TA techniques for the rubber laboratory.
Thermal analysis has traditionally been in the domain of the research scientist, effectively used in the identification of new materials and the determination of reaction mechanisms. As more sophisticated thermal analysis equipment has become available, utilizing computerized control and data collection, this technology has gained increasing acceptance in both development laboratories and production facilities. The techniques involved can accurately measure basic physical changes with powerful utility. The advantages of these techniques are numerous. For instance, only small material samples are required for testing, the properties that are obtained are not dependent on part geometry, testing can be performed on actual production samples and a quick turnaround time can be achieved.
Thermal analysis involves the determination of a material property as a function of temperature and time. The property as a function of temperature and time. The property is monitored during a controlled temperature program. While the number of thermal analyst techniques is increasing, and new methods are being developed, there are three basic techniques that are most commonly used:
* Differential Scanning Calorimetry (DSC) measures the heat flow to the sample that is required to maintain a temperature equivalent to a reference cell. The physical property measured is enthalpy, and the data obtained indicates phase changes and reactions as a function of temperature.
* Thermogravimetric Analysis (TGA) monitors the mass of a sample during a controlled temperature scan. Changes in mass can be related to material decomposition and the evolution of product gases and volatile components.
* Thermomechanical Analysis (TMA) measures geometric changes in a sample as a function of temperature and/or time. Thermal expansion characteristics and changes in volume caused by phase changes are detected. In addition, material softening can be monitored by probe penetration.
The operating procedures for each technique are specific to the class of material being tested, and procedures specific to elastomeric materials will be outlined here. The thermal analysis setup is built around a Du Pont #2100 controller. The DSC, model #912, is a dual cell unit; the TGA unit used is model #951; the TMA used is model #943. Cooling of the DSC chamber is achieved using an LNCAII liquid nitrogen cooling accessory.
For demonstration purposes, a number of generic polymer formulations have been used to generate the data which is discussed below. This data illustrates the thermal analysis techniques. These generic formulations are given in table 1 and referred to in the text by polymer name.
Table : Table 1 - reference material formulations
EPDM CR EA CSM A B C A B A A B Polymer 100 100 100 100 100 100 100 100 Filler 80 150 100 155 75 80 55 85 Plasticizer 35-65 85 60 45 25 0-20 15 40 Process aids - - - 2.5 - 3 5 5 Curatives 11 20 10 10.5 8 6 9 30 Additional - 6 - 4 2 - - -
Differential scanning calorimetry
In this technique the samples are heated at a constant rate, and the net heat flow to the material sample is measured against a reference. The heat flow is recorded in Watts/gram (W/g). It can be directly related to the specific heat of the material. Also, distinct deviations in the heat flow rate illustrate endothermic or exothermic conditions, which are caused by phase changes or chemical reactions. A sample curve, obtained from a chloroprene (CR) stock, is shown in figure 1. The slight negative slope shows that, over this temperature range, the specific heat of the material is decreasing with increasing temperature. Between -50[degrees]C and -30[degrees]C, however, a sharp increase in the magnitude of the slope indicates a phase change, the glass transition.
A basic DSC is designated as single cell and has two chambers, one for the sample and one for the reference chamber. The model #912 DSC has three chambers. One is a reference, while the other two are available for samples. In this configuration, two material samples can be tested simultaneously. All chambers are heated at the same constant and predetermined rate.
The calibration procedure includes a check of the thermocouples and the sensing circuitry, which is performed using the melting points of two materials - indium for a high temperature mark of 156.6[degrees]C, and mercury for a low temperature mark of -38.8[degrees]C. This calibration should be performed once per week if the equipment is used regularly.
Another important parameter is the cell constant. This constant represents the ratio of the theoretical to the experimental heat of fusion of the chamber itself and is used during the calculations of the heat flow rate. The software packages available with the DSC will automatically calculate the cell constant as part of the calibration routine, and then use this value during subsequent experimentation. The current value is displayed to the operator and is typically between 1.0 and 1.2. If the cell constant increases above these typical values, then more serious problems are possible, and the equipment manufacturer should be consulted. Once per year the equipment manufacturer can provide a more complete calibration check which tests the hardware and the software systems.
Sample preparation is important for consistency of results. The necessary sample size is approximately 10-15 milligrams. A large surface area is beneficial since it will help to maintain an even heating rate through the entire sample, and also allows faster heating. The sample should be as flat as possible on the bottom, since this side will be in contact with the thermocouple and a flat surface will give a more accurate reading. The material sample is placed in an aluminum pan, which is crimpled on all sides. An empty pan is placed in the reference chamber.
At the initiation of the test, the computer triggers the LNCAII computerized cooling assemble; this unit uses liquid nitrogen to reduce the temperature of the chamber to the required starting point. Throughout the test, a nitrogen purge of 50 m1/min. is maintained in the test cell. The entire temperature cycle is computer controlled. Prior to starting the test, the heating cycle is programmed into the computer. This includes start and finish temperatures, and scan rate. The potential temperature range is -180[degrees]C to 725[degrees]C. The programmed start and finish temperatures are chosen depending on the material being used and the transition that is being investigated. A scan rate of 20[degrees]C/min. has been determined to be the most effective for elastomers. A slower rate will produce a curve that shows significant noise in the signal, while a faster rate will not give a clear picture of the transitions that occur. A comparison of curves obtained using different sampling rates on a CR compound are shown in figure 2. The three curves were obtained at 10[degrees]C/min., 20[degrees]C/min. and 50[degrees]C/min. Note that not only will the scan rate affect the resolution of the curve, it also affects the heat flow rate.
For any of the thermal analysis techniques, a real time plot can be displayed by the #2100 controller during the test. If a problem appears, the test can be terminated at any time. The final curve is retained in the computer memory, and the graph can be designed and viewed on the terminal before plotting. Available features include rescaling the graph and zooming in on portions of the curve that are most significant. The curve can also be saved on a disk for later comparisons and for record keeping.
Frequently, the DSC is used to determine the glass transition temperature of a material (Tg), and in this case the temperature cycle is usually started at -100[degrees]C. The temperature is then increased at a rate of 20[degrees]C/min. until the maximum temperature is reached, which for standard tests is typically 50[degrees]C. Since the Tg signifies the transition of the polymer from a glassy solid to an elastic phase, the transition involves a drop in enthalpy and shows up in the DSC curve as an endothermic reaction.
As seen in figure 3, this endothermic condition appears as an increase in the slope of the curve. The available software can determine the Tg. The operator has to identify both baseline portions of the curve on the screen, which indicates the constant power flow before and after the phase transition. Actual determination of the Tg can then be done in a number of ways, as illustrated in figure 4. The recommended method is to determine the inflection point, or the point on the curve with the steepest slope, and identify this point as the Tg. Another option allows the Tg to be determined simply as the midpoint of the curve between the two baseline tangents.
This method is not recommended because it places significant importance on the initial temperature, where the endotherm begins, and the final temperature, where the heat flow again levels out. Both of these points are directly related to many variables, including sample preparation, heating rate and the sampling rate of the equipment. For instance, as the scan rate is increased, the initial temperature is shifted to the right, as can be seen by referring back to figure 2. Such dependence may falsely skew the Tg value. Also, the midpoint calculation does not allow for highly variable slopes in the Tg region.
In a production environment, the DSC analysis is beneficial for quality control. The curves essentially provide the fingerprint of a compound, and any deviations in compound mixture or even raw material content will be clearly visible. Subsequent batches can be checked against a well-established curve. For instance, figure 5 shows four curves obtained from a generic compound of EPDM. The repeatability of the shape of these curves is important. The curve will be dependent on the specific machine used as well as the operating procedures: therefore, a standard curve must be obtained and kept on file. Any subsequent comparisons are made against this control. For instance, a change in polymer grade would shift the Tg, or a change in quantity of plasticizer may affect the slope of the Tg region of the curve. An increased level of plasticizer causes a lowering of the Tg in an ethylene acrylic compound.
If time and resources permit, a series of curves can be determined that illustrate the effects of specific changes in ingredient levels. Most commonly, the operator learns through experience to recognize deviations that are critical and to propose explanations.
In a development laboratory, the DSC can be used to give an indication of the low temperature properties of a compound. The Tg indicates the point at which the material enters its elastic phase. Therefore, the brittleness point of the material must fall at a lower temperature than this temperature. In figure 6, a cure is shown that indicates a Tg of -52[degrees]C. The compound that was tested was found independently to have brittleness point of -55[degrees]C, which is slightly lower than the Tg indicated by the DSC. Although this technique has not been used to actually determine a discrete brittleness temperature, it does provide an accurate and quick screening process to use during new compound development. For instance, a comparison of plasticizers in a compound can be performed and comparative data obtained.
TGA is the measurement of the mass of a sample vs. temperature. The mass measurement can be reported as grams or as percentage mass change. As the temperature increases, volatile components are liberated, and thermal degradation occurs, both resulting in a loss of mass. The TGA curves are unique to each material. TGA can be performed on raw polymers, and as an illustration, a representative curve for an ethylene-propylene polymer (EPDM) is shown in figure 7.
For compounds which are essentially mechanical mixtures, each component has a distinguishable effect on the curve. Figure 8 shows a rather typical compound curve, obtained on a compound of EA. The first mass loss in the curve represents the volatilization of any plasticizer that is present. The next component to degrade is the polymer itself. A percentage mass change will indicate the percent polymer present in the compound. The temperature at which the polymer starts to decompose, the shape of the curve and the temperature at which the curve levels out are all points that are unique to the polymer. The addition of air to the chamber is required to allow oxidation of the carbon black. The curve will level out when all organic ingredients are removed. A small amount of inorganic residue will remain.
The sample holder for this piece of equipment is a platinum "boat" or open container, which is hung at one end of a quartz rod. The quartz rod acts as a balance. Any movement of the opposite end of the rod is monitored, and indicates changes in mass of the sample. The sample is surrounded by a furnace, designed for a controlled heating rate.
Calibration on this equipment is performed approximately once per week during regular use. Calibration of the sensors and software is confirmed by running a sample of calcium oxalate. The resulting curve is well documented and very reproducible. A separate calibration for thermocouples can be performed by hanging a piece of indium on the quartz beam. The indium has a melting point of 156.6 [degrees]C. As the indium reaches its melting point, it will start to liquefy and drop off of the hanger. The resulting data curve should then show a sharp drop off at the appropriate melting point to confirm calibration of the thermocouples.
The sample size required is approximately 10-15 grams. Of the three thermal analysis techniques discussed here, this technique has the least stringent sample requirements. If possible, the sample should have maximum surface area to allow for equal heating rates throughout the sample, but sample shape is not critical.
The sample is placed in the open platinum boat, which hangs in the test chamber. Just prior to placing the sample in this boat, the placement of the quartz beam must be checked to insure proper alignment, and then the balance must be zeroed. General operating procedures are as follows. A nitrogen purge is maintained in the chamber at approximately 50 ml/min. The heating cycle typically ranges from room temperature to 1000 [degrees]C, although the equipment can reach 1200 [degrees]C. When 650 [degrees]C is reached, air is introduced into the chamber, and heating is continued up to 1000 [degrees]C. The introduction of oxygen allows for oxidation of the remaining components of the sample. The recommended scan rate for elastomers is 20 [degrees]C/min.
The TGA procedure has often been used to identify unknown polymers and unknown compounds. By having a record of TGA curves obtained on raw polymers, it is possible to compare curves with these standards and identify the material. General curve characteristics can be seen for each family type of polymer. For instance, an EPDM shows a very clean curve, as does a fluoroelastomer. These two polymers can be differentiated by temperature of decomposition. Typical CSM and CR curve differences include a small dip present for CSM at the start of the polymer region.
As with the DSC, this procedure is an effective means of checking quality control, specifically to check compound consistency. The curve in figure 9 shows the repeatability of a curve from batch to batch of an EPDM compound when mixed under controlled conditions.
Any changes in the ingredients of a compound will affect the curve. This includes factors such as the material source and the amount present. For example, changes in the plasticizer level may have an effect not only on the recognized plasticizer portion of the curve, but also may cause overlap into the polymer portion of the curve.
TGA curves were obtained on an ethylene/acrylic (EA) compound that was mixed deliberately with varying levels of plasticizer. The three curves in figure 10 show that as the amount of plasticizer increases, it becomes more difficult to differentiate between the plasticizer component and the polymer component. After an initial curve is obtained, the volatilization temperature of the plasticizer is determined. Then a second scan can be run with a programmed isotherm at the volatilization temperature to remove the plasticizer. In this case, the volatilization temperature is 325 [degrees]C. The remaining temperature scan can then be completed, and the polymer portion of the curve will be more distinct. The resulting curve will clearly illustrate the isotherm.
When a change in the curve appears, it is difficult to correlate directly to a change in the amount or type of any one ingredient. Often experience with the compound, and then to determine where the problem lies. However, this method can be used to flag a change in the compound, and then combined with other techniques and plant records to determine the cause of the inconsistency.
In order to obtain a complete record of production batches, the most effective practice is to perform both the DSC and TGA techniques. This allows the two techniques to back each other up. Neither one alone will provide a complete picture. For instance, if the DSC indicates an increase in Tg of a compound, the TGA may also indicate a lower than normal level of plasticizer. Both techniques require small samples and a short amount of time to perform.
TMA measures geometric changes in a sample as a function of temperature. The resulting curve will show dimension change in [mu]m versus temperature or time. Unlike the DSC and the TGA, the applications for this technique are varied and operating conditions are often directly related to specific applications. One basic test is to measure the thermal expansion of a material.
The cell required for a controlled heating sequence is similar to those for the other thermal techniques. TMA utilizes a probe which is in direct contact with the sample to measure material dimensional changes. The type of probe varies depending on the expected response of the material and different types include expansion, penetration, macroexpansion and flexural modulus.
The expansion and macroexpansion are used for tests measuring bulk dimensional changes. The penetration probe indicates softening of the material. The expansion probe has a flat surface which rests on the sample. The macroexpansion probe is similarly designed but has a larger surface. The penetration probe has a pointed tip which extends into the sample. The penetration probe, as described, is often used for plastics, which are typically harder materials than rubber.
Problems have been encountered when the tip of the penetration probe has completely immersed in the rubber, and the base of the probe starts to penetrate the sample also. This condition is not accounted for in the supplied software, and results in a variable weight distribution across the sample. Therefore, the penetration probe is not recommended for elastomers.
The expansion probe is often used in place of the penetration probe. For this application, the sample diameter should be larger than the foot of the probe, thereby assuring accurate penetration into the sample. If the measured thickness of the sample decreases at higher temperatures, then softening has occurred and penetration is likely. The most often used probes in the rubber laboratory are the macroexpansion and the expansion.
This unit can be calibrated once per year by the equipment manufacturer. Approximately once per week, a check of the thermocouple should be done in the laboratory, by using the melting point of indium, as discussed with the TGA apparatus.
Sample preparation for the TMA is critical, since the sample geometry itself is measured. Samples are in the shape of a disk. The diameter is not required to be any certain dimension, but for the work described here a 3/8" diameter was convenient because of the availability of a die of that size.
The two sides of the sample should be as flat as possible to insure intimate contact with the probe. If uncured samples are prepared, the stock should be milled carefully before samples are cut. The thickness of the material can vary slightly, since it is measured before the test is run and thickness corrections are provided. However, the thickness should be uniform across the sample itself.
After placing the sample in the chamber, the correct probe is chosen. A load is applied to the probes, usually by weighting the top of the shaft with 2g for the expansion probes and 5g for the penetration probe. The mass can be varied depending on the sample hardness and the extent of dimensional change expected. As the sample is subjected to the heating cycle, an LVDT is used to measure the displacement of the probe from its zero position. This displacement can then be related to thermal expansion or softening of the material.
The TMA unit can supply temperatures from -160 [degrees]C to 1200 [degrees]C. Because this test is often used to simulate conditions which the material will see in service or during processing or handling, the typical temperature scan covers from room temperature to 200 [degrees]C. A nitrogen purge is used at a rate of 50 ml/min. The scan rate will vary considerably depending on the conditions to be simulated. The choice of a rate can be based on the speed at which any physical changes are expected to occur, and also the ambient temperatures expected in actual conditions. For instance, a constant temperature can be held for a length of time to indicate a static processing condition.
TMA will also indicate thermal softening, and often, extensive softening can result from an undercured sample. The extent of softening can be related to the degree of undercure. The curves in figure 11 show a comparison between an uncured and a completely cured CSM compound. The softening in the uncured compound is dramatic. Even less significant levels of undercure can be distinguished. The three curves show approximately 50%, 75% and 100% of total cure. The amount of softening increases as cure time decreases. If this type of trend can be well documented for a single compound, then this technique can be used routinely for an indication of undercure.
Another application for this technique is in the evaluation of sponged material. The process of activation of the blowing agent and subsequent sponge formation can be monitored. Figure 12 shows a series of scans performed on an EPDM compound. The first scan was taken on masterbatch material, the second on masterbatch with curatives and the third scan was taken on the complete compound with the blowing agent incorporated. The rate of expansion can be clearly seen as well as the initial temperature of activation. Studies such as this one can be used in screening blowing agents as well as cure systems. Through experimentation, the optimum scan rate for these studies was found to be 50 [degrees]C/min., which corresponded to heating rates found during cure. This scan rate also gave the most accurate correlations with the total expansion found when the same compound was cured in LCM.
Still another application is in determining the amount of flow that an uncured sample will undergo when subjected to low shear. The sample can be placed under a low loading and placed at the appropriate temperature for an extended period of time. This type of analysis is useful for examining flow that may occur in wrapped hoses during the time prior to autoclave curing, or any process where a formed shape has to sit for any length of time before the curing reaction starts. It also could predict cold flow of mixed stock during storage or transport. Although these results may not correlate directly with physical changes under these conditions, it can be used as a comparative tool during compound development.
Applications for thermal analysis techniques in the development and the production rubber laboratories are increasing in number and scope. The wide range of applications, as well as the ease of operation and quick turnaround possible, have resulted in an increase in the number of thermal analysis units in the rubber industry. Different modules are employed to investigate different parameters, with potential uses in analytical and process related programs. The equipment is accurate, reproducible and quick, although careful attention to the test procedures and calibration is important. Overall, the use of thermal analysis is becoming widespread, and will be an important tool in the rubber laboratory in the future.
PHOTO : Figure 1 - DSC chloroprene "A" formulation
PHOTO : Figure 2 - DSC sampling rates/chloroprene "A" formulation
PHOTO : Figure 3 - DSC Tg calculation/chloroprene "A" formulation
PHOTO : Figure 4 - DSC step transition midpoint options
PHOTO : Figure 5 - DSC repeatability/EPDM "A" formulation
PHOTO : Figure 6 - DSC brittleness temperature/chloroprene "B" formulation
PHOTO : Figure 7 - TGA EPDM polymer
PHOTO : Figure 8 - TGA EA formulation
PHOTO : Figure 9 - TGA repeatability/EPDM "A" formulation
PHOTO : Figure 10 - TGA varying plasticizer levels/EA formulations
PHOTO : Figure 11 - TMA green vs. cured polymer/CSM "A" formulation
PHOTO : Figure 12 - TMA sponge formation/EPDM "B" formulation
[1.] D.W. Brazier, Rubber Chem. Technology, 53, 3 (1980). [2.] P.S. Gill, American Laboratory, January 1984. [3.] Louis Little, Elastomerics 121, 2 (1989). [4.] Wesley Wm Wendlandt, Thermal Analysis, 3rd Edition, John Wiley & Sons, 1986. [5.] Richard L. Zeyen III, presented at a meeting of the Rubber Division, American Chemical Society, Dallas, TX, April 19-22, 1988.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||thermal analysis|
|Date:||Jan 1, 1990|
|Previous Article:||Accident investigation.|
|Next Article:||Stiffness measurement for evaluating heat resistance of rubber compounds.|