Note 88: Analysis of Silicone Contaminants on Electronic Components by Thermal Desorption GC-MS

by Eric D. Butrym, Scientific Instrument Services, Ringoes, NJ
6/23/00

Abstract:

Silicone based polymer and oligomer contamination on electronic components is assessed. Methodology is suggested for qualitative and semi-quantitative analysis of silicone-based contaminants on electronic and flow path parts for critical applications. A portable gas sensing apparatus was evaluated to determine the source of fugitive volatile silicone compounds. Results show that TD-GC-MS can be used to target silicone emitting parts and determine the nature and amount of contamination being generated. Butylated Hydroxytoluene (BHT) was also measured as an off-gassing product. Drawbacks to the method include silicone artifacts from the GC column and inlet. Ways of minimizing these artifacts for more precise determination of contaminants are discussed.

Introduction:

Silicone compounds have become necessary and nearly ubiquitous components of a wide range of manufactured products and processes. Uses of silicone polymers range from lubricants, adhesives, films and barriers to process additives like surfactants or plasticizers. One of the characteristics of silicone-oxygen oligomers is their high vapor pressure. Some compounds with molecular weights of over 800 amu can be analyzed by GC at relatively low temperatures due to their high silicone content. While this aspect has been used to great advantage in the derivitization of less volatile analytes, the tendency of silicone polymer components and oxidation products to off-gas from their matrix can pose problems for certain devices. Particularly, magnetic and chemical sensors are affected by deposition of silicone compounds on their surfaces, and for this reason manufacturers of electronic data storage systems and chemical sensors need to carefully control and monitor potential sources of silicone contaminants in their systems. For example, parts-per-million (ppm) silicone oligomer emission from components inside a portable hydrocarbon detector can polymerize on the surface of the heated catalyst sensor and severely impede its performance in a matter of hours. Likewise given the small area of hard disk read-write heads, minimal exposure to silicone compounds can result in errors or drive failure. Information about the amount and types of silicone containing materials present in an assembly can be the key to controlling their migration to sensitive components, as well as an aid to setting manufacturing material specifications.

Methods and Instrumentation:

Two general methods were used to determine the sources of volatile silicone materials present in a portable multi-function gas sensing apparatus. Direct Thermal Extraction-Gas Chromatography-Mass Spectrometry (DTE-GC-MS) was employed to assess polymer components used for keypads and rubberized 'boot' pieces used in the assembly. Purge & Trap-Thermal Desorption-Gas Chromatography-Mass Spectrometry (P&T-TD-GC-MS) was also used to evaluate whole assemblies and larger components for the presence of silicone compounds and siloxane oligomers. In the DTE-GC-MS experiments, a small (<10 mg) sample of the polymer was cut and deposited into a glass lined thermal desorption tube (GLT). The GLT was plugged with silanized glass wool to keep the sample in place. Both the GLT and the glass wool had previously been conditioned for 2 hours at 350 >°C with clean nitrogen flow.

The portable hydrocarbon sensor was brought in by a potential customer to evaluate the suitability of Thermal Desorption as an analytical tool to determine the source of siloxane materials which had been found to polymerize on the surface of a catalytic element. Several materials were proposed as likely sources of the silicone material, including a rubberized 'boot' that enclosed a liquid crystal display, a gasket that served to weatherproof the unit, and a plastic overlay that protected the instrument from dirt and excess moisture.

Figure 1. Heated Sampling Chamber (A) and Purge & Trap vessel (B)

The liquid crystal display was also considered a likely source of the silicone contamination, and a single display was provided for Purge and Trap testing. Two of the handheld instruments were also sampled in their entirety by P&T. The objective was to evaluate the use of a potting resin to halt fugitive siloxane emissions from the LCD. P&T-TD-GC-MS was performed by isolating the pieces to be studied in a heated sampling chamber or a glass Purge & Trap vessel (both from SIS, Inc.) and purging with nitrogen onto a GLT packed with 100 mg Tenax® TA (SIS, Inc.). Tenax GLT traps were conditioned for 2 hours at 300 >°C with nitrogen flow. Whole units were each sampled for 60 minutes at 100 >°C with a flow of 100 ml/min through the P&T vessel (Figure 1B). The vessel was kept at temperature by placing it in a laboratory oven and plumbing purge gas and sampling ports outside of the oven. A sample was collected from the LCD by placing it in a heated sampling chamber (Figure 1A) set at 80 >°C and purging at 100 ml/minute for 30 minutes. The exposed LCD was considered likely to emit more siloxane materials at lower temperature than the enclosed and assembled instruments.

Thermal Desorption and Direct Thermal Extraction were performed using the AutoDesorb™ model 2000 automated thermal desorption system (SIS Inc. Ringoes, NJ) for analysis. Separation and detection were performed with and Agilent (Palo Alto, CA) 6890 GC and 5973 MSD in EI mode. Thermal desorption, GC and MS parameters are given in Table 1.

Results and Discussion:

Figure 2

The polymer materials were analyzed by DTE, with sample sizes of 4.8 mg, 5.0 mg, and 7.4 mg for the boot, gasket, and overlay respectively. The analysis was intended to be qualitative, so no standard was added to the desorption tubes. Results are presented in terms of area response per milligram of sample to relate approximate relative emission rates. Four cyclic siloxane oligomers were consistently seen in the DTE samples. They are listed as compound #s 1,4,6,and 11 in Table 2. Spectra are presented in Figure 6. BHT was also found in all of the samples, and area counts for this compound are also reported. BHT levels did not appear to correlate with siloxane emissions in this small sample, however this may be expected as the samples were of different materials. Results are presented in Figure 2.

Figure 3

Figure 3 shows a blank run by the P&T method and Figure 4 is the Total Ion Chromatogram from the LCD sample. Results of the whole-instrument and LCD analyses are presented in Figure 5. Interestingly, the potted sensor unit displayed a higher amount of siloxane material in the sample than the unpotted unit. This discrepancy may be due to random variation in silicone materials used in the instruments, or it may have been the result of an unnoticed leak in the sampling device. The relative amounts of silicone compounds and BHT, however, remain similar between the samples. The LCD yielded the most volatile material of any of the samples. In addition to greater quantity, there was also a greater variety, as the LCD emitted several siloxane compounds that were not detected in the other samples. This is probably due to the efficiency of the heated sampling chamber relative to the glass P&T vessel, as well as the fact that the LCD possessed the largest pool of silicone of any component in the sensor. Relative amounts of BHT did not correlate with the whole-instrument samples. Because standards were not used, the results are merely qualitative, however the gross differences between samples and the large amount of siloxane materials found in the LCD provide evidence that it is this part that contributes most of the contaminating material. The manufacturer of the detector can use this information verify that silicone-free specifications are being met by suppliers.

Figure 4

Quantitative and semi-quantitative analysis is possible with these methods with only minimal extra work. Once a response factor can be established for the siloxanes with respect to an internal standard, the DTE samples can be spiked with the internal standard for quantitative analysis. Likewise, the P&T samples may be spiked either during sampling by placing a standard in with the sample as it is being collected. Alternatively, the adsorbent tube may be spiked after sampling is complete to provide a known level of standard on column. The major difficulty in establishing quantitative methodology for silicone materials by GC is that most of the capillary columns that yield good peak shapes for these substances are themselves manufactured with silicon bonded phases. Several steps can be taken to minimize the background contribution from the GC system. First , an alternative to silicone-based inlet septa should be in place. The AutoDesorb™ system works well with the Merlin Microseal inlet device, provided care is taken in installation, and injection speed is not too fast. Additionally, any bonded phase should be kept away from the hot inlet. A suitable length of deactivated fused silica guard column should be used, and whole-oven cooling should be avoided. Blanks should be run frequently to assure consistently low levels of siloxane materials from the GC inlet and column.

Figure 5

Table 2. - Siloxane Compounds from Components of Handheld Hydrocarbon Sensor.

* Compounds 7 and 9 are discrete peaks, but present spectra that are identical to compounds 6 and 8 respectively. These may be structural isomers, or possibly evidence of sample overload. They are detected only in the LCD sample.

Figure 6. Representative Spectra of Common Siloxanes

Conclusion:

A simple, sensitive method is described to determine the source and extent of volatile off-gassing of silicone and other volatile substances from small electronic devices or components. The method can be expanded with minor changes to provide quantitative results suitable for process monitoring, quality control, and setting engineering specifications. Successful execution depends on control of background silicone artifacts from the GC.