Recent developments in NR curing systems.

Legislation enacted in Germany over the last four years concerning the level of certain N-nitrosamines in the workplace has had a major impact on the choice of curing systems for both NR and synthetic polymers. This impact is reflected in the volume of literature and patents from research and development groups worldwide.

The concern engendered extends from Germany to manufacturers and suppliers wishing to export rubber products and raw materials to Germany and to responsible companies, worldwide, concerned with health and safety in the workplace.

This article discusses the current status concerning compounding and goes on to show how such concerns have stimulated the application of sophisticated techniques to the problem. This in turn is producing new information which is improving the understanding of the vulcanization process.

A list of the chemical abbreviations used in this article is given in table 1.

[Tabular Data Omitted]

N-nitrosamine generation and avoidance

Nitrosating species can originate from some elastomers, from fillers and from the atmosphere. Their reactions with secondary amines can take place both in the rubber and following amine diffusion, in the atmosphere. Stevenson and Virdi (ref. 1) have reviewed the toxicological aspects of this problem. Loadman (ref. 2) has shown that NR is virtually clean as an elastomer in both solid and latex forms. Both white and black fillers can serve as sources of nitrosating agents during cure (refs. 3-6), following adsorption of oxides of nitrogen during their production or in their storage. Chemical treatment of fillers, e.g. N220 black, by ammonia or amines has been claimed to reduced N-nitrosomorpholine generation to 3% (ref. 7).

Twelve N-nitrosamines are included in the German MAK list (ref. 8) giving atmospheric limit values for chemical substances in the workplace. Rubber additives containing secondary amines which, when decomposed, can be sources of the designated N-nitrosamines (refs. 9 and 10) are listed in table 2.

Table 2 - additives which may produce N-nitrosamines
in rubber compounding when
decomposed
Chemical class     MBS, DIBS, OTOS, OTTBS
Sulfenamides       MTD, TMTM, TETD, MPTD, DPTT
Thiurams           ZDMC, ZDEC, ZDBC, Z5MC, ZEPC,
Dithiocarbamates   ZMPC, TeDMC, CuDMC
Sulphur donors     DTDM, MBSS

The compounder has three choices in meeting these regulations:

* Continue to use formulations containing these additives when atmospheric levels meet the regulations.

* Use an additive which will reduce the levels of relevant N-nitrosamines, to meet the regulations.

* Change the formulation to eliminate additives derived from secondary amines.

The provision of adequate ventilation will play a significant part in the first and second approaches, if only in moving the N-nitrosamines elsewhere.

Several additives have now been proposed to inhibit N-nitrosamines generation in rubber. Whether compounders use this approach will depend partly on attitudes towards the level of N-nitrosamines found in relation to regulations and partly to whether a system not containing secondary amines can satisfy the cost and performance requirements for the final product.

Lheureux (ref. 11) et al. carried out a major study on the inhibition of the formation of methylphenylnitrosamine in rubber by considering products which could:

* Block or reduce nitrogen oxide species;

* Degrade the nitrosamine;

* Block the methylphenylamine.

Polyoxymethylene was the most active product found, reacting with the amine via its aldehyde groups. However, while a 96% reduction in N-nitrosamine was found in the rubber, no significant decrease was observed in the atmosphere surrounding the press. This was attributed to reaction of the amine with oxides of nitrogen in the atmosphere. These workers, as several others (ref. 2), failed to find any reductions in N-nitrosamines using ascorbic acid.

Table 3 lists products claimed to produce reductions in N-nitrosamine levels, when added to formulations.

[Tabular Data Omitted]

Significant variations in the efficacies of these products have been observed. This approach has restrictions in that it does not eliminate N-nitrosamines generated in the atmosphere, as evidenced by Lheureux's work, the additives can affect the properties of the formulation and can be expensive.

The only sure approach to eliminating the N-nitrosamine problem is to eliminate either the nitrosating species or to use additives which do not contain the secondary amines which are precursors of the listed hazardous N-nitrosamines. The selection of appropriate products is determined by the extent to which processing safety, cure characteristics and ultimate properties meet cost and performance criteria.

As yet there are no products which are direct substitutes for TMTD, ZDC and DTDM, three important chemicals for the rubber goods market in particular. Compromises must be made since most of the newer products developed contribute by activating slower curing systems based on the established "safe amine" range.

Three of the established products which have increased in usage are DPG, as an activator for thiazoles and sulfenamides, ZBPD for activation and reversion resistance in NR and CTP as a means of adjusting processing safety. The more important products developed in recent years are zinc dibenzyldithiocarbamate (ref. 15), dibenzylthiuramdisulfide (ref. 16), diaminediiso-cyanato zinc (ref. 17), xanthogenpolysulfide (ref. 18) and N-t-butyl-2-benzothiazylsulfenamide (ref. 19).

The first four products are used essentially in formulations requiring short processing safety and fast cure, whereas the latter (TBSI), gives long delay and slow cure.

Streit (ref. 20) and Lupfert (ref. 21) have both referred to problems in obtaining N-nitrosamine free semi-EV and EV curing systems. In such cases the choice for medium to longer cure time needs lies in using either DTDC or, more frequently, a low sulfur-high accelerator combination. A set of starting point formulations directed at the rubber goods manufacturer is available from Monsanto (ref. 22).

The compounder wishing to change his current formulation to one which is N-nitrosamine-free has therefore a choice from both existing and new products. For such studies the use of experimental design (ref. 23) and of the desirability concept (ref. 24) is often imperative in order to cope with the variables of product and loading involved. Additionally, there will be a lower level of familiarity for any new products involved. However, the use of these techniques can bring solutions to a problem through optimization, which have escaped the traditional approach.

While rubber fabricators were expecting a stream of new products from additive manufactures to test, when legislation was imminent, certain producers have given some of the reasons why this did not happen. For example, Davies et al (ref. 8) pointed out that the potential new sulfenamides TABS and TOBS were unlikely to be commercialized because of inadequate amine availability. Engels (ref. 9) reported adverse toxicological results for the decomposition products of an attractive thiophosphoryl replacement for TMTD.

In this situation as outlined, a key factor in the research necessary to both continue the manufacture of rubber products and to develop new additives is the ability to analyze vapor emission from hot rubber compounds, as a means of more fully understanding the reactions of the additives involved.

Developments in analytical techniques

Environmental concerns have prompted a number of studies into the nature of volatile emissions during vulcanization. These vary from the specific analysis of nitrosamines adsorbed by an air sampler cartridge, as used by Chaser (ref. 4), to more complete analysis using the RAPRA gas transfer mold (ref. 25), headspace analysis from open molds (ref. 26) and the use of sintered glass plates as a pressure pad allowing gases to permeate through (ref. 27). These workers used GC or GC/MS techniques to analyze the fractions collected.

Aarts and Davies (ref. 28) used isotachophoresis and GC to examine samples of condensate and vapor trapped from an air stream flowing over a standard NR formulation heated at 160-170 [degrees] C. For curing systems based on 2.5 phr sulfur plus either 0.6 phr CBS or TBBS, they found 2.5-4% and 10-13%, respectively, of the theoretically free amine available for release. This was in general agreement with earlier studies showing that a proportion of the amine was bound into the polymer backbone (ref. 29). However, the studies only accounted for about 10% of the weight loss in the compounds and they suspected that water was a major component.

With A. Gergely they went on to demonstrate a new approach whereby the fumes were led directly into a mass spectrometer.

These preliminary results showed, for a conventional S/CBS cure in NR, that water was a major component of the vapor, reaching its maximum at a time corresponding to the onset of cure. Cyclohexylamine emission was essentially complete by the end of the cross linking process. With a S/MBTS/DPG system, minor amounts of aniline arising from the DPG were observed.

The attraction of this technique is in identifying species as they are being produced, i.e. in "real" time and to give an insight into the mechanism of vulcanization reactions, not possible with extractive techniques.

This potential is demonstrated using three examples.

Application of mass spectrometry to compound vapor analysis

Experimental

A rubber sample of about 100 mg is heated in a sealed sample chamber in a stream of helium carrier gas. Evolved volatiles are continuously sampled and passed to a VG 7070 EQ mass spectrometer ion source, where they undergo electron impact ionization. The mass spectrometer scans the total mass range every three seconds and the data are recorded and stored. Subsequently it is then possible to produce from these data a total ion trace and each individual ion chromatogram as a function of time.

Compounds used were based on a masterbatch of 100, SMR L; 50, N 330 black; 4, ZnO; 2, stearic acid. No oil was included in order to reduce the number of peaks in the mass spectrum. 6PPD, when used, was included in the masterbatch. The attribution of selected peaks used is given in table 4.

Table 4 - attribution of
peaks in mass
spectra
 m/z    Product
value   Water
  18    DMA
  45    TBA
  72    [CS.sub.2]
  76    Morpholine
  87    CHA
  99    TMT
 132    BTH
 135    MBT
 167    TBBS
 238    Sulphur ([S.sub.8])
 256    6PPD

Basis of interpretation of mass spectra

Individual peaks in a mass spectrum are produce by the ionization of a molecule or of its fragments and represent a particular mass charge, m/z. A fragment pattern is characteristic for a given molecule and involves both intensity and m/z values; however the same value of m/z can arise from different structures. The intensity of a given peak is basically determined by the ease of ionization and the stability of the resulting ion and not by concentration per se.

In order to decide whether a particular ion originates from a particular product present in the vapor or by fragmentation it is necessary to examine:

* The mass spectrum of a pure standard;

* The mass spectrum of a mixture with other ingredients of the mix;

* A comparison of the intensity/time curve with that of other fragments present.

At this stage the results should be treated qualitatively but there is very good agreement between duplicate runs in the position and timing of maxima and inflections. For the same species it is possible to make a general comparison.

Results

Figure 1 shows the MS spectrum measured after about 17 minutes at 140 [degrees] C for the NR masterbatch containing 2.5 phr sulphur, 0.5 phr TBBS and 2 phr 6PPD. A number of the m/z values have been characterized to indicate both the power of the technique and the complexity of the analysis.

Figure 2 shows the intensity/time traces for sulfur, TBBS and 6PPD for the compound shown in figure 1. The three appear as the temperature of 140 [degrees] C is reached. 6PPD shows a rise and a plateau, as expected from a product present in the bulk. TBBS shows a very sharp maximum which has disappeared by the estimated rheometer [t.sub.S1] value (allowing for sample heating).

Sulfur reaches it peak at a time corresponding to the [t.sub.S1] value and then drops steadily, showing an inflection in the second half of the period of vulcanization. Similar observations were made previously by Wise (ref. 30) and by Sullivan (ref. 31) and this inflection coincides with the point at which polysulfides reach their maximum concentration.

The trace for water emission, (figure 3), shows a significant increase from the time of heating, reaching a maximum in the early part of the cure curve. The secondary maximum reported by Davies (ref. 23) is also apparent. Water can arise from moisture in the rubber and filler as well as from the zinc oxide/stearic acid reaction. This curve gives emphasis to the potential role of this component in the initial stages of vulcanization. Our studies suggest that the second peak is associated with the zinc oxide/stearic acid reaction.

Figure 4 shows the development of some of the reaction products during cure, BTH appearing first. Sullivan (refs. 31 and 32) et al. in their recent papers utilizing HPLC chromatography to study vulcanization reactions, showed that BTH was produced at an early stage and attributed it to the possible hydrolysis of the sulfenamide by water present. It is noticeable that MBT, TBA, [CS.sub.2] and sulfur all show a second peak or inflection towards the end of the vulcanization period and in the same time zone as the second water peak.

The data produced stimulated a study of amine generation in an EV system containing three different additives, CBS, DTDM, TMTD (1:1:1), both with and without 2 phr 6PPD in the masterbatch.

Figure 5 shows the evolution of free amine for the compound held at 120 [degrees] C, a temperature chosen to approach processing conditions. Superimposed on these curves is the cure curve measured on the moving die rheometer for the same temperature profile as for the sample in the mass spectrometer.

The MS technique can differentiate between the amines in their times of maximum emission. It is apparent that the morpholine curve shows an inflection and a maximum, i.e., two stages are involved. The inflection corresponds to the maximum observed for cyclohexylamine, ex CBS, and the maximum to that of dimethylamine from TMTD. The second peak in the morpholine curve corresponds to the second half of the cure curve. One might have expected a reversal in the order of emission of CHA and DMA, since TMTD is a more reactive vulcanizing additive than the sulfenamide, CBS. However, this could reflect the easier scission of the amine in thiocarbonyl sulfenamides as compared with the amine in a thiuram (ref. 4).

Figure 6 shows some of the reaction products. The curve for water shows a steep peak at the beginning of the scorch period, with evidence of a second maximum in the region of the DMA maximum. BTH appears early with MBT in the first part of the cure curve. TMT and [CS.sub.2] both show peaks towards the end of the vulcanization time, with an inflection apparent on the [CS.sub.2] curve. This is in the same time period as the morpholine maximum. In the presence of 6PPD, amine emissions all occur earlier, consistent with faster vulcanization reactions.

Sullivan et al. showed significant difference between TBBS and TBSI for conventional, semi-EV and hi-lo Ev cures.

Figure 7 compares MBT emissions for conventional NR stocks based on 2.5 phr sulfur and 0.5 and 0.55 phr, respectively, of TBBS and TBSI with hi-lo EV stocks based on 0.5 phr sulfur and 5.0 and 4.25 phr, respectively, TBBS and TBSI. The loadings were chosen to give equal modulus. All contained 2.0 phr 6PPD.

Curves for MBT emission from the conventional compounds are very similar. The EV show major differences in that the MBT arising from TBBS is nearly 30 times more intense. Both show a second maximum, but differing in relative intensities, as compared with the first. Both EV cure curves show inflections in the early part of cure, that with TBSI being more dominant. However, in this work the MBT is appearing sooner than is apparent in HPLC studies. We are currently examining whether such results are related to the different techniques used to remove the MBT or are a reflection of the state in MBT is present in the compound.

The evolution of BTH, (figure 8), shows subtle differences from that reported by Sullivan, but the accelerator/sulfur ratios are slightly higher in this case. There is little difference between conventional cures, but in the EV a significant peak is observed for the TBBS and one at a third of that intensity from TBSI which, in turn, is about five times greater than for the conventional systems. Sullivan found negligible amounts of BTH in his EV system.

These results would indicate that the combination of mass spectrometry with extractive techniques could be used as a means of obtaining a greater understanding of the vulcanization process.

Conclusions

This article has examined the choices open to the compounder in the light of the latest developments concerning the N-nitrosamine issue. Essentially these are:

* Rely on adequate ventilation;

* Reduce the level of N-nitrosamine generated;

* Change the compound.

If the compound remains unchanged then adequate ventilation covering the whole of the processing line is a prerequisite.

Additives which inhibit N-nitrosamine formation pose a number of problems and are most likely a short term solution as their efficiency is not complete in terms of the environment surrounding the rubber.

The replacement of potentially hazardous additives based on secondary amines is the safest solution, but maintaining processing, cure and physical properties may be difficult. It is advisable to make use of the latest statistical approaches for such studies. This in turn can lead to benefits which the traditional approach to compounding can overlook.

Using sophisticated MS analytical techniques, it is now possible to monitor the emission of amines and other volatiles on a "real" time basis. This is particularly relevant to the conditions occurring along a processing line, where the scorch delay in a compound is being used up as the additives present react.

The results of this work using the MS technique show that it is possible to highlight aspects of the curing process which are not identified using other approaches, such as the monitoring of water and complement earlier techniques such as HPLC.

From such studies we would anticipate not only a better insight into what is happening in production but also advances in our understanding of vulcanization mechanisms.

References

[1.] A. Stevenson and R.S. Virdi, Elastomerics, 123 (6) 22 (1991). [2.] M.J.R. Loadman, Kautsch. Gummi, Kunstst. 42 201 (1989). [3.] R.H. Schuster, F. Nabholz and M. Gmunder, Kautsch. Gummi, Kunstst. 43 95 (1990). [4.] D.W. Chaser and P.D. Matheny, Kautsch. Gummi, Kunstst. 44 435 (1991). [5.] R. Freise, J. Konig and K. Gammann, Kautsch. Gummi, Kunstst. 44 429 (1991). [6.] D.W. Chasar and P.D. Matheny, Kautsch. Gummi, Kuntstst. 46 462 (1993). [7.] German Patent, Dl 4012797 A1 Oct. 1991. [8.] Technische Regal fur Gefahstoffe, "Nitrosamine" 552, BAr Bl. (September 1988). [9.] K.M. Davies, D. G. Lloyd and A. Orband, Kautsch. Gummi, Kuntstst. 42 120 (1989). [10.] H.W. Engels, Kautsch. Gummi, Kunstst. 43 992 (1990). [11.] M. Lheureux, Th. Kuhlmann and V. Siekermann, Kautsch. Gummi, Kunstst. 43 107 (1990). [12.] H.J. Graf, Kautsch. Gummi, Kunstst. 46 486 (1993). [13.] M. Abele and H.W. Engels, Kautsch. Gummi Kunstst. 45 638 (1992). [14.] D.W. Chaser, Kautsch. Gummi, Kunstst. 45 18 (1992). [15.] Robinson Brothers Ltd., Rhein Chemie Rheinau GmbH. [16.] Akzo Chemicals S.a.r.l., Rhein Chemie Rheinau GmbH. [17.] Rhein Chemie Rheinau GmbH. [18.] Robinson Brothers Ltd. [19.] Monsanto Chem. Co. [20.] G. Streit, Kautsch. Gummi, Kunstst. 42 289 (1989). [21.] S. Lupfert, Kautsch. Gummi, Kunstst. 42 16 (1989). [22.] Curing systems in typical IRP formulations, Monsanto Europe S.A. [23.] TR 70. Experimental design. an introduction. Monsanto Europe S.A. [24.] TR 67A. Compound optimization using statistical experimental design and the desirability concept. Monsanto Europe S.A. [25.] K.G. Ashness, G. Lawson, R.E. Wetton and B.G. Willoughby, Plastics Rubber Proc. Appl. 4 69 (1984). [26.] A.J. Aarts, G.B. Benson, N.L. Dunchateau and K.M. Davies. Intern J. Environ. Anal. Chem. 38 38 (1990). [27.] R.H. Shuster, H. Linde and G. Wunsch, Kautsch. Gummi, Kunstst. 44 222 (1991). [28.] A.J. Aarts and K.M. Davies, "The fate of rubber chemicals in use and their contribution to environmental issues," paper presented at Rubbercon 92, Brighton, U.K., 15-19 June, 1992. [29.] A.J. Aarts, K.M. Baker and K.M. Davies, Kautsch. Gummi, Kunstst. 37 439 (1984). [30.] R.H. Campbell and R.W. Wise, Rubber Chem. Technol. 37, 635 (1964). [31.] A.B. Sullivan, C.J. Hann and G.H. Kuhls, "Vulcanization chemistry-fate of elemental sulfur and accelerator during scorch delay as studied by modern HPLC," Rubber Div. ACS Mtg., Toronto, Canada, May 21-24, 1991. [32.] C.J. Hann, A.B. Sullivan, B.C. Host and G.H. Kuhls, "Vulcanization chemistry - comparison of the new accelerator TBSI with TBBS," Rubber Div. ACS Mtg., Detroit, U.S.A. Oct. 8-11, 1991.