Direct non-destructive characterization of tire materials by spectroscopy.

Identifying ingredients in cured elastomer compounds is an important task of analytical chemists in the tire and rubber industries. A wide variety of instrumental techniques such as infrared and ultraviolet absorption spectroscopy, mass spectrometry, nuclear magnetic resonance spectroscopy thermal analysis, and gel permeation and size exclusion chromatography are useful tools in rubber analysis (ref. 1). Of particular interest is the nondestructive characterization and the direct instrumental analysis of rubber compounds and rubber products since chemical and physical analyses are difficult and samples may only be available in limited amounts. Several surface characterization techniques have been successfully applied to directly study rubber compound surfaces without the need for extensive sample preparation such as microtoming, grinding and/or extracting with solvents. Infrared spectroscopy (refs. 2 and 3) and mass spectrometry (refs. 4 and 5) are two of the most widely used techniques.

Proton induced x-ray emission has been used sparingly to determine the trace elements present in organic polymers (ref. 6), but has been applied to study rubber compounds (refs. 7 and 8). Interest in the direct characterization of rubber ingredients using non-destructive instrumental methods that require little sample preparation led to the present investigation. Photoacoustical Fourier transform infrared (PA-FTIR) spectroscopy and proton induced x-ray emission (PIXE) spectrometry are used to characterize tread lugs sectioned from two worn off-the-road tires from similar service which displayed significantly different performance in that chipping/chunking was visible in one tire brand. PA-FTIR and PIXE information bases were established by characterizing various rubber ingredients, such as elastomers, fillers and curatives, and model tire compounds prepared with varying ingredients at different levels.

Experimental Materials

Natural rubber (SMR 5), cis-l,4-polyisoprene (Natsyn 2200), polybutadiene (Budene 1203), silica (Hi-Sil 233), clay (Natka 1200), talc (Mistron Vapor), zinc oxide and sulfur were used as received. All rubber compounds were prepared by mixing in two stages in an internal mixer, milling into sheets and press curing at 150[degrees]C for approximately 20 minutes. Table 1 is a summary of the elastomers and the filler levels used in the compounds studied; all compounds contained a processing oil, stearic acid, zinc oxide, accelerator and sulfur.

Model OTR tire tread compounds were prepared by mixing all ingredients and press curing milled sheets at 135[degrees]C for approximately 30 minutes. Formulas are shown in table 2.

Instrumentation

Fourier transform infrared (FTIR) spectra were obtained with a [N.sub.2] purged Mattson Cygnus 25 spectrometer ungraded with a Rev. 8 board and First software. The photoacoustic detector was a helium purged MTEC Model 200. The sample cup was 2 mm x 10 mm. The instrument was operated at a mirror velocity of O. 115 cm/s, a resolution of 8 [cm.sup.-l] and using 64 scans. Infrared spectra shown are smoothed and baseline corrected. Proton induced x-ray emission spectra were obtained from Element Analysis Corp., Tallahassee, FL. Thermal gravimetric analyses were preformed using a Du Pont 9900 thermal analyzer programmed for a temperature rise of 5[degrees]C per minute from room temperature to 800[degrees]C and using a 60 [cm.sup.3] per minute nitrogen gas flow rate, switching to air at 600[degrees]C.

Results and discussion

Infrared spectra

PA-FTIR spectra of two areas (each side of a visible ply interface) of off-the-toad tire treads manufactured by two different companies (tread #1 and tread #2) are shown in figures 1 and 2, respectively. Analysis of the spectra of OTR tire tread #1 (figure 1) indicates that these two infrared spectra are similar to one another in that distinct infrared bands are present at 834, 1107, 1378, ca. 1450, 2856, ca. 2923 and ca. 2956 [cm.sup-1] in both spectra. The spectra of OTR tire tread #2 (figure 2) afford very similar spectra to one another with infrared bands present at 571, 590, ca. 837, ca. 1376, 1447, 1667, 2856, ca. 2958 [cm.sup.-1]. Spectra of treads #1 and #2 differ in that tread #1 has a very prominent 1107 [cm.sup.-1] band not present in spectra of tread #2.

PA-FTIR spectra for natural rubber (cis-1,4-polyisoprene) and polybutadiene (cis-1,4-polybutadiene), figure 3, show that several individual infrared bands of NR and BR have the same vibrational frequencies, but that each polymer has a distinct infrared spectrum that permits its identification. For example, when a compound is prepared containing 28% carbon black in a blend of 33.7% BR and 22.5% NR (60/40 blend), compound #1, infrared bands of both polymers are present, figure 4. The presence of cis1,4-polybutadiene is specifically indicated by the diagnostic band at 733 cm-1 along with the band at 3010 [cm.sup.-1]. Cis1,4-polyisoprene is easily identified by the vibrational bands at 841 [cm.sup.-1] and 1378 [cm.sup.-1], neither of which are present in the infrared spectrum of polybutadiene, figure 3. It is calculated that approximately a 10% blend of polybutadiene in cis-polyisoprene could be observed by PA-FTIR based on detection of the characteristic 733 [cm.sup.-1] polybutadiene band. Table 3 is a summary of the infrared band assignments of vibrations in compound #1. Inspection shows that many of the infrared bands for the NR (figure 3) correspond to bands on the infrared spectra of treads #1 and #2 (figures 1 and 2), but not to those of BR.

Table 4 summarizes vibrational frequencies of NR, tread #1 and #2, and cis-1,4-polyisoprene obtained using transmission. i.n. frared. spectroscopy. Note that corresponding band positions of NR obtained via PA-FTIR and transmission infrared are within 4 [cm.sup.-1], and also correspond to bands present in the PA-FTIR spectra of the two tire treads. These data thus indicate that the rubber used to prepare the treads in the two different OTR tire brands has a cis-1,4-polyisoprene structure.

Tread #1 has a strong band at 1107 [cm.sup.-1] not representative of NR/cis-1,4-polyisoprene. This band is attributed to the presence of a Si-O stretching vibration indicating the presence of a non-black filler. PA-FTIR spectra of powdered silica, clay and talc show that silica and clay have vibrational bands near 1100 [cm.sup.-1].

To confirm the polymer structure in the two OTR tire treads, PA-FTIR spectra were obtained for two cured carbon black-filled compounds having a blend of NR and cis-1,4-polyisoprene with different levels of carbon black (ca. 25% and 33%), compounds #2 and #3 respectively (figure 5). Infrared characterization of compounds #2 and #3 is important since: (1) their preparation to obtain PA-FTIR spectra is similar to that required by treads #1 and #2, and (2) the infrared band frequencies and signal-to-noise levels obtained are expected to be similar to treads #1 and #2, since both sample sets are sulfur vulcanized and carbon black filled.

Although higher noise levels are present, the major infrared bands remain readily identifiable including the medium to strong intensity bands listed by ASTM (ref. 10) as being diagnostic of natural rubber: 833, 1370 and 1665 [cm.sup.-1]. Thus, infrared spectroscopy identifies the elastomer structure in both plys of tire treads #1 and #2 as cis-1,4-polyisoprene. However, infrared spectroscopy is not a useful technique in determining whether these elastomer vibrations are from NR or a synthetic cis-1,4-polyisoprene since both polymer microstructures, and therefore molecular vibrational characteristics, are identical. PA-FTIR spectra could be obtained at 38.4% (compound #4), but not at 50% (compound #5) carbon black (ref. 8).

Finally, spectra shown in figure 5 were obtained by direct interrogation of the residual rubber compound adhered to a wire tire cord after extracting from the composite (the rubber coverage) prepared for an adhesion study. This demonstrates the simplicity in preparing cured rubber compound samples for PA-FTIR investigations since it was not necessary to remove the rubber from the matrix prior to analysis.

Proton induced x-ray emission spectra PIXE spectra were obtained by direct analysis of:

* each ply area of OTR tire treads #1 and #2;

* NR and cis-l,4-polyisoprene (Natsyn 2200);

* powdered samples of carbon black, clay, silica and talc.

The elements present in each area of the two OTR treads were identified and quantified. These data were used to determine whether the elastomer type is a natural rubber or a synthetic cis-1,4-polyisoprene and to confirm the identity of the non-black filler type. Figures 6 and 7 are examples of spectra obtained for treads #1 and #2 and figures 8 and 9 are spectra of NR and cis-l,4-polyisoprene, respectively, obtained using energy dispersive analysis of the emitted x-rays. Table 5 is a summary of the most abundant elements present in these treads and polymers that are detectable by PIXE spectroscopy.

PIXE data suggest that the elastomer in both layers of treads #1 and #2 is probably natural rubber since a very high level (1140 ppm) of titanium (4.511 KeV) is present in the synthetic polymer Natsyn 2200 (cis-1,4-polyisoprene; figure 9), probably as a residual of the Ziegler-Natta polymerization catalyst, but not in SMR-5 natural rubber (figure 8) or the two areas of the OTR tire treads from either tire manufacturer, figures 6 and 7. One limitation of PIXE is that the elastomer could still be a synthetic rubber if the catalyst used to polymerize it were a low atomic number element such as lithium. This limitation arises because the x-rays emitted by sodium (which can be detected, but with an intensity dependence) and lower atomic number elements would not be observed using a PIXE instrument equipped with an energy dispersive detector that uses a standard beryllium window since these frequencies are absorbed by the window.

A second application of PIXE spectroscopy to rubber analysis is the detection of relatively high silicon (1.740 KeV) levels (35,600 - 43,380 ppm) in tread #1, but not in tread #2 (1470 - 1690 ppm), indicating that a non-black filler is added to the compound in both layers of tread #1. Direct PIXE analysis of powdered samples of carbon black (N-220), clay, silica (Hi-Sil 233) and talc reveals very high levels of silicon for each non-black filler type, table 4. PIXE results on the fillers show that for carbon black the sample is 99.6% low atomic number elements, presumably carbon, hydrogen and oxygen. Clay has 57.2%, talc has 50.0% and silica has 52.9% low atomic number elements. For each of these three fillers oxygen is a very abundant element. Analysis also shows that clay has aluminum (1.487 KeV) and talc has magnesium (1.254 KeV) concentrations that are the same order of magnitude as are the silicon levels. Since no significant aluminum or magnesium concentrations were detected in tread #1, the non-black filler is thus identified as silica. This result permits confirmation of the assignment of the ca. 110 [cm.sup.-1] infrared band in the PA-FTIR spectra of tread #1 as the Si-O stretching vibration of silica.

PIXE analysis further provides quantitative information on the zinc (8.631 KeV) and sulfur (2.308 KeV) concentrations present in rubber compounds and tire treads (table 5). Were other elements to be present such as a cobalt salt adhesion promoter, they would also be quantified in a single analysis.

One limitation is that PIXE information is only elemental in nature, thus for example, while the presence of sulfur is measured quantitatively, it can originate from a variety of sources such as powdered sulfur ($8), accelerators or impurities. Alternatively, since PIXE results are quantitative, tread #1 appears to consist of two rubber compounds, such as a cushion (tread base) and tread (cap), since silicon levels in the two areas (table 5) vary significantly (4.33% versus 3.56%).

Thermal gravimetric analysis 

Thermal methods were used to determine the NR, carbon black and inert (silica/zinc oxide) contents in each layer of the OTR tire tread lugs by determining the percent-weight loss as a function of temperature and assigning four regions used to calculate the relative percentages of volatile components (oils, antidegradants), elastomer, carbon black and inert species. A rubber compound of known composition (#1) was used as a TGA standard for this determination. Table 7 is a summary.

TGA data indicate that the areas above and below the ply line of OTR tread #2 are very similar in composition, but that two compounds are probably present in OTR tire tread #I.A tread (cap) and a cushion (tread base) compound are probably used in the manufacture of tread #1. Analysis of the data in tables 6 and 7 and using a total formula weight of approximately 170 phr permits calculation of a general formula containing natural rubber, carbon black, silica, zinc oxide and sulfur for each area of both tire treads.

Compounding

The OTR tire tread compounds #1 and #2, whose formulas are shown in table 2, were prepared and selected physical properties tested. Data show that use of silica in an OTR tread compound (#2) increases cut growth resistance (reduces DeMattia flex cracking) by 30% and significantly increases the tear strength of the compound when compared to its corresponding non-silica-filled control compound (#1) at equal hardness and abrasion resistance values. This is consistent with the observed chip/chunk problem visible for tread #2 that is not present in tread #1. Analysis shows that tread #1 contains moderate levels of silica, i.e. 12-15 phr. Since the cure time ([T.sub.90]) of compound #2 is 50% longer than that of the carbon black-filled compound #1, additional compounds were studied. Compound #3 is a carbon black-filled compound having the same total filler volume as compound #2. Compound #4 has the same carbon black and silica content as longer cure time observed for compound #2. Table 8 is a summary of physical property results. Comparison of data for compounds #1, #3 and #4 shows that:

* [T.sub.90] cure times are all similar;

* tear strength is increased by using an additional 13 phr of carbon black;

* tear strength is further increased by using 15 phr of silica;

* cut growth resistance is significantly increased by addition of 15 phr silica, but not upon addition of 13 phr of carbon black.

General considerations The infrared examination of carbon black-filled rubber compounds presented a challenging problem since carbon black is a strong absorber of infrared radiation and also causes extensive scattering of the incident radiation. Many infrared techniques have been tried. Previous work generally dealt with carbon black loadings of 25% or less. In photoacoustic FTIR spectroscopy, infrared radiation modulated by the interferometer of the FTIR impinges a sample in a sealed compartment. Absorbed radiation is re-emitted in the form of heat which causes a modulation on the gas pressure within this compartment. The pressure modulation is an acoustic signal which is detected with a microphone. To obtain pressure modulation frequencies in the range where microphones are most sensitive, interferometer mirror velocities of the order of 0.1 cm/s must be used as compared with 0.3 cm/s for a TGS detector of 1 cm/s for a MCT detector. Soundwaves are used to detect infrared absorption frequencies. It should also be noted that it is the absorbed radiation which is being measured as opposed to the transmitted or reflected radiation in the typical infrared measurement. Photoacoustic FTIR spectroscopy proved to have a number of attractive advantages:

* The rubber sample does not have to be pliable enough to make good contact with an ATR crystal; be microtomed thin enough for transmission; or be ground for use in a pellet or mull. The only sample preparation required is cutting it to fit the sample cup.

* The sample may be of any shape.

* It does not suffer from the Christiansen effect (band distortion from changes in index of refraction in regions of an absorption band) observed when using KBr pellets and mulls.

* It does not require Kubelka-Munk or Kramers-Kronig transformations of the data as do diffuse and specular reflectance infrared spectroscopy, respectively. The main disadvantages are that the signal-to-noise ratio is relatively low and the quantitative aspects are not always straightforward.

Particle induced x-ray emission uses an MeV ion beam (electron or proton) that penetrates the substrate, gradually loses energy and stops at a characteristic depth for the specific particle in a specific material matrix. In surface experiments such as Auger electron spectroscopy or x-ray fluorescence, the energetic particle is an electron. In the present experiments the particle is a proton. The transferred energy causes various reactions that may produce radiation (x-rays) which can be monitored by appropriate detectors. Penetration ranges are approximately one or two to ten microns, dependent upon the matrix material (metal, metal oxide, polymer). PIXE has the same disadvantages of other elemental methods such as x-ray fluorescence (tel. 11), flame atomic absorption and inductively coupled plasma emission spectroscopies. Information obtained by these methods is elemental not chemical in nature, and the low atomic number elements are not detected, generally those elements below sodium in atomic number. In addition, a PIXE set-up is expensive and more difficult to operate. Alternatively, PIXE has the advantages that:

* the sample specimen does not require pretreatment;

* it is a powerful method for analyzing trace elements in organic materials;

* elements are determined with high sensitivity, generally 0.1-0.2 ppm;

* a precision and accuracy of >95% is routinely obtained:

* the technique is generally non-destructive;

* it has true multi-elemental capabilities;

* it can be automated and easily handle on

the order of 100 analyses per hour (tel. 6).

Summary

Both photoacoustic infrared and proton induced X-ray emission spectroscopy proved to be valuable techniques in identifying ingredients in carbon black-filled rubber compounds. This is primarily due to their nondestructive natures and significantly reduced sample preparation requirements. Used in conjunction with one another, concise information was obtained on elements and molecular vibrations present in specific compounding ingredients. Sample preparation for each technique consisted only of cutting the sample to fit the mount.

PA-FTIR has a reported limited use a approximately 15-25% carbon black; however, spectra with up to 38.4% carbon black were obtained. Vibrational information identified the elastomer structure and suggested the nonblack filler type, even at these high levels of carbon black. Spectra were obtained from a compound (33% carbon black) adhered to the surface of a wire tire cord thus showing the flexibility in obtaining useful information without extensive sample preparation. PA-FTIR was used to examine treads from worn OTR treads for two manufacturers, determining that cis-1,4-polyisoprene was the elastomer used in both treads, but that the tire which did not display a chip/chunk problem contained a non-black filler, probably silica. PIXE data were used to establish that the cis-1,4-polyisoprene in each tire tread was natural rubber and to confirm the presence of silica in the tread which did not display a tear problem. Quantitative data on zinc and sulfur levels were also obtained. With PA-FTIR, PIXE and thermal gravimetric data the general tread formula was established for the two tire treads studied: Polymer type and percentage, approximate carbon black level, non-black filler type and level and zinc oxide and sulfur (from all sources) levels. It was determined that the tire that did not display a chip/chunk tear problem was manufactured using two compounds, probably a cushion (tread base) and tread (cap), both of which contain silica at 12-15 phr.

The OTR tire that had a visible tread tearing problem is constructed with two layers of a similar stock, neither of which contains silica. Lab studies of physical properties of model OTR tire tread compounds prepared with/without silica indicate that significant increases in tear strength and decreases in cut growth can be obtained by using 15 phr silica.

References

1. A. Krishen, Anal. Chem., 61, 238R (1989).

2. W. Cooper, F.C.J. Poulton and P.R. Sewell in "Synthetic elastomers," Vol. 12, Encyclopedia of Industrial Chemical Analysis, F.D. Snell and L.S. Ettre, eds., Interscience, 1971.

3. H. Ishida, ed., "Fourier transform infrared characterization of polymers," Plenum Press, New York, 1987.

4. R.P. Latimer and R.E. Harris, Rubber Chem. Technol., 62, 548 (1989).

5. W.H. Waddell in "Applications of analytical techniques to the characterization of materials," D.L. Perry, ed., Piemum Press, New York, 1992.

6. S.A.E. Johansson and J.L. Campbell, "PIXE: A novel technique for elemental analysis," John Wiley & Sons, Chichester, 1988.

7. J.G. Gillick and W.H. Waddell, United States patent 4,824,682, April 25, 1989.

8. W.H. Waddell and J.R. Parker, Rubber Chem. Technol., accepted for publication.

9. "The Aldrich library of FTIR spectra," Edition I, C.J. Pouchert, ed., 1985.

10. ASTM Method D3677-83.

11. R. Klockenkamper, B. Raith, S. Divoux, B. Gonsior, S. Bruggerhoff and E. Jackwert. h, Fresenius Z. Anal. Chem., 326, 105 (1987).