Determination of the styrene content of styrene-butadiene rubber using TGA.

Quantitative determination of the styrene content of SBR using standard methods requires extensive chemical treatment of the samples. The presence of carbon black, additives and inorganic fillers hampers rapid determination of the styrene content of SBR. Production laboratories routinely use the ASTM D 1416 (ref. 1) test method to measure the styrene content of solvent-extracted rubber. This test method determines the styrene content of SBR by correlating the amount of styrene with the refractive index of the sample. Though the sample-to-sample reproducibility of the technique is reasonably good, the applicability of the method is limited by the fact that the method cannot be used on rubber compounds composed of polymer blends and carbon black, which are normally present in a tire tread formulation.

This project was undertaken to develop a methodology for determining the styrene content of SBR in the presence of emulsifiers, carbon black and other elastomers by thermogravimetric analysis (TGA). The test method presented in this study has the potential to eliminate the use of hazardous organic solvents routinely employed in rubber extraction.

Thermal analysis techniques are often used to characterize elastomers. TGA, which measures weight changes of a relatively small sample (5-50 mg) as a function of temperature or time, is quickly becoming a popular technique for product characterization in the rubber industry. Common rubber formulations consist of a base polymer or a blend of polymers, plasticizers and fillers. As the temperature is increased, the synthetic rubber components begin to volatilize and the sample decomposes. The major rubber components volatilize at different temperatures, which allows for the quantitative determination of the main constituents simultaneously in one short TGA run. Oil-extended polymers have also been analyzed by TGA (refs. 2-5), but pose some problems since the oil decomposition range overlaps the region where the elastomer decomposes. One method used to overcome this problem is vacuum TGA (ref. 6); however, this method is usually accompanied with an isothermal step that lengthens the analysis time. The main components of an unextended SBR can be completely isolated and quantitated using TGA.

In this study, TGA was used to analyze SBR samples with a styrene content ranging from 5 to 60%. The pyrolysis temperature required to obtain a specified % weight loss was monitored as the styrene content of the SBR samples was varied. The effects of emulsifiers, carbon black, curing aids and butadiene microstructure are also presented.

Experimental

Instrumentation

Refractive index (RI) measurements were obtained using a Reichert-Jung Abbe Mark II digital refractometer with a digital readout. A MGW Lauda T-1 ethylene glycol/water bath was used to maintain the temperature of the refractometer. The refractive index of samples with styrene content less than 30% was measured at 25 [degrees] C. For samples with styrene content higher than 30%, measurements were made at 55 [degrees] C and the results were corrected to 25 [degrees] C using equation (1):

(1) [RI.sub.25] = [RI.sub.t] + 0.00037(t-25)

where: [RI.sub.25] = refractive index at 25 [degrees] C; [RI.sub.t] = refractive index at temperature t; and t is the sample temperature in [degrees] C.

SDT 2960 from TA Instruments was used to obtain the TGA curves. The instrument was equipped with an accessory for switching the purge gas from nitrogen to compressed air.

Samples and reagents

Samples of emulsion styrene-butadiene rubber (ESBR) and emulsion butadiene rubber (EBR) were obtained from Ameripol Synpol. Samples of high vinyl (57%) solution-SBR and high cis (98%) BR were obtained from JSR and Bayer, respectively. Polystyrene standard (F-380) was obtained from Tosoh. BR (36% cis, 55% trans and 9% vinyl) was obtained from Scientific Polymer Products. SBS was obtained from Shell Chemical.

Binary blends of SBR/EBR were prepared by dry mixing the appropriate ratios of the elastomers using a laboratory mixer. Black-filled samples were prepared by mill mixing 50 and 75 phr of N234 type carbon black into the rubber.

HPLC grade isopropanol and toluene were obtained from Fisher Scientific. Compressed nitrogen and air were obtained from Air Liquide and were used without further purification.

Experimental procedure

Approximately 1 g of each of the SBR samples was cut into small pieces and extracted twice using a 75/25 (v/v) isopropanol and toluene mixture. The extracted rubber was then dried in an oven. Approximately 0.1 g of the extracted and dried rubber sample was pressed onto aluminum foil using a hydraulic press at 126 [degrees] C/40,000 psi for five minutes. A strip of the aluminum foil containing the sample was cut and placed on the prism of the refractometer. Three RI measurements were made and the styrene content was calculated as described in the ASTM D 1416 test method (ref. 1).

The purge gas ([N.sub.2]) flow rate to the TGA was set at 100 mL/min. To improve the resolution of small variations in the TGA decomposition curves due to changes in the styrene content of the samples, the heating rates were slowed during significant weight changes in the TGA curves (refs. 7 and 8). Table 1 shows the general temperature profile used to pyrolyze the SBR samples. After data acquisition was completed, the oven was quickly heated to 1,200 [degrees] C to remove any residual material from the sample pan. The total analysis time was approximately 1.5 h. Throughout this study, sample weights were kept between 10.5 and 11.5 mg to eliminate any sample size effects.

Table 1 - general TGA temperature profile

Step 1   Ramp 20 [degrees] C/min. to 300 [degrees] C
Step 2   Ramp 5 [degrees] C/min. to 600 [degrees] C
Step 3   Ramp 20 [degrees] C/min. to 700 [degrees] C

Results and discussion

Pyrolysis of polybutadiene has been studied by a number of workers (refs. 9 and 10) and is known to decompose via a two-stage mechanism between 370 [degrees] C and 470 [degrees] C. Microstructural changes in the polymer, which include cistrans isomerization, cyclization, and crosslinking are known to occur in the 200-300 [degrees] C temperature range, before the evolution of volatile pyrolyzates (refs. 9 and 10). Evolution of volatile products ensues above 300 [degrees] C. Numerous gaseous hydrocarbon products have been characterized by GC and Pyrolysis-GC/MS. The major products are 1,3-butadiene and 4-vinylcyclohexene (ref. 11). Above 400 [degrees] C, the cyclized and crosslinked residue undergoes further degradation to produce volatile products (refs. 9 and 10).

Figures 1 and 2 indicate that the first stage in the BR degradation increases with increase in the heating rate. The same figures also show that the major TGA peak (450-500 [degrees] C) in the decomposition of SBR shifted towards lower temperature as the styrene content of the samples was increased. In this study, the styrene content of SBR samples was measured by correlating the magnitude of this shift with the styrene content of the samples.

[GRAPHS OMITTED]

Determination of styrene content

As described above, the styrene content of the SBR samples was determined by monitoring shifts in TGA curves with change in the styrene content of the samples. Figure 3 shows the TGA curves for the standards used in this study. The styrene content of the samples shown in the thermogram decreases from left to right. A linear regression fit of the plot of the temperature at 30%, 50% and 70% weight loss vs. the styrene content of the samples gave a correlation coefficient of [r.sup.2] = 0.98 (figure 4).

[GRAPHS OMITTED]

The effects of emulsifier (rosin acid and/or fatty acid) on the thermograms of the ESBR samples were studied by comparing the TGA curves of the solvent-extracted and unextracted samples. Figure 5 shows that the shift of the slope (plot of the pyrolysis temperature vs. styrene content) of the samples was more pronounced for the 30% weight loss than for the 70% weight loss. The styrene values obtained from the unextracted samples using the 70% weight loss were in good correlation ([r.sup.2] = 0.99) with those obtained from the solvent-extracted samples. This indicates that when determining the styrene content of SBR samples using the TGA decomposition temperature at the 70% weight loss, extraction of the sample prior to analysis may not be necessary.

[GRAPH OMITTED]

Determination of the styrene content of SBR in SBR/BR blends

Since the base polymer of most tread compounds in passenger tires is normally either SBR or a blend of SBR/BR, it was deemed necessary to evaluate the feasibility of using the technique described in this work for blend analysis. To this end, several blends of SBR/BR with different proportions of the two elastomers were prepared using a laboratory mixer. The styrene content of the blends was calculated from the amount of SBR in the blend and the styrene content of the SBR. Table 2 compares the styrene content measured by TGA with the results obtained by RI and the calculated values. The % styrene determined by TGA showed good correlation with the RI as well as with the calculated results ([r.sup.2] = 0.98 for TGA vs. RI and [r.sup.2] = 0.99 for TGA vs. calculated values).

Table 2 - comparison of the styrene content of
SBR samples determined by TGA with the values
determined by RI and calculated styrene levels

Sample           TGA     RI     Calculated

SBR #1          25.79   23.55       --
SBR #2          22.30   23.78       --
SBR #3          37.23   38.29       --
80/20 SBR/EBR   18.98   19.79     19.06
60/40 SBR/EBR   15.31   14.02     14.34
50/50 SBR/EBR   12.80   11.44     11.98
40/60 SBR/EBR   10.52    9.95      9.62

Determination of styrene content of CB-filled SBR samples

Most tire tread formulations contain carbon black as filler. Carbon black has been known to hamper the determination of styrene content of SBR using the well-accepted test methods, such as refractive index and infrared spectroscopy. Though the presence of carbon black shifted the TGA curves, one can estimate the styrene content of carbon black-filled samples using the appropriate calibration curves. Table 3 shows the TGA temperature profile used for the analysis of the carbon black-filled samples.

Table 3 - TGA temperature profile used for carbon
black-filled SBR samples

Step 1   Ramp 20 [degrees] C/min. to 300 [degrees] C
Step 2   Ramp 5 [degrees] C/min. to 600 [degrees] C
Step 3   Ramp 20 [degrees] C/min. to 400 [degrees] C
Step 3   Gas 2 (Air)
Step 4   Keep isotherm for 1 min.
Step 5   Ramp 20 [degrees] C/min. to 800 [degrees] C

The % styrene of the carbon black-filled samples was determined by first calculating the weight of the rubber at a pre-specified weight loss, [W.sub.RSL], using equation (2):

(2) [W.sub.RSL] = C x [W.sub.R] + (%CB x [W.sub.T])/[W.sub.T] x 100%

where C is a factor employed to compensate for the rubber content of the samples at a specific weight loss (e.g., C = 0.30 for 70% weight loss). [W.sub.T] represents the total sample weight, %CB represents the percent carbon black of the sample, and [W.sub.R] is the weight of the rubbery portion of the sample. [W.sub.R] is calculated using equation (3):

(3) [W.sub.R] = [W.sub.T] - (%CB x [W.sub.T])

The styrene content of the black-filled BR is then determined from the temperature that corresponds to [W.sub.RSL] and the calibration curve shown in figure 4. The TGA curves for a carbon black-filled SBR sample are shown in figure 6. For comparison, the styrene content of the carbon black-filled and unfilled SBR samples was determined. The results obtained are shown in table 4. The deviation between the filled and unfilled samples was [is less than or equal to] 1.7%. Based on these results, carbon black does not appear to dramatically affect the accuracy of the test method.

[GRAPH OMITTED]

Table 4 -styrene content of carbon black-filled
SBR determined by TGA

Sample ID   Carbon black (phr)   % styrene

    1               --             25.79
    2               50             24.27
    3               75             27.47

Effect of curing agents on CB-Filled SBR

As stated above, the main concern when selecting an analytical test method for characterizing polymers is the potential interference from rubber additives, such as curing agents, oils, carbon black and other fillers. In order to determine the effects of curing agents and fillers on the TGA curves of SBR, a sample of compounded SBR (per ASTM recipe given in table 5) was analyzed by TGA using the temperature profile shown in table 3. Decomposition of the gum rubber sample was compared to the gum SBR with vulcanizates and the vulcanized SBR. Figure 7 shows that the derivative curves (DTGA) of the cured SBR and the gum rubber with vulcanizates exhibit a significant shift from gum SBR at temperatures lower than 425 [degrees] C. In this study, the weight loss temperature of the cured (442 [degrees] C) and gum with vulcanizates (441 [degrees] C) SBR samples was in a region that showed minimal shift due to the presence of curing agents. Table 6 shows the styrene content calculated from the 70% weight loss temperature for the cured, gum with vulcanizates, and gum SBR samples. The deviations in styrene content are comparable to results obtained for the carbon black-filled gum SBR samples.

[GRAPH OMITTED]

Table 5 - styrene-butadiene compounded by
ASTM recipe

Ingredient      Phr

SBR 1502       100
Sulfur           1.75
Stearic acid     1
Zinc oxide       3
TBBS(*)          1
Carbon black    50
Total          156.75

(*) N-tert-butyl-2-benzothiazolesulfenamide
Table 6 - effect of curing agents on the TGA curves

Sample                        Temperature      % Styrene
                            ([degrees] C)(*)

Gum SBR                          443.00          25.79
Gum SBR with vulcanizates        441.30          29.14
Cured SBR                        442.24

(*) @ 70% weight loss

Effect of butadiene microstructure on TGA decomposition

Since changes in butadiene microstructure are known to affect the decomposition temperature and the composition and amount of the pyrolyzates of BR, it was found necessary to evaluate the effect of changes in diene microstructure on the determination of styrene content of SBR by TGA.

The TGA curves of samples of high (98%), intermediate (36%) and low (~ 13%, ref. 12) cis-BR were studied. In addition, the TGA curves of samples of high (57%) and low (~ 16%, ref. 12) vinyl SBR with the same styrene content were investigated (table 7). The results clearly show that significant changes in the diene microstructure shift the TGA curves, and therefore appropriate standards have to be employed to avoid large errors. If the weight loss temperature used to determine the styrene content is in a region where there is no significant shift of the decomposition curves, small changes in microstructure do not cause significant errors.

Table 7 - effect of butadiene microstructure on the
TGA curves

Sample               % Cis     Temperature ([degrees] C)(*)

  1                   98                  458.58
  2                   36                  456.94
  3                   13                  452.91

         % Vinyl   % Styrene

  4        57         15                  451.31
  5        16         15                  449.83

(*) @ 70% weight loss

Effect of random vs. block copolymers

Commercial SBR is routinely produced as either random or block depending on the end-use. The effects of the styrene distribution on the determination of styrene content by the TGA technique were developed in this study by analyzing SBS block copolymer with a styrene content of 31%. The styrene content of the sample was determined using the temperature that corresponds to the 70% weight loss and the calibration curve generated using SBR samples with a random styrene distribution. The styrene content obtained by this method was significantly lower than the expected value, indicating the need for using standards with the appropriate styrene distribution.

Summary

Tire tread formulations often consist of SBR, curing agents, inorganic fillers and other elastomers, which hamper the styrene determination by the traditional methods. The styrene content of SBR is an important measured parameter as it affects the physical, mechanical and dynamic properties of SBR vulcanizates.

TGA was used to monitor shifts in the decomposition temperature of SBR as a function of styrene content. In this study, a methodology has been developed for determining the styrene content of SBR from the magnitude of shifts in the TGA curves. The test method is quick and does not involve the use of hazardous solvents.

The effects of emulsifiers, carbon black, curing agents and other elastomers on the technique were also investigated. Ways of minimizing the effects of changes in these experimental parameters on the determination of styrene content of SBR were also presented.

Factors that affect the accuracy of the test method include significant changes in styrene distribution (random vs. block) and the microstructure of the diene. In these instances, it was found necessary to use standards with the correct microstructure and styrene distribution.

[GRAPHS OMITTED]

Acknowledgements

"Determination of the styrene content of styrene-butadiene rubber using TGA" is based on a paper given at the May, 2000 Rubber Division meeting.

"Advances and developments in NR" is based on a paper given at the September, 1999 Rubber Division meeting.

"A new grade of natural rubber latex: NC360" is based on a paper given at the September, 1999 Rubber Division meeting.

References

(1.) Annual Book ASTM Stand. 09.01 (1990).

(2.) A.K. Sircar, J. Sci. Ind. Res. 41, 536 (1982).

(3.) S.J. Swarin and A.M. Wims, Rubber Chem. Technol. 47, 1193 (1974).

(4.) J.J. Mauer, "Thermal Analysis: Vol. 1," R.F. Schwender and P.D. Gam (ed.). Academic Press, 1969, p. 373.

(5.) J. Harris, Synthesis 8, 20 (1977).

(6.) I. Groves and L.C. Thomas, Research & Development 133 (February 1988).

(7.) S. Sorensen, J. Therm. Anal. 13, 429 (1978).

(8.) J.M. Criado, Thermochim. Acta. 28, 307 (1979).

(9.) R.P. Lattimer, R.E. Hams, C.K. Rhee and H.R. Schulten, Rubber Chem. Technol. 61, 639 (1988).

(10.) H.R. Schulter, B. Plage and R.P. Lattimer, Rubber Chem. Technol 62, 698 (1989).

(11.) G.N. Ghebremeskel, J.K. Sekinger, J.L. Hoffpauir and C. Hendrix, Rubber Chem. Technol. 69, 874 (1996).

(12.) J.L. Binder, Ind. Eng. Chem. 46, 1727 (1954).