index.html_txt_mastonw_cmp.gif (7337 bytes)

CONTRIBUTED ARTICLE

7. A Guide to Identifying Common
Inorganic Fillers and Activators
Using Vibrational Spectroscopy.

Kevin D.O. Jackson

T.A.R.R.C.
Brickendonbury, Hertford
SG13 8NL
United Kingdom

Abstract
LINE_RED_SHORT.GIF (602 bytes)

Vibrational spectroscopy, (FT-IR and FT-Raman spectroscopy) has been used to study a range of inorganic fillers and activators. The paper demonstrates the information available from vibrational spectroscopy and suggests areas in which useful data can be gained.

Three specific examples of the use of FT-Raman spectroscopy in identifying fillers in some elastomeric products are included, as are peak positions (IR and Raman) used to identify the materials present.

LINE_RED_SHORT.GIF (602 bytes)


 Introduction
LINE_BLACK.GIF (1118 bytes)

FT-Raman and FT-IR spectroscopy have both been used to investigate and identify [1-5] a large range of components associated with the rubber industry. In this paper we apply the combined and complementary techniques of FT-Raman and FT-IR spectroscopy to the analysis of inorganic fillers and activators and demonstrate where useful information can be gained from one or both of these techniques.

IR spectroscopy has been extensively used to study gum vulcanizates, one of the earliest examples of this being by Corish [6] in 1960 using microtomed sections of vulcanizate. However it has been much more common to use this technique to identify the polymer, with the filler giving an inconvenient mask to the polymer spectrum, than to use IR to identify the filler.

It has already been demonstrated [7,8] that FT-Raman spectra can be obtained from elastomers reinforced with inorganic fillers and the resulting FT-Raman spectrum used to identify the elastomer. The elastomer identification is possible because the relatively low Raman signals derived from most of the inorganic species used as rubber fillers result in a much lower degree of spectral masking than is the case with IR spectroscopy. However, as the FT-Raman spectra presented here show, where a FT-Raman spectrum is obtainable this can lead to additional information being gained such as crystalline structure or form as well as an identification of the filler.

The FT-Raman experiment is especially useful within a quality control environment, where a quick, reliable answer is required, because of the relatively trivial sampling procedure (the sample is simply placed in the exciting laser path) and the non-destructive nature of the test. This latter characteristic also makes FT-Raman identification useful where individual testing is required to ensure compliance before use (eg medical items).

In some instances neither the FT-IR or the Raman can be reliably utilized alone to give a precise identity to the inorganic fillers and/or activators present in a vulcanizate but the combination of both of these techniques provides valuable additional information on which to base an assignment.

It should be noted that carbon black filled samples cannot be studied by FT-Raman spectroscopy and require specialist reflection IR techniques [9,10] to obtain even the most rudimentary spectra. No carbon black filled samples were examined in this study.

Experimental
LINE_BLACK.GIF (1118 bytes)

All spectra were obtained using a Perkin-Elmer system 2000 combined FT-IR/Raman instrument with gold coated optics. For the FT-Raman experiment a quartz beam splitter and InGaAs detector were used with sample illumination by a diode laser operating at 1.064µm (near infra-red). In order to obtain FT-IR spectra a KBr beam splitter was used together with a TGS detector.

FT-Raman spectra were recorded by direct reflection from the sample surface. For solid samples a clamp was used to hold the sample in position and for powders a cavity was filled with the powder and then placed at the focus of the excitation laser beam. FT-IR spectra were all recorded in the form of Kbr discs of 10mm diameter. Discs were produced by grinding the samples with dried Kbr and then pressing with an effective weight of 10,000 Kg.

25 Scans at 4 cm-1 resolution were used to collect both FT-IR and FT-Raman spectra, unless otherwise noted, requiring less than 5 minutes acquisition time per sample.

All samples were standard materials used in the preparation of commercial products with no purification prior to examination.

Results and Discussion
LINE_BLACK.GIF (1118 bytes)

Zinc Oxide

The most common inorganic additive to be found in any vulcanizate is zinc oxide. The FT-Raman spectrum of zinc oxide is simple with a fairly sharp peak at 441 cm-1 and a much less intense peak at 335 cm-1. Unlike many other additives the commercial material examined herein did not show any signs of an organic material which is often added to commercial products to help prevent removal of these ingredients into extraction systems or to reduce the risk of powder explosions (dampeners). The IR spectrum is also fairly simple with an effective cut off at about 600 cm-1 . The FT-Raman and FT-IR spectra of zinc oxide are shown in Figure 1.

section3.htm_txt_mjfig1_cmp.gif (5253 bytes)

Figure 1. Absorbance FT-IR (a) and FT-Raman
(b) Spectra of Zinc Oxide

Titanium Dioxide

Titanium dioxide is usually added to elastomeric goods as a white colour agent and as a brightener. Titanium dioxide is used in two crystal forms; anatase and rutile. The FT-Raman spectra of these two forms are quite different, the anatase form having peaks at 640, 515 and 397 cm-1, whilst the rutile form has peaks at 610 and 446 cm-1. The FT-IR spectra of the two materials are much more similar to each other, both having a broad peak between 800 and 470 cm-1 but with the anatase form having a valley in this peak centred around 580 cm-1. Figure 2 shows the FT-Raman spectra of an NR/SBR (Natural Rubber/ Styrene Butadiene Rubber) latex foam that has titanium dioxide added and overlayed reference spectra of the anatase and rutile forms of titanium oxide. It can be clearly seen from these spectra that the titanium oxide present is in the rutile form.

section3.htm_txt_mjfig2_cmp.gif (5973 bytes)

Figure 2. FT-Raman Spectra of
(a) NR/SBR Foam Filled with Titanium Dioxide
(b) Rutile Titanium Dioxide and (c) Anatase Titanium Dioxide

Other Oxides

Silica (silicon dioxide) has an extremely weak Raman spectrum, in fact Raman spectra can be collected from samples in glass containers. The IR spectrum is well known [1], with the most intense peak around 1100 cm-1 and is shown in Figure 3 . Variations in the peak positions and the number of peaks can be seen with different forms of silica from amorphous silica to quartz. Amorphous clay, for example, has a very intense, broad peak centred at 1095 cm-1 and an additional much less intense broad peak at 800 cm-1. whilst quartz exhibits an intense broad band around 1100 cm-1 combined with sharper bands at 795, 775 and 690 cm-1.


section3.htm_txt_mjfig4_cmp.gif (4611 bytes)

Figure 3.
  Absorbance FT-IR spectrum of finely ground silica

Occasionally very broad underlying bands are observed in the FT-Raman spectra (increasing background towards lower wavenumbers) of elastomers containing silica, this is indicative of a slightly fluorescent component in the sample mix, other than this there is no indication of the added filler. The FT-IR ATR spectrum is, however, dominated by the silica filler with the 1100 region being particularly affected [7,8].

Magnesium oxide has an IR spectrum very similar to that of zinc oxide, with a broad ‘cut-off’ starting at approximately 700 cm-1. There are, however, additional weak bands at 1476 and 1413 cm-1 which can be used to differentiate between the two materials. Again, the polar nature of the magnesium oxide molecule precludes a Raman spectrum [11].

Commercial antimony oxide (used as a fire retardant and white colour) has both a Raman and IR spectrum. The IR spectrum is dominated by a broad peak centred at 738 cm-1, with additional unresolved peaks at 580, 537 and 508 cm-1. The FT-Raman spectrum has 5 fairly sharp peaks at 716 (small), 453, 375,257 and 192 cm-1, the most intense peaks being at 257 and 192. It is these last two peaks that can be used to determine the presence of antimony oxide in a commercial product. Figure 4 shows a NR based foam with added antimony oxide. The presence of the fire retardant is clearly demonstrated, with the peaks at 257 and 192 cm-1 showing above that of the polymer.

section3.htm_txt_mjfig3_cmp.gif (5840 bytes)

Figure 4. FT-Raman Spectrum of an NR Foam containing
Antimony Oxide Fire Retardant

Carbonates

Calcium, zinc, sodium carbonates and dolomite (calcium magnesium carbonate) are used as additives to elastomer mixes. The most common use for these materials is as a white pigment and as a ‘bulking’ agent. The IR spectrum of calcium carbonate is quite characteristic with a very intense broad band centring at 1430 cm-1. In addition there are sharp bands at 1795, 876 and 712 cm-1. The Raman spectrum of calcium carbonate consists of three main, sharp bands at 1089, 716 and 285 cm-1. The identification of calcium carbonate filler in NR is demonstrated in Figure 5 which shows a FT-Raman spectrum of a vulcanized SMRL (Standard Malaysian Rubber Light) with 40 phr# calcium carbonate and a reference FT-Raman spectrum of calcium carbonate. The three bands listed above can be clearly observed in this spectrum making a positive assignment.

# [In the Rubber Industry parts per hundred by weight is frequently used - abbreviation phr. Editor's Note]

section3.htm_txt_mjfig5_cmp.gif (5539 bytes)
Figure 5. FT-Raman Spectra of
(a) an NR Vulcanizate filled with 40phr Calcium Carbonate
and (b) Calcium Carbonate

Magnesium carbonate gives a Raman spectrum consisting of a single intense line at 1122 cm-1 making Raman spectroscopy an unreliable source of identification alone. Supporting evidence for the presence of magnesium carbonate can be gained from the IR spectrum which is characterized by major absorptions at 1482 and 1420 cm-1 . Additional, lower intensity absorptions at 1120, 886, 853, 803, 719 and 593 cm-1 can be observed from some samples of magnessium carbonate but these are unlikely to be resolved in commercial materials, as shown in Figure 6.

section3.htm_txt_mjfig6_cmp.gif (4428 bytes)
Figure 6. Absorbance FT-IR spectrum of Magnesium Carbonate

The IR spectrum of dolomite is, understandably, very similar to that of calcium carbonate but the bands are even broader#. The most intense peak is centred at 1440-1450 cm-1 whilst the other main bands are at 887 and 734 cm-1. The Raman spectrum is a combination of the peaks found for magnesium and calcium carbonate.
# [Why are the Raman or Infrared bands of a range of carbonates different? After all the vibrators are the 4 atoms of the carbonate ions. We shold explain this in a future edition. Editor's "as awkward" note.]

The Raman spectrum of sodium carbonate has peaks at 1081, 1071 and 701 cm-1. There is also a peak near the cut off point of the filters (150 cm-1) but this is an unreliable source for identification. The IR spectrum can be characterised by a broad absorption at 1431 cm-1 and two sharp bands at 875 and 713 cm-1.

Zinc carbonate has a Raman spectrum unlike the other carbonates examined in that the peaks observed are broad. The peaks occur at 1551, 1375, 1065, 985, 738, 712, 391 and 229 cm-1 the most intense of these being at 1065 cm-1. It is unlikely that any peak other than the one at 1065 cm-1 would be observed in an elastomer mix as the Raman response is very poor for this material. The IR spectrum of zinc carbonate reveals the true nature of the material which is that it is actually a mixture of both the carbonate and the hydroxide and is thus more properly known as basic zinc carbonate or zinc carbonate hydroxide. The IR spectrum has sharp bands in the finger print region at 1046, 951, 834,738 and 708 cm-1. In addition there are two bands in the area associated with the C=O stretch, the 1384 cm-1 band, a position normally associated with inorganic carbonyls and the 1507 cm-1 band associated with the presence of the hydroxide. It should also be noted that the OH stretch centred at 3365 cm-1 is also greatly increased in intensity beyond that of the other carbonates due to the presence of the hydroxide.

Sulphates

The FT-Raman spectra of calcium and barium sulphate are very similar in composition but with the peaks slightly shifted from each other. The FT-Raman spectrum of calcium sulphate has bands at 1137, 1010 (largest band) 673, 622, 497 and 417 cm-1 whilst barium sulphate has bends at 1140, 989 (largest band) 648,618, 463 and 453 cm-1. The FT-Raman spectra of calcium and barium sulphate are shown in Figure 7. The FT-Raman signal from these materials is fairly strong, producing good quality spectra, but practical application of the FT-Raman technique in identifying these materials in elastomers will be difficult for SBR’s as the major bands will be very close to that of the styrene peak (1000 cm-1).

section3.htm_txt_mjfig7_cmp.gif (5282 bytes)
Figure 7. FT-Raman Spectra of (a) Barium Sulphate
and (b) Calcium Sulphate

The IR spectra of calcium and barium sulphate are also fairly similar, with a broad absorption centred around 1150 cm-1 for calcium sulphate and multiple unresolved absorptions with maxima at 1180, 1120 and 1080 cm-1 for barium sulphate. The remainder of the spectrum of calcium sulphate depends on whether the anhydrous or hydrated (2H2O) form is present. The hydrated salt has additional bands at 659 and 600 cm-1 as well as the water bands at 3540, 3390 and 1617 cm-1 whilst the anhydrous form has sharper bands at 670, 609 and 590 cm-1. Barium sulphate shows bands at 982, 633 and 610 cm-1.

Zinc Sulphide

Zinc sulphide is included in this summary as it is a possible vulcanization product rather that a normally added ingredient although it is, of course, present in lithopone. The FT-Raman spectrum of zinc sulphide has sharp peaks at 992 and 352 cm-1 and broader, multiple peaks centred at 645 and 419 cm-1. The IR spectrum of zinc sulphide is limited to a peak at 316 cm-1. This is below the normal range scanned and is therefore rarely detected by IR spectroscopy.


Conclusions
LINE_BLACK.GIF (1118 bytes)
The vibrational spectra of inorganic materials studied in this paper were chosen to represent the more commonly used inorganic fillers, activators and pigments but is not an exhaustive list. The purpose of the paper is to demonstrate the information available from vibrational spectroscopy (both Raman and IR) and to suggest areas in which useful information can be gained.

The three most obvious examples of the use of vibrational spectroscopy in the identification of inorganic elements within a vulcanized elastomer studied herein come from the examination of titanium dioxide, antimony oxide and calcium carbonate. With these three ingredients simply running the FT-Raman spectrum of the whole material would normally be sufficient to demonstrate the presence of these compounds within the vulcanized material. For other components, especially the more polar materials such as silica, the IR spectrum can also be used to identify the inorganic components either by reflectance techniques or by degradation of the sample to leave the inorganic residue.

References
LINE_BLACK.GIF (1118 bytes)

  1. F.K. Scholl, Infra-red Analysis of Polymers, Resins and Additives; An Atlas. Volume II Carl Verlag, München (1972)    
  2. W.C. Wake, B.K. Tidd and M.J.R. Loadman, Analysis of Rubber and Rubber-like Polymers (3rd Edition), Applied Science, London (1983)     
  3. ISO 4650 Rubber Identification, Infra-red Spectrometric Method 1st Ed. (1984)     
  4. G. Ellis, P.J. Hendra, C.H. Jones, K.D.O. Jackson and M.J.R. Loadman, Kaut + Gummi Kunst. 43, 118 (1990)     
  5. P.J. Hendra, C.H. Jones, P.J. Wallen, G. Ellis, B.J. Kip, M van Duin, K.D.O. Jackson and M.J.R. Loadman, Kaut + Gummi Kunst. 45, 910 (1992)     
  6. P.J. Corish Analyt. Chem 31, 1298 (1959)     
  7. K.D.O. Jackson, M.J.R. Loadman, C.H. Jones and G. Ellis, Spectrochim. Acta 46A, 217 (1990)     
  8. K.D.O. Jackson and P.J. Hendra, Spectrochim.Acta 50A, 1987 (1994)     
  9. M. Claybourn, P. Colombel and J. Chalmers, Appl. Spectrosc. 45, 279 (1991)
  10. W.W. Hart, P.C. Painter, J.L. Koenig and M.M. Coleman Appl.Spectrosc. 31, 220 (1977)
  11. C.V. Raman, Nature 21, 619 (1928)

Received 16th June 1998, received in revised format 25th August 1998.
Accepted 26th August 1998

Line_red.gif (1317 bytes)
[IJVS Home]  [Hot Sources]  [Links]  [Bookshelf]  [Contents]
[Publication Details]  [Help]  [Back Issues]  [Feedback]Line_red.gif (1317 bytes)
Copyright © 2004 John Wiley & Sons, Ltd