Coatings Clinic: thermogravimetric analysis.

Thermogravimetric analysis (TGA) is a technique for following weight change of materials on heating. The basic instrument is a balance with a pan loaded with the specimen, an oven for heating the specimen, and a thermocouple to measure the temperature. Air or an inert gas flows over the specimen and removes gaseous degradation products, which in turn may be captured and analyzed. The specimen may be heated at a programmed rate (usually linear) or heated to a given temperature and then held there. Weight is measured continuously and plotted versus temperature or time. The relatively smooth set of descending curves in Figure 1 is an example of a percent weight-temperature plot. Because of the complexity of many plots, it may be difficult to pick out events of interest. Therefore, investigators often turn to first derivative plots, (%/[degrees]C or %/min) versus temperature, for help. The plot with the peaks in Figure 1 is an example of a derivative plot. Some industries, such as pharmaceuticals, routinely use such traces to fingerprint materials and compare batches. Derivative plots from coatings can be useful, but they must be interpreted with great care. It is safe to say that relatively large, sharp derivative peaks indicate weight loss rate maxima and side by side peaks probably point to sequential events, but small peaks are tricky. They may be due to small but meaningful events, but are just as likely to be the result of electrical noise, trapped solvent escaping from the specimen, or some heavy footed person walking by the instrument. Just because there are several derivative peaks, do not assume that several reactions or other events occurred.

[FIGURE 1 OMITTED]

TGA may be used to characterize a number of material properties, including coating composition (volatiles, nonvolatiles, and ash), coating and component stability or volatility, amount of retained solvent or moisture, and the stability of conveyor lubricants and other materials that go through ovens along with painted ware. Processes such as drying, baking, curing, and degradation may be monitored and a combination of DSC and TGA can be used to follow the unblocking of blocked crosslinkers and catalysts.

TGA in air is useful for following the evolution of volatiles during baking. A typical experiment will involve a simulated bake with a ramp to the bake temperature. Sometimes weight loss continues beyond the normal bake time, which may mean that reformulation with faster solvents is needed or that a higher bake temperature may be necessary. However, another possibility is that degradation is occurring or supposedly non-volatile components are being lost. For example, I have tested can coatings where the main polymer degraded at or near the bake temperature and automotive coatings where a considerable amount of UV absorber volatilized during the bake. We can test the resistance to degradation or volatilization by doing TGA measurements on individual components as a function of temperature or, in bake simulations, as a function of time.

A major problem with measurements on wet coatings is that it is difficult to prepare good specimens. A 1-2 mil (25-50 [micro]m) coating spread over a TGA pan makes for a very small specimen that produces almost no volatiles to collect for analysis and increases the possibility of weight measurement errors. However, a larger specimen will be too thick for realistic solvent evaporation. A way to get around this is to replace the pan with a small aluminum sheet or foil coupon (1-1.5 cm on a side) with a small hole for hanging from the instrument rod or hanger. This minipanel can be preweighed and tared in the TGA, then coated by virtually any application method and placed back in the instrument. This gives a specimen with a reasonable weight and a correct coating thickness. It works well with any TGA where a specimen can be hung.

A common technique for characterizing polymers or coatings has been to make isothermal measurements at a series of elevated temperatures, calculate the rates of weight loss, then plot log rate versus reciprocal temperature in Kelvin (K). The slope of the resultant Arrhenius plot gives the activation energy for whatever process or reaction is occurring and extrapolation can be used to determine rates at lower temperatures. The latter is useful for estimating coating lifetimes and for making comparisons between formulations. I realize the dangers of extrapolation, but this technique has proved useful, especially for heat resistant coatings such as silicones, fluorocarbons, and silicone alkyds and for the evaluation of the effect of different pigments on heat stability. If more than one process or reaction is going on or if successive events occur, the rate-reciprocal temperature plot is likely to be a curve rather than a line, which complicates things. Thermal and thermal oxidative effects can be separated by doing separate experiments in an inert gas such as nitrogen and in air. There is a newer technique, Modulated TGA, which allows activation energies and rate constants to be obtained from single experiments rather than the multiple-temperature isothermal method.

"Coatings Clinic" is intended to provide a better understanding of the many defects and failures that affect the appearance and performance of coatings. We invite you to send your questions, comments, experiences, and/or photos of coatings defects to Cliff Schoff, c/o "Coatings Clinic," CoatingsTech, 527 Plymouth Rd., Ste. 415, Plymouth Meeting, PA 19462; or email [e-mail].

By Clifford K. Schoff

Schoff Associates