that there was no forced air circulation and no chlorine gas in the chamber, only sulfur vapor. The resistances of the thin films were measured using potentiostats to pump known values of currents through the thin films and measuring the voltage drops across them. Thin film temperatures were monitored using thermocouples attached to a data logger. Resistance and temperature readings were taken simultaneously every 10 minutes over the 5-day period of each test.

Three conformal coatings were tested: acrylic coating 39-45 um thick; silicone coating 100 um thick; and atomic level deposition (ALD) coating 0.1 um thick. ALD coatings are ultra-thin (1-200 nm), stochiometric, dense and highly uniform in thickness. The ALD process is performed in a vacuum reactor at relatively low temperatures, typically 80-300°C, depending on the material deposited and the substrate thermal budget. To characterize/evaluate the effectiveness of the conformal coatings, the corrosion rates of the coated thin films were measured and compared with uncoated (bare) thin films.

Four tests were run with the chamber under the conditions shown on the psychometric chart of Figure 2: (1) 15% relative humidity, 40oC; (2) 15% relative humidity, 50oC; (3) 31% relative humidity, 50oC; and (4) 75% relative humidity, 50oC. The durations of various electrical and temperature conditions are listed in Table 1.

The effect of the chamber temperature is clear: a 40oC chamber environment is less corrosive than a 50oC environment. On the other hand, higher relative humidity makes the air less corrosive. At higher humidity, the sulfur concentration in the chamber decreases because of sulfur vapor absorption by surfaces with higher amounts of adsorbed moisture. Lower sulfur concentration makes the air less corrosive.

Figure 3 summarizes the corrosion rates of copper and silver serpentine thin films coated with acrylic or silicone and compares them to corrosion rates of bare (uncoated) copper and silver films. ALD coated thin film corrosion rates were not included because their corrosion rates were too low — they were within the limits of the experimental error.

The acrylic coating protected copper thin films from corrosion to some extent. In a 40oC FoS environment, the corrosion rate of acrylic coated copper was two orders of magnitude less compared to bare copper. Increasing the FoS environment temperature to 50oC increased the acrylic coated copper thin film corrosion rate by approximately an order of magnitude. Of the three humidity test conditions (15, 31 and 75%), the highest corrosion rate of acrylic coated copper was at 75% relative humidity. Acrylic coatings did not protect the underlying silver thin films. The increase in temperature from 40o to 50oC increased the silver corrosion rate somewhat. 
 
The silicone coating provided no corrosion protection to the underlying copper thin films. At 40oC and 15% relative humidity, silicone coating provided some corrosion protection to silver thin films. At 50oC, silicone coating provided no corrosion protection to the underlying silver thin film over the whole range of relative humidity tested. 

Another interesting observation is that Ag2S whiskers grew on the bare silver thin films and that these whiskers were prevented from growing by the three conformal coatings.
 
Predicting field performance often requires accelerated testing involving higher temperatures and harsher environments to shorten the test times to convenient durations. A downside of harsher conditions is that they may unduly overstress the hardware, thus changing the failure mechanism from one actually occurring in the field. A better approach is to keep the test conditions close to or the same as the field conditions and to shorten the test time by improving the sensitivity of detection of the hardware degradation. The latter approach was taken in the INEMI study by characterizing conformal coatings based on the corrosion rates of coated thin films that can be measured with down to +/-1 nm sensitivity. As a result, there is no need to accelerate the test conditions. This and earlier studies indicate that the environmental conditions in a flowers of sulfur chamber at 40oC are adequate for testing conformal coatings for many applications.

Conclusions

INEMI's study provides additional proof that conformal coatings can be evaluated effectively by the degree of corrosion protection they provide to coated copper and silver films. The corrosion rates of metal thin films can be very effectively and accurately measured by the 4-point technique and by taking into account the effect of temperature on film electrical resistance. The film temperature must be measured using a thermocouple with fine wires.  The film electrical resistance corrected for temperature is inversely proportional to the film thickness remaining.  The rate of decrease of remaining film thickness is the corrosion rate of the film and is used to characterize the corrosion protection provided by the conformal coating. Test environmental conditions in a modified INEMI FoS chamber at 40oC are aggressive enough for conformal coating testing. The advantage of not using conditions that are too aggressive is that the degradation mechanism in the test is the same as that in the field. This proposed approach will help evaluate conformal coatings much faster and more effectively than presently used methodologies, thereby enabling continuing advances in conformal coatings development.