Storage stability of FKM compound based on a bisphenol AF/onium cure system and its potential as a standard reference compound.

Curemeters are the most widely used quality instruments in the robber industry today. ISO 9001, Section 4.11.2, states "... identify all inspection, measuring and test equipment that can affect product quality, and calibrate and adjust them at prescribed intervals, or prior to use, against certified equipment having a known valid relationship to internationally or nationally recognized standards...." (ref. 1). Also QS-9000 Quality System Requirements, Section 4.11.1, states "... establish and maintain documented procedures to control, calibrate and maintain inspection, measuring and test equipment ... to demonstrate the conformance of product to the specified requirements" (ref. 2). For any good quality measurement system, certain characteristics (or sources of variation) should be monitored and corrected if necessary. This is especially important if a statistical quality control program is being implemented (ref. 3). These measurement characteristics are discussed below.

* Accuracy

* Repeatability

* Reproducibility

* Stability

* Linearity

Accuracy for curemeters can primarily be verified with a torque standard traceable to NIST or National Institute for Standards and Technology. On the other hand, to verify the accuracy while testing a robber compound will require a standard reference compound. No nationally recognized standard reference compound has been established yet because rubber compounds in general do not have sufficient storage stability. Generally, robber compound cure properties will change in storage at room temperature. The purpose of this study was to determine if such a stable standard reference compound can be found.

Repeatability is the variation in the measured property using one instrument (curemeter) and one operator. It can be checked with a uniform factory or laboratory robber compound. Repeatability can be checked as part of a gage R&R study or by participating in an ASTM interlaboratory crosscheck. In gage R&R terminology, this variation is referred to as equipment variation (EV) (ref. 4).

Reproducibility has two definitions. In gage R&R terminology, it is the variation in measurements by different operators and is referred to as appraiser variation (AV). With an oscillating disk rheometer (ODR) this variation can be significant. However, with a modern rotorless curemeter such as the MDR 2000 moving die rheometer, the AV is usually small because there is only a negligible effect from different operator techniques (ref. 5). In ASTM terminology, the term "reproducibility" is defined as the ability of multiple instruments to agree with each other (ref. 6). This measurement is made through an interlaboratory cross check with the same reference compound. In these studies, it would be very useful to have a "stable" reference compound. The cure properties of most conventional robber compounds will change even after being exposed only a few days at room temperature. It is not practical to ship a standard reference compound from one laboratory to another in dry ice!

Stability with a curemeter is very important when implementing a quality test program. In routine operations, the cure measurements can change or drift with time. Therefore it is important to periodically monitor the statistical stability of a curemeter through the use of a control chart (ref. 7) with a standard reference compound. Of course, it is very important that the reference compound itself remain stable over time. With conventional reference compounds, the only way to achieve compound stability is to keep it under refrigeration. However, a standard reference compound which remains stable at room temperature is preferred. Using a stable standard reference compound with control charts over time will provide a means of detecting and separating variation from "special causes" (out of calibration, dirty dies, worn seals, die gap out-of-tolerance, etc.) from normal random variation. When a curemeter falls out of "a state of statistical control," then a corrective action must be implemented, such as cleaning the dies, replacing seals, recalibrating, etc.

Linearity is defined as the difference in accuracy over the expected operating range of the instrument. This characteristic can be monitored by standard reference compounds with properties at different ends of the operating range.

Vulcanization chemistry

The effect of stock aging on vulcanization chemistry is presented below for different types of cure systems.

Conventional sulfur cures

Sulfur vulcanization is a chemical reaction which responds to laws of reaction kinetics just as other chemical reactions do. Decreasing the cure temperature causes the reaction rate to decrease exponentially. At room temperature, this reaction is continuing, but at a very slow rate. The reaction rate is dependent on the compound's energy of activation ([E.sub.a]) in Kilojoules per mole (ref. 8). This principle is based on the Arrhenius equation given below in equation 1.

(1) k = [Ae.sup.-Ea/RT]

Where:

A = a preexponential factor;

R = gas constant;

T = absolute temperature;

k = conversion rate.

This explains why the scorch safety and cure times as measured by a curemeter become shorter with aging of stored stock at room temperature. It is also observed that the curemeter maximum torque ([M.sub.H]) will also decrease after a stock is stored at room temperature for a while. This is probably due to a reduction in the cure efficiency from aging which decreases the ultimate crosslink density at optimal cure. For sulfenamide cures, this might be explained from the known decomposition of the sulfenamide accelerator to mercaptobenzothiazole (MBT) and the respective amine (ref. 9).

Peroxide cures

The decomposition rate of a peroxide curative such as dicumyl peroxide and the formation of carbon-carbon crosslinks have a relatively high activation energy (approximately 155 kJ/mol) (ref. 10) compared to a conventional sulfenamide-sulfur cure (approximately 100 kJ/mol or less (refs. 11 and 12).

AF/onium cures

The bisphenol/onium cure system has been used to cure fluorocarbon elastomers since the early 1970s. The crucial ingredients are a bisphenol as the crosslinker, an organo-onium compound as a phase-transfer catalyst, and an inorganic base [e.g. Ca[(OH).sub.2]] as an acid acceptor.

The cure mechanism was described by P. Venkateswarlu (ref. 13) and is described in figure 1. As shown, this cure reaction proceeds via the formation of a double bond in the back bone of the fluorocarbon polymer. The resulting unsaturated polymer can then be crosslinked by the bisphenol. It's the lack of formation of these double bonds at compounding and storage temperatures that makes this cure system so stable and consistent from lot to lot. The cure rate of the compound may change slightly if it absorbs carbon dioxide or moisture from the air. This would then affect the activity of the acid acceptors, which would change the cure rate. This can be prevented if the compound is stored in a sealed polyethylene container. Fully formulated fluorocarbon elastomer compounds containing the bisphenol/onium cure system are known to be stable for years when stored under the proper conditions. It is important to note that greater instability may result if this AF/onium cured FKM compound is stored in refrigeration. This is because the moisture which condenses on the surface of this compound when it is removed from refrigeration will affect the cure. Therefore it is important to store this compound at room temperature for the best stability.

[Figure 1 ILLUSTRATION OMITTED]

Discussion

Initial screening

Initial work was performed by members of the ASTM Task Group of Subcommittee D11.20 to determine relative bin storage stability of compounds based on conventional sulfur, efficient vulcanization, conventional peroxide, high MW peroxide, and peroxide/coagent cure systems. Some of the formulations which were used are given in table 1. Figure 2 shows the percent change in cure times after 10 days of aging at 110 [degrees] F as measured with an MDR 2000 curemeter. None of these compounds showed good storage stability. SBR compounds based on methylol terminated phenol-formaldehyde curing resins were studied. These resins were tested with zinc chloride as well as 5 phr polychloroprene (CR) with 5 phr zinc oxide. Aging studies did show a better storage stability for the resin cured compound containing CR, but these cure systems required higher cure temperatures and displayed marching modulus.

Table 1

Ingredients                        A      B        C       D

SBR1606                          162    162      162     162
6PPD                               2      2        2       2
TMQ                                -      -        -       -
Stearic acid                       2      2        2       2
Zinc oxide                         4      4        4       4
TBBS accelerator                   1    2.5        3       1
Rubber makers sulfur            1.75      -        -       -
Insoluble sulfur (80%)            -     1.5     1.25     0.5
DTDM                              -      -        -        2
Dicumyl peroxide (40%)            -      -        -       -
  on CaC[O.sub.3] carrier
N,N'-m-phenylene                  -      -        -       -
  dimaleimide
2,5-Dimethyl-2,5-bis              -      -        -       -
  (t-butylperoxy)hexane-3
  (45%)

Total                         172.75    174   174.25   173.5

Ingredients                       E      F        G       H

SBR1606                          162    162      162     162
6PPD                               -      -        -       -
TMQ                                1      -        1       1
Stearic acid                       -      -        -       -
Zinc oxide                         -      -        -       -
TBBS accelerator                   -      -        -       -
Rubber makers sulfur               -      -        -       -
Insoluble sulfur (80%)             -      -        -       -
DTDM                               -      -        -       -
Dicumyl peroxide (40%)             4      4        4       -
  on CaC[O.sub.3] carrier
N,N'-m-phenylene                   -      1        2       -
  dimaleimide
2,5-Dimethyl-2,5-bis              -      -        -        4
  (t-butylperoxy)hexane-3
  (45%)

Total                            167    168      169     167

[Figure 2 ILLUSTRATION OMITTED]

Initial testing of FKM compound

In 1995, work began in the ASTM task group to evaluate a new fluoroelastomer compound based on Bisphenol AF/ onium cure system. Also, another compound based on EPDM and a high molecular weight peroxide was evaluated. Tables 2 and 3 give the formulations for these compounds. Figure 3 shows the MDR maximum torque ([M.sub.H]) stability for these compounds after four months of aging based on test results using one instrument. The EPDM compound, while showing significantly better stability than compounds based on conventional cure systems, did show a significant decrease in the [M.sub.H] after four months aging. The fluoroelastomer (FKM) compound, on the other hand, showed basically no change in [M.sub.H] after four months storage.

Table 2 - flouralastomer compound with
bisphenol AF/onium cure

Ingredients                             PHR

FC-2152 fluoroelastomer MB               80
FC-2178 fluoroelastomer MB               20
N550 carbon black                        10
N326 carbon black                         5
Calcium hydroxide                         6
Magnesium oxide                           3
Camaba wax                              0.5
Total                                 124.5
Table 3 - peroxide cure of EPDM compound

Ingredients                                                 PHR

Vistalon 606 EPDM                                           100
R.P. of diphenylamine and diisobutylene                       1
Zinc oxide                                                    5
N550 carbon black                                            65
Paraffinic oil                                                5
Zinc salt of high MW fatty acid                             1.5
Lithene AH                                                  4.0
40% [Alpha] [Alpha]' bis                                      7
  (t-butylperoxy)diisopropyl benzene on
  clay carrier

[Figure 3 ILLUSTRATION OMITTED]

ASTM interlaboratory crosscheck

The most effective way to determine the real changes that occur with a rubber compound over a six month time period is to measure these changes as grand averages of the curemeter parameters from many laboratories. If just one instrument is used to measure small changes in a rubber compound over time, one can never be certain that these changes are not due to random instrument variations. However, when we compare the changes in the mean calculated from many laboratories testing the same compound under approximately the same storage conditions over time, the Central Limit Theorem works to our advantage and enables us to measure real changes in the compound properties over time. (The theorem states that as the population of laboratory test values increases, the distribution of the mean [Mu] for these values will approach a normal distribution and the standard deviation for this mean will be equal to the standard deviation of the individual test values divided by the square root of the number of individual test values n) (ref. 14). In this study, a total of 17 laboratories participated.

Each laboratory participant was instructed to clean the dies, change the die seals and recalibrate the instrument before starting the testing. All tests were performed under the following test conditions on MDR 2000 moving die rheometers:

* 0.5 degrees arc strain;

* 175 [degrees] C temperature setting;

* 12 minute cure time.

It is important to note that all participants were instructed to use the furnished nylon film sheets for all tests.

Nylon 6,6 sheets were used as a barrier film to protect the dies and avoid a "first test" effect afterwards. (Without the film, the first test effect occurs where the fluoroelastomer cleans the die surfaces and affects the surface adhesion with conventional rubber compounds.) This barrier film was used in testing both the fluoroelastomer compound as well as the SBR control compound (with a conventional sulfur cure system).

Each laboratory was instructed to test fluoroelastomer (FKM) standard reference compound and an SBR standard reference compound (given in table 4) on the first and second week after receiving the samples. Then they were to store the remaining samples at room temperature in a furnished, sealed, polyethylene bag and repeat this two weeks of testing three months later and six months later.

Table 4 - SBR standard reference compound

Ingredients                 PHR

SBR 1606                    162
6PPD                         2
Stearic acid                 2
Zinc oxide                   4
TBBS accelerator             1
Rubbermakers sulfur         1.75
Total                      172.75

The SBR compound has been used privately as a standard reference compound for curemeters (both moving die rheometer and oscillating disk rheometer type) for many years. The problem with this SBR compound is that it can be used at best for only a few weeks when stored at room temperature. However, when it is stored at -10 [degrees] C in a freezer, it can be used well over three months (ref. 15).

Figures 4 through 10 compare the changes in the grand averages for each of the curemeter parameters with the aging of both compounds at room temperature. Figure 4 shows that the FKM compound with the bisphenol AF/onium cure system had very little change in maximum S' torque after six months of shelf aging compared to the SBR reference compound which dropped almost two in.-lb, torque units (15% of the initial maximum). The linear regression analysis for the FKM compound shows no significant difference between the original, three-month and six-month test results ([R.sup.2] - 0.63). On the other hand, if it is argued that this regression is significant, it shows a decrease in [M.sub.H] for the FKM compound of only 0.2 in.-lb, in torque after six months of aging (less than one-tenth of the decrease for the SBR compound with the sulfenamide cure).

[Figure 4 ILLUSTRATION OMITTED]

Figures 5 and 6 show the stability of the MDR dynamic property measurements for the FKM compound vs. the SBR compound. As the conventional SBR compound ages, it becomes more hysteretic as seen from the significant increase in the cured viscous (loss) torque S" and cured tan [Delta]. By contrast, the cured S" and tan [Delta] for the FKM compound show practically no increase for the same six month aging period.

[Figures 5-6 ILLUSTRATION OMITTED]

Figures 7 through 10 compare the changes in scorch and cure times for the FKM compound vs. the SBR compound. These figures show the cure times of the FKM compound to be significantly more stable than the SBR compound. However, there is a small significant change detected for the FKM compound with time.

[Figure 7 ILLUSTRATION OMITTED]

The FKM compound appears to have good stability for maximum torque ([M.sub.H]) and cured dynamic properties after six months of room temperature shelf aging; however, its cure times do change with aging. Therefore, the FKM compound should be considered as a good room temperature standard reference for maximum torque ([M.sub.H]) and dynamic properties (S" and tan [Delta]), but probably not for scorch and cure times such as [t.sub.c] 10, [t.sub.c] 50 and [t.sub.c] 90, which do change slowly with bin storage. A batch of the FKM compound can probably be stored at room temperature for up to six months and used as a reference for maximum torque and cured dynamic properties. It has not been determined yet whether this FKM compound can be used for a time period greater than six months.

Figure 11 shows the shape of the fluoroelastomer compound cure curve with the bisphenol AF/onium cure system. An advantage of this compound over a peroxide cured compound is that it has fairly good scorch safety. Its cure curve shape is similar to a sulfenamide cure. Moreover, the bisphenol AF/onium cured fluoroelastomer reaches a very flat plateau with virtually no reversion or marching modulus.

[Figures 8-11 ILLUSTRATION OMITTED]

Other industry reference material standards

Table 5 shows a list of the standard reference materials and industry reference materials (IRMS) currently used in the rubber industry. ASTM D4678 describes the procedures necessary to establish an IRM while ASTM D5900 describes the specification requirements for IRMs (ref. 16). Many of these IRMs are used as standard compounding materials in standard ASTM test recipes to reduce test variation. On the other hand, the IRM 902 and IRM 903 (both oils) are used in ASTM D471 for oil immersion testing.

Table 5 - standard reference materials and industry reference
materials used in the rubber industry

No.              Identification
                 June 15, 1997

SRMs
SRM 383          Mercaptobenzothiazole (MBT)
RM 8384          N-tert-butyl-2-benzothiazolesulfenamide

IRMs
IRM 1a           Tetramethylthiuram disulfide (TMTD)
IRM 2b           Benzothiazyl disulfide (MBTS)
IRM 43a          Naphthenic process oil
IRM 91a          Zinc oxide
IRM 241b         Butyl rubber ML 1+8 @100 [degrees] C 72
IRM 902a         Petroleum oil
IRM 903b         Petroleum oil
IRM 341a         Magnesium oxide

No.              Current source

SRMs
SRM 383          NIST
RM 8384          NIST

IRMs
IRM 1a           Andy Tisler, Exxon
IRM 2b           Akron Rubber Dev. Lab.
IRM 43a          R.E. Carroll, Inc.
IRM 91a          R.E. Carroll, Inc.
IRM 241b         Akron Rubber Dev. Lab.
IRM 902a         R.E. Carroll, Inc.
IRM 903b         R.E. Carroll, Inc.
IRM 341a         C.P. Hall

An ASTM web page on IRMs is available on http://www.astm.org

IRM 241, butyl rubber, is used as a standard reference material to keep the Mooney viscometer as well as the RPA2000 rubber process analyzer in a state of statistical control. Approximately every four years, a new lot of butyl is set aside, checked carefully for uniformity, and used in an ASTM interlaboratory cross-check with at least 20 laboratories to establish an agreed upon reference value for Mooney viscosity. A recent laboratory study showed the IRM 241 butyl to be quite stable after four years (no significant change in viscosity) (ref. 17).

Figure 12 shows an example of a control chart which has been used to keep a Mooney viscometer in a state of statistical control. If a Mooney viscometer is found to be statistically out-of-control (not stable), one should investigate for special causes such as (1) out-of-calibration, (2) dirty dies and/ or rotor, (3) bad seal, (4) change in rotor height, (5) change in air pressure (closing force), etc.

[Figure 12 ILLUSTRATION OMITTED]

IRM 241 butyl rubber is not normally used to check the stability of a moving die rheometer, or an oscillating disk rheometer (ODR) because this material does not generate sufficiently high torque values in these rheometers' normal operating range. (The most common strain for an MDR is [+ or -] 0.5 degrees arc and [+ or -] degree arc for ODRs. An oscillating frequency of 100 cycles per minute is normally applied.) However, figure 13 shows an example of a control chart which has been used to keep a rubber process analyzer in a state of statistical control using this IRM 241 butyl robber. An RPA is quite different from an MDR in that it can generate higher applied strains and higher frequencies which results in higher torque responses. (On the other hand, the standard reference compounds we have been discussing for the MDR can also be used for the RPA, usually at 0.5 degrees arc strain, 100 cpm and 175 [degrees] or 177 [degrees] C cure temperature.) The RPA test configuration used for the data shown in figures 13a, b and c is presented in table 6. If a control chart, using this butyl rubber, detects an RPA stability problem(s), one should look for special causes such as (1) out-of-calibration, (2) worn upper and/or lower die seals, (3) dirty dies, (4) rubber under the seal plate, (5) die gap setting off, (6) etc.

Table 6 - RPA20 configuration COOP1 00

Subtest                   0 ready
              Temp.                               100 [degrees] C

Subtest                   1 timed
              Time                                         2.5m.m
              Temp.                               100 [degrees] C
              Freq.                                        30 cpm
            Strain                                       0.2 deg.
Subtest                   2 frequency sweep
              Temp.                               100 [degrees] C
            Strain                                         1 deg.
Subtest                   3 strain sweep
              Temp.                               100 [degrees] C
              Freq.                                         5 cpm

[Figure 13 ILLUSTRATION OMITTED]

Conclusions

The fluoroelastomer standard reference compound showed much better maximum torque ([M.sub.H]) stability than the SBR standard reference compound or the other compounds evaluated.

The fluoroelastomer standard reference compound showed much better aging stability for dynamic property measurements (S" @ [M.sub.H] and tan [Delta] @ [M.sub.H]) than the SBR compound.

The FKM cure times did change with aging, but not as much as the SBR compound.

Sometimes, the cure times for the FKM compound would slightly increase (lengthen) for a while with aging before starting to decrease (shorten).

References

(1.) ANSI/Q9001-1994 "Quality systems--model for quality assurance in design, development, production, installation and servicing," American National Standards Institute, NY, p. 7.

(2.) Quality System Requirements, QS-9000, Chrysler Corp., Ford Motor Company, General Motors Corp., August, 1994.

(3.) Measurement Systems Analysis Reference Manual, Automotive Industry Action Group (AIAG), 1990, p. 15.

(4.) "Fundamental Statistical Process Control Reference Manual" Automotive Industry Action Group (AIAG), 1991, p. 121.

(5.) J.A. Sezna and J.S. Dick, "The use of rheometers for process control," 140th meeting of the ACS Rubber Division, Fall, 1991.

(6.) D4483, "Standard Practice for Determining Precision for Test Method Standards in the Rubber and Carbon Black Industries," 1997 Annual Book of ASTM Standards, Vol. 09.01, ASTM, West Conshohocken, PA.

(7.) Measurement Systems Analysis Reference Manual, Automotive Industry Action Group (AIAG), 1990, p. 28.

(8.) Kenneth A. Connors, Chemical Kinetics, The Study of Reaction Rates in Solution, VCH Publishers, Inc., NY, 1990, pp. 13-14.

(9.) J.J. Luecken and A.B. Sullivan, "Stability of benzothiazole sulfenamide accelerators," Elastomerics, August, 1981, p. 34.

(10.) DIN 53 529 Part 2, "Testing of Rubber and Elastomers, Measurement of Vulcanization Characteristics (Curometry), Evaluation of Cross-linking Isotherms in Terms of Reaction Kinetics," German Standards (KIN-Normen), p. 9.

(11.) H.G. Burhin, "Testing of cure kinetics - a tool to improve tire quality," presented at Tyretech '90, Brighton, England, November 5, 1990, p. 14.

(12.) John S. Dick and Henry Pawlowski, "Applications for the curemeter maximum cure rate in rubber compound development, process control and cure kinetic studies," Polymer Testing 15 (1996) 207-243.

(13.) P. Venkateswarlu, R.E. Kolb, R.A. Guenthner and T.A. Kestner, "Elucidation of chemical events occurring in the solid phase during the curing of fluoroelastomers with bisphenol AF," presented at the Detroit Meeting of the ACS Rubber Division, October 17-20, 1989 (Paper No. 123).

(14.) T.H. Wonnacott and R.J. Wonnacott, Introductory Statistics, 2nd. Ed., John Wiley & Sons, Inc. 1972, p. 125.

(15.) H. Pawlowski and D.C. Ranson, internal report, December, 1990

(16.) 1996 Annual Book of ASTM Standards, Vol. 9.01, Rubber, Natural and Synthetic--General Test Methods; Carbon Black, ASTM, West Conshohocken, PA.

(17.) John S. Dick, results from interlaboratory crosscheck for Mooney stress relaxation for five different polymers, Nov. 27, 1996, presented at the Dec., 1996 Meeting of ASTM D11 Committee on Rubber, New Orleans.

John S. Dick, Alpha Technologies and Allan T. Worm, Dyneon Corp.