Mold fouling during rubber vulcanization.

Mold fouling is a commonly seen phenomenon in the rubber processing industry. During vulcanization, a layer is deposited on the wall of the mold which builds up gradually with each subsequent production cycle. This layer impedes an adequate heat transport, is the source of release problems and results in products with irregular surfaces. Mold fouling makes regular cleaning of the molds necessary. As this costs extra time and money, it is a problem to which a solution is eagerly anticipated.

Factors

The degree of mold fouling depends on various factors. The object of the investigation described in this article was to determine the effect of several of these factors. The selected parameters were the mold, the process and the compound. These parameters determine aspects of a product such as form, material use and properties.

Various earlier studies gave rise to conclusions about the origins of mold fouling. Possible causes have been able to be identified on the basis of these conclusions. Probably these various causes occur concurrently.

* Chemical reactions may occur between the compound and the metallic surface of the mold;

* Low molecular components may be separated from the compound and be deposited on the mold surface;

* Solid components in the compound may adhere to the metallic surface.

In the possibilities mentioned above, mold fouling is caused by components in the compound or reaction products. The main (process) parameter in these cases is the vulcanization temperature, as illustrated by a study performed at the IKV in Aachen, Germany (ref. 1). This study revealed that mold fouling increases exponentially as a function of vulcanization temperature at a particular injection speed, as is shown in figure 1. Mold fouling was also shown to increase linearly with the injection pressure, as is illustrated in figure 2. It was, however, not possible to determine how the roughness affects fouling on the basis of this study.

[FIGURE 1 AND 2 OMITTED]

Preliminary experiments

Prior to the actual experiments, a preliminary test was conducted to establish the nature of the fouling in the molds. In samples obtained from the processors, the type of fouling was determined for different compounds by means of infrared spectroscopy (IR). The results showed that the deposits contained both organic and inorganic components. The samples were subsequently analyzed via rontgen micro analysis (RMA). Primarily zinc, sulfur and carbon were detected, but no reaction products. This may be explained by the fact that the accelerators evaporate at the vulcanization temperature and do not precipitate onto the (hot) mold. This is borne out by a study carded out by Rapra (ref.2) in which vulcanization vapors were extracted and subsequently analyzed by means of gas chromatography and mass spectrometry (GC/MS). The vapors proved chiefly to contain reaction products of accelerators, which therefore leave the mold in gaseous form.

Mold fouling investigation

It was assumed that such factors as the type of steel applied and the roughness of the mold would play an important role. In order to test this in a straightforward and simple manner during every phase of the investigation, a multi-functional mold was constructed (figure 3). Eight different inserts could be used with this mold, each easily being replaceable. The inserts were made of different types of steel, roughness or surface coating.

[FIGURE 3 OMITTED]

The mold was furthermore constructed in such a way to be self-releasing; the cavities were injected centrally. The mold itself was made of chromium steel 17% with inserts of C45 steel, chromium steel 12% and chromium steel 17%. Dependent on the type of steel, the roughness of the inserts varied between 0.1 and 1.6 Ra.

The impact of the process parameters

Two compounds were selected to investigate the impact of the process parameters on the development of the mold fouling during injection molding with rubber:

* A polar nitrile butadiene compound, containing a sulfur vulcanization system (NBR-s, table 1); and

* a non-polar ethylene propylene compound, containing a peroxide curing system (EPDM-p, table 2).

The increase in fouling was visually assessed as a function of the number of cycles at two settings of the injection temperature. The inserts were assessed visually and examined with RMA after the tests were finished.

Fouling first became apparent for both the NBR-s and the EPDM-p compound on the C45 steel after about 20 cycles. There was almost no relation to the setting of the injection temperature. Fouling continued to accumulate in the form of a deposited layer as a function of time. The phenomenon was observed on the higher-grade steels (chromium steel 12% and 17%) only after a greater number of cycles. The roughness of the mold had no perceptible effect on the stage at which fouling occurred, nor on the degree of fouling.

The deposit on the inserts was later examined microscopically (figure 4) and with RMA. The accumulated fouling was clearly visible as grey crystals in the case of the NBR-s compound. Elementary analysis detected primarily zinc and sulfur (figure 5). It was concluded on the basis of the zinc sulfur ratio that the crystals were primarily made of insoluble zinc sulfide. The EPDM-p compound yielded virtually identical results (figure 6), as this compound also contained zinc and sulfur, in addition to peroxide.

[FIGURE 4, 5 AND 6 OMITTED]

The effect on the use of semi-permanent release agents.

During this phase of the investigation, the inserts were coated with various semi-permanent release agents. These were mainly crosslinked polymers based on a polytetrafluoroethylene or polysiloxane with a temperature range up to 250 [degrees] C. The selected agents were all water-based.

In accordance with the supplier's instructions, the semi-permanent release agents were applied to the inserts using an atomizer. A brief curing time was sufficient to obtain an adequate bond with the metallic surface of the inserts. Injection conditions were the same as during the tests before.

The test was expanded to include, next to the NBR and EPDM compounds, compounds based on natural/butadiene rubber (NR/BR), polychloroprene rubber (CR), fluorocarbon rubber (FKM) and silicone rubber (VMQ). The results are provided in table 3 and figure 7. In the table, the number of cycles until the first observation of mold fouling is given. It illustrates that a coating based on polysiloxane yields the best results for low temperature (<100 [degrees] C) compounds such as NBR-, EPDM- and CR. High temperature (>100 [degrees] C) compounds such as FKM and VMQ give the best results, no fouling was visible. This is probably due to the absence of zinc oxide and the lower oxidation sensitivity. Mold fouling was in no way related to the polarity of the compounds tested.

[FIGURE 7 OMITTED]

The RMA of the fouling produced by compounds based on NBR, EPDM and NR/BR pointed towards zinc sulfide crystals (table 4). Elementary analysis of the CR compound revealed that the deposit formed consisted of carbon (50%), zinc chloride, magnesium chloride and ironoxichloride (FeOCI).

Permanent coatings

Coating specialists provided the inserts with a permanent, in most cases metal-based, coated layer of about 5 microns in thickness. Besides synthetic coatings with a thickness of 5 and 50 microns were used. Injection molding was again carried out to a fixed schedule and the release of the products was assessed. The type of fouling analyzed on the metal coatings and the thin synthetic coatings strongly resembled the fouling observed earlier on the uncoated inserts and semipermanent release agents (table 5). Closer study revealed that here again zinc sulfide (zinc oxide) was responsible to a large part for the mold fouling. No fouling was seen on the thick synthetic coating, but the coating was irregular (rough) and partly destroyed due to the high shear stresses at high temperatures.

Follow-up

In addition to this study, a follow-up has been started to investigate the causes of mold fouling. There are several hypotheses on this subject and if the cause is known, it should be possible to solve the problem of mold fouling. Possible study areas are:

* Acquiring a better idea of the reaction mechanism causing zinc sulfide to form a deposit and investigating ways to prevent this;

* investigation of the possibilities of producing high-quality rubbers without using zinc oxide;

* evaluating thicker metal and plastic coatings; and

* mold fouling of peroxide compounds.

Conclusions

The results of the experiments establish zinc sulfide as the major source of mold fouling. The zinc oxide reacts with the sulfur to form the (insoluble) zinc sulfide (inorganic fouling). The zinc sulfide forms condensation centers in which low-molecular weight components such as softeners are deposited (organic fouling). The low-molecular material resinifies and oxidizes to carbon as a function of time.

Mold fouling can be reduced by the use of more high-grades of steel. The roughness did not have a noticeable influence on the degree of fouling.

Another option is to protect the metal by using semi-permanent mold release agents or permanent coatings. Polysiloxane-based release agents in particular are able to prolong standing time. Permanent coatings based on metal alloys are applicable to a limited degree, because the costs are very high and the results are not better than semi-permanent release agents. Plastic coatings with a layer thickness of around 50 microns appear to offer sufficient protection against depositing zinc sulfide, but lack sufficient mechanical strength.

Polymers for low-temperature (<100b [degrees] C) applications such as NR/BR, NBR, CR and EPDM cause more fouling than polymers for high-temperature (>200 [degrees] C) applications such as FKM and VMQ. This is also caused by the presence of zinc oxide in the compound and the sensitivity to oxidation of unsaturated polymers. The polarity of the polymers does not appear to affect the degree of mold fouling.

Table 1 - NBR compound

NBR acn 33%            100 phr
Stearic acid              1
Zinc oxide RS             5
Vulkanox NKF              2
N660 carbon black        50
Kaoline (Dixie clay)     30
DOP                      15
TMTD                      1
TETD                      1
CBS                      1.5
Sulfur (80%)             0.4
Table 2 - EPDM compound

EPDM                100 phr
Stearic acid           1
Zinc oxide UM          5
MBI                    1
N 550 carbon black   67.5
Sunpar 2280           30
Perkadox 14/40 MB      8
Sulfur                0.3
Table 3 - influence of release agent

Number of cycles until first observation of mold fouling

Polymer   Blanco   Siloxane   Fluor

NBR-s      40         200       50
EPDM-p     20         100       10
NR/BR      20         150       10
CR         20         100       10
VMQ       >600       >600     >600
FKM       >600       >600     >600
Table 4 - element analysis balance insert

Polymer   Analysis     Formula     Remark

NBR-s     Zn, S, Ca      ZnS       Crystals
EPDM-p    Zn, S, C       ZnS       Crystals
NR/BR     Zn, S, Si      ZnS       Crystals
CR        Zn, Mg, S   ZnS, ZnC12   Crystals
                        MgC12
          Fe, O, Cl     FeOCl      Pitting
VMQ           -           -           -
FKM           -           -           -
Table 5 - permanent coatings

Number of cycles until first observation of mold fouling

Coating      NBR-s   EPDM-p   NR/BR     CR

Blanco        20       20       40      30
Chromium      20       20       40      40
Nickel        <10      20      <10     <10
Nitrids       50       20      40      40
Teflon 5m     20       20      40      30
Teflon 50m   >600     >600    >600    >600

References

(1.) G. Menges, W. Benfer, "MoM fouling in injection molding of elastomers, " Gummi Asbest Kunststoffe, 36, April 1983.

(2.) B.G. Willoughby, Rubber Fume Ingredient/Emission Relationships, Rapra Technology Ltd, 1994.