On October 22, 1967 the first bale of Keltan® EPDM was produced in Geleen, the Netherlands. During the last five decades Keltan has continuously realized improvements regarding catalyst, process, product and application technology. Innovation was often achieved in close collaboration with external partners, i.e. customers, institutes and academia. Today, Arlanxeo´s Business Line Keltan has a truly global presence with EPDM plants in Changzhou (China), Geleen (the Netherlands), Orange (USA) and Triunfo (Brazil). Keltan EPDM experts can be found in all major sales regions, giving local, dedicated technical support to their customers. This paper presents a technical overview of the past, the present and the future of Keltan EPDM (Fig. 1).

Fig. 1. Time line of Keltan EPDM: top: start-up of plants and company history;
bottom: technical innovations.
Copyright: all Arlanxeo

Fig. 1. Time line of Keltan EPDM: top: start-up of plants and company history;
bottom: technical innovations.
Copyright: all Arlanxeo

50 years of EPDM history

In the south of the province Limburg in the Netherlands, coal has been mined for centuries. In the 19th century coal mining was industrialized and scaled-up [1], and in 1902 the various coal fields and concessions were eventually concentrated under the flag of the Dutch States Mines (DSM). In the first half of the 20th century, chemical industry was set up in the Geleen area with raw materials coming from (by-) products of coal production, first resulting in cokes and gas but later on also in aromatics, ethylene, hydrogen, ammonia and fertilizers. After the 2nd World War DSM further intensified and diversified their chemical industry, including the building of naphtha crackers and the polymerization of ethylene and propylene to polyethylene (PE) and polypropylene (PP), resp. The last coal mine was closed in 1973.

Fig. 2: left: Structures of most common EPDM third monomers (only one stereo-isomer shown); right: general structure of ethylene/propylene/diene random ter-polymer (EPDM) with ENB as diene.
Copyright: all Arlanxeo

Fig. 2: left: Structures of most common EPDM third monomers (only one stereo-isomer shown); right: general structure of ethylene/propylene/diene random ter-polymer (EPDM) with ENB as diene.
Copyright: all Arlanxeo

In the 1940s DSM worked on the production of its own synthetic rubber, viz. Stamikol®, which was based on ethylene dichloride and sodium polysulfide. In the 1950s DSM explored the production of synthetic isoprene and butadiene rubbers (IR and BR, respectively). By 1961 DSM started a research project to manufacture synthetic EPM rubber and in the early stages obtained a license allowing the use of a Ziegler catalyst. Soon after the start-up of a pilot plant for EPM, DSM started building its first, synthetic EPM rubber plant in Geleen in 1965, which was the first of its kind in Europe, having a design capacity of 12.000 ton/year with a capital investment of 18 mln. dutch guilders (equivalent to 8 mln. €).

In 1966 the brand name of Keltan was created from a list of 10.000 possible names with the aid of a Univac III computer. On October 22nd 1967, the first bale of Keltan EPM rubber was produced. Starting with an EPM copolymer, dicyclopentadiene (DPCD) was soon introduced as a third monomer, quickly followed by 5-ethylidene 2-norbornene (ENB) (Fig. 2). Both DCPD and ENB are asymmetric, non-conjugated dienes with the two unsaturations having different reactivities for insertion polymerization. The norbornene unsaturation in DCPD and ENB is strained and is used for incorporation of the third monomer in the EPDM polymer backbone. The residual unsaturation in the final EPDM product facilitates sulfur vulcanization and enhances peroxide curing.

The initial idea was to replace NR by EPDM in rubber applications, such as tires for cars and bicycles. However it quickly became clear that EP(D)M rubber is not well suited to these applications, so Keltan needed to find other more appropriate applications for this outstanding, synthetic rubber product. Because of the excellent weathering resistance (EPDM is highly resistant to heat, oxygen, ozone, UV and water), opportunities to use EPDM in automotive and building and construction applications were explored and developed. As early as 1968, an EPDM roofing based on K720 (today renamed as K6460D) was applied by one of our earliest customers, i.e. Hertel B.V. in Kampen, carrying the brand name Herlatan®, which was based on a combination of the (brand) names of Hertel and Keltan. Worldwide, this is the oldest (evaluated) single ply EPDM-based roof, and it is still in service today, having been monitored now for several decades [2]. The elongation at break (eab) has decreased with time and the tensile strength has increased, both due to weathering (Fig. 3). However, the current eab is still within specification, i.e. the EPDM roof sheet is still in fair shape.
Over the years the EPDM market has grown tremendously. Today EPDM rubber is one of the largest and most important synthetic elastomers with a total volume of around 1200 kton/yr, making it the largest volume non-tire rubber. Keltan demand has grown accordingly, requiring the building of more production units. In 1973 a second EPDM rubber line, a copy of the first line, was started up in Geleen, and was debottlenecked in 1978 through the introduction of ammonia deep-cooling, increasing the capacity to

Fig. 3: Elongation at break (eab) and tensile strength (TS) over time of a single ply K720
(today renamed as K6460D) EPDM roof sheet implemented in 1968 [2].

Fig. 3: Elongation at break (eab) and tensile strength (TS) over time of a single ply K720
(today renamed as K6460D) EPDM roof sheet implemented in 1968 [2].

30 kton/yr. For several years the first line was mothballed, due to a temporarily lack of sales. The EPDM plants in Geleen have subsequently been further debottlenecked several times by means of the implementation of several process technology improvements, giving significant increases to production throughput and substantial improvements to product quality. An example is the installation of a crumb-buffer vessel in 1984, that decouples the polymerization and the finishing units of the plants and hence, reduces significantly the scrap levels. Further developments saw Keltan succeeding in overcoming the infamous gel formation issues and in­troducing improved catalyst systems. The performance of the first, classical Ziegler-Natta (ZN) catalyst systems (vanadium compounds in combination with aluminum alkyls) were improved via so-called promotors (organochloro compounds), increasing their activity by one order of magnitude, while at the same time modifying the EPDM micro-structure (improved homogeneity) and hence, performance. Controlled long-chain branching was introduced in the late1990s for optimum mixing and processing of EPDM compounds. In the last decade, the Keltan ACE™ advanced catalyst technology was developed in the Geleen research center. In 2003 a third even larger EPDM plant was started up in Geleen and in 2013 this new EPDM plant was converted from ZN to ACE catalysis (Fig. 1).

In 1989 DSM acquired the EPDM plant in Addis (USA) from the Copolymer Rubber and Chemical Corporation; a plant that was later closed in 2004. In 1991 a joint venture was initiated be­tween Idemitsu and DSM, and an EPDM plant was started up in Chiba (Japan), which was closed in 2004. In 1996 DSM acquired the EPDM plant of Nitriflex in Triunfo (Brazil), which is the only EPDM plant in South America and gives Keltan a strong, local position. In 1997 DSM acquired the Sarlink® thermoplastic vulcanizates (TPVs) product line from Polysar, which was a major step towards forward integration in the EPDM technology chain.

In 2011 Keltan EPDM was divested by DSM and acquired by Lanxess. In a way Lanxess continued on the road, which was paved by Bayer A.G., with the invention of synthetic rubber by Fritz Hofmann in 1909, the commer­cialization of Buna® SBR in 1936 and the first production of chlorinated and brominated IIR in 1961 and 1975, respectively. In 2011 Lanxess had a very broad, synthetic rubber portfolio, including not only BR, SBR, IIR, CR, NBR, hNBR and EVM rubbers, but also EPDM rubber. The latter EPDM products with the brand name of Buna EP were produced in two plants, i.e. one plant with two lines in Orange (USA) and one plant in Marl (Germany). The plant in Orange was originally started up by BF Goodrich in 1972, later acquired by Polysar in 1982 and finally acquired by Bayer in 1990. The plant in Marl was started up by Hüls in 1972 and also acquired by Bayer. Soon after the Lanxess acquisition of Keltan, the Buna EP and Keltan and EPDM product portfolios were integrated under the Keltan brand name. In 2011 the Sarlink TPV portfolio was separately divested by DSM to Teknor Apex, giving Keltan more freedom in supplying the TPV market. In 2015 the largest EPDM plant worldwide was started up in Chang­zhou (China) based on ACE catalysis. In 2016 the Marl plant was closed. At the moment, the four Keltan EPDM plants are assets owned by Arlanxeo, a joint venture between Lanxess and Saudi Aramco, founded on April 1, 2016 with its headquarters in Maastricht (the Netherlands). Arlanxeo is a globally leading producer of synthetic rubbers with Keltan being a market leader in EPDM rubber with a market share of over 20 percent.

In summary, the history of Keltan EPDM rubber has been quite hectic, not only from a production process but also from a business viewpoint. When Keltan became part of Lanxess in 2011, the production and application know-how of many different rubber products, such as (S)BR and IIR, became available, allowing synergy and strengthening further innovation. Today we are looking forward to a bright future with Keltan being part of Arlanxeo, partly owned by Saudi Aramco, and with several new technical innovations in the pipeline.

Fig. 4: Simplified, typical lay-out of a traditional EPDM  plant with polymerization in solution combined with steam recovery.

Fig. 4: Simplified, typical lay-out of a traditional EPDM plant with polymerization in solution combined with steam recovery.

Global partnership

Fig. 4 shows a simplified lay-out of a typical, traditional EPDM plant using solution polymerization combined with steam recovery. First, the monomers and the solvent are purified by removing water and other polar species. Cooling of the reactor feed is required to facilitate adiabatic operation of the reactor for the strongly exothermic polymerization reaction, which is initiated by injecting the catalyst system into the homogeneous solution in a continuously stirred tank reactor. After so-called killing of the catalyst, the acidic catalyst residues are removed via washing with water. Next, stabilizer and optionally extender oil are added to the rubber solution. Then, steam is injected into the rubber solution resulting in precipitation of the rubber and evaporation of the solvent, i.e. wet recovery. Any non-reacted monomers and the solvent are recycled. The solid rubber is recovered as crumb by removal of the water and volatile organic components (VOC) in a series of process steps, including filtering, expelling, expanding and drying. Finally, the rubber crumb is pressed into bales and packaged. Typically, the recycling and recovery sections are the largest parts of an EPDM plant.
With the general flow chart of an EPDM plant in mind, one may distin­guish several aspects of EPDM production technology, i.e. ZN vs. (advanced) metallocene catalysis, solution vs. slurry polymerization, single vs. multiple reactor set-up, and wet vs. dry recovery of the final, solid rubber product. The four Keltan EPDM plants differ in technology as shown in Fig. 5. The  largest production line in Geleen and the plant in Changzhou are fully based on proprietary Keltan ACE catalyst technology, whereas the other Keltan EPDM lines/plants exploit various different types of state-of-the-art ZN catalysis, including the Triunfo plant, which was originally based on JSR (formerly Japan Synthetic Rubber) technology. All these facilities produce EPDM rubber in a solution process. The Orange plant is one of the few EPDM slurry (or suspension) processes worldwide, which enables the production of extremely high molecular weight rubber and was originally based on BF Goodrich technology. All the Keltan EPDM plants use wet finishing, i.e. recovery of the EPDM polymer from a solution or slurry via steam stripping of the solvent. This not only results in a high EPDM product quality, but also allows for the production of EPDM products with a (very) high Mooney viscosity and/or (very) high oil extension. In Triunfo a world scale reactive extrusion (REX) line is in operation for the production of down-sheared and maleic anhydride grafted EPM grades. Altogether, this brings Keltan today to a unique position, not only as the EPDM producer with the largest number of EPDM production plants, but also with the most diverse production lines, having the widest range of state-of-the-art technologies available.

Fig. 5: The Keltan EPDM plants on four continents are a valuable asset underlining our global partnership.

Fig. 5: The Keltan EPDM plants on four continents are a valuable asset underlining our global partnership.

High quality products

The EPDM polymer plays a crucial role in achieving, amongst others, an optimum filler dispersion, consistent compound processing, desired surface aesthetics, the best elastic performance of the vulcanizate and low rates of compound waste. Keltan has invested much throughout the years to optimize its EPDM manufacturing processes on all production sites. The target was and still is to produce a very high and consistent quality product within narrow specification limits. Any change in the EPDM plant can have an effect on the product’s quality and consistency, so to control these changes a strict Factor Change procedure is followed (Fig. 6). A multi-disciplinary team reviews any proposed changes and judges the likely consequences for e.g. the customers.

Fig. 6: Keltan factor change procedure

Fig. 6: Keltan factor change procedure

Not only are narrow specifications for the main characteristics of the EPDM product important, i.e. Mooney viscosity, ethylene and diene contents and the amount of extender oil, Keltan is also looking for ways to further define test methods by actively taking part in relevant standardization committees and meetings, such as those of the NEN (dutch organization for standards), the International Organization for Standardization (ISO) and ASTM International. In the past, Keltan has made considerable contributions to the development of standards on EPDM chemical composition and branching (for more details see below).

The formation of gel was for a long time considered an inherent feature of the EPDM manufacturing process, which could lead to high levels of rejected products where surface quality is critical, e.g. in the production of weather-strips. In the late 1980s a strip extrusion test, originally developed by our customer Metzeler in Lindau (Germany), was further developed to determine the gel type (hard, elastic, soft and soft-soft) and the corresponding number of gels in a 20 meter non-vulcanized rubber strip. This test was further aligned with the requirements of our customers in the mid 1990s and has been used as an internal quality check ever since. Two origins of gel formation are distinguished, i.e. i) reactor gels: gels made during the polymerization process and ii) oxidative gels: material that gets stuck in the manufacturing process, i.e. so-called hang-up, and becomes oxidized by exposure to heat, air and humidity. In the early days, reactor gels were infamous for shutting down the EPDM production process , with so-called gel outbreaks occurring after a period of con­tinuous production. The manufacturing process then needed to be stopped to allow the reactor to be cleaned. Certain high molecular weight and high ENB grades were notorious for reactor gel formation. Over the last fifty years much has been invested by Keltan in reactor control and dosing, stirring and filtra­tion. The introduction of ammonia suppresses gel formation, the use of promotors (for more details see below) leads to more homogenous EPDM products and the Keltan ACE technology (see below) yields products that are intrinsically gel-free. However, oxidative gels can still be formed in the plants and therefore, a lot of attention, extensive inspection and very frequent cleaning of equipment is required.

In all Keltan EPDM plants con­tinuous improvement sits high on the agenda after safety. Most obviously, this pertains to products made in the Keltan slurry plant in Orange. Historically, products made in a slurry process were in general more susceptible to gel formation, had problems with the dispersion of extender oil, had more
issues with color and smell, and were prone to product inconsistency. As a result of major investments in the Orange plant a tremendous step forward in quality improvement has recently been made. The use of an improved ZN catalyst system with a new activator, similar to that used in the Geleen plants, has resulted in less reactor fouling and significantly reduced gel counts, lower vanadium catalyst residuals, and improved odor and color. Further investments in hardware and feedstock contributed to even better process control, higher bale consis­tency, improved oil dispersion and higher overall product quality.

Fig. 7: Electrical conductivity on a logarithmic scale versus volume fraction of carbon black for an EPDM radiator
hose [3].

Fig. 7: Electrical conductivity on a logarithmic scale versus volume fraction of carbon black for an EPDM radiator
hose [3].

Technical service

Over the last 50 years the colleagues at Keltan not only developed a high level of understanding about how to effi­ciently polymerize EPDM polymers and how to produce high quality products, but also gained considerable experience of the best ways to apply EPDM in what are sometimes very demanding applications. A good example of this is the work carried out in the late 1990s to develop an optimum EPDM compound for automotive radiator hoses. Confronted with the formation of cracks that developed on the inside surface of the hoses during service, the Technical Service and Application Development group in Geleen decided to investigate the origin of these cracks [3]. It was found that the formation of cracks was due to electro-chemical degradation (ECD) caused by the dif­ference in electrical potential between the engine and the hose, with the coolant inside the hose acting as an electrolyte to create a galvanic coupling (redox) reaction. Failed hoses were collected and examined using electron microscopy, allowing dendrite and striae micro-crack structures to be observed, which were the result of ECD. These micro-cracks can grow in the rubber matrix until the textile reinforcement is reached, which is suscep­tible to degradation by hydrolysis when it comes into contact with the coolant, causing weakness of the hose. To overcome damage by ECD, several compounding approaches were developed, resulting in certain rules that need to be considered. First of all, the electrical conductivity of the compound mainly depends on the structure and the loading level of the carbon black filler (Fig. 7). Secondly, white fillers may be used to reduce conductivity. Finally, in order to test the compounds, Keltan developed an ECD test apparatus as an alternative to the so-called Brabolyzer test [4].

Fig. 8: Generic scheme for alkylation and activation of vanadium catalyst by aluminum alkyl and promotor, respectively.

Fig. 8: Generic scheme for alkylation and activation of vanadium catalyst by aluminum alkyl and promotor, respectively.

A more recent example of the application experience is demonstrated through the ongoing development of EPDM compounds for drinking water applications. Keltan has several EPDM grades and compound recommendations, which are suited for both food contact as well as potable water applications. Globally, approvals for final rubber articles are given by certifying bodies, and when applicable, Keltan submits the information needed for these approvals to the relevant certifying bodies. Keltan is also actively supporting the work done by the Wirtschaftsverband der deutschen Kautschukindustrie (WDK) in Germany in their efforts to qualify ingredients applicable for the new Elastomer Guideline (positive list 1). The positive lists, containing ingredients that are allowed to be used for these types of applications are under severe scrutiny, and only one single accelerator (2-mercaptobenzothiazole: MBT) is currently on positive list 1, leaving rubber compounders no practical choice of ingredients. This trend to stricter regulations can also be seen in France and other European Community member states.
Other recent examples of our application development studies include the development of an EPDM compound with dynamic performance to allow the replacement of NR in engine mounts (see below), “green” compounding and the development of green TPV’s (combining recyclability with bio-materials), both based on Keltan Eco EPDM (see below), the activation of resol cure of (EPDM) rubber with zeolite [5], the use of EPDM as a blend component with polydiene rubbers for enhanced ozone resistance for e.g. tire sidewalls, the compounding for improved UV and heat resistance and the compounding methodology for EPDM articles with low VOC.

To facilitate the exchange of information with the customers, Keltan in­troduced in 2004 an on-line compound database (connect.keltan.com). Currently, this database contains more than 200 relevant starting recipes for EPDM compounds, 400 EPDM product-related documents such as product and safety data sheets, and a large number of scientific and technical studies on all kinds of EPDM-related topics, which have been presented by Keltan at rubber symposia and/or published in scientific and technical journals. Finally, 80 extensive answers on some of the most Frequently Asked Questions in the EPDM industry can be accessed. The information from this database is available in English as well as in Chinese.

Finally, next to a global presence with our four EPDM plants on four different continents, Keltan has EPDM experts in all major sales regions, giving dedicated technical support to their local customers. Technical support, customer service and marketing groups are based in Geleen (the Netherlands) for Europe, Africa and the Middle East, in Pittsburgh (USA) for North America, in Sao Paolo (Brazil) for South America, in Changzhou (China) for Greater China and in Singapore and India for the rest of Asia, Australia and the Pacific. Experimental development work is done in laboratories in Germany, China, Canada and Brazil with all relevant and up-to-date analytical, mixing, processing and testing facilities at the experts’ disposal, enabling fast and effective customer support. In Germany even a twin screw extruder for TPV developments and a complete, continuous vulcanization line for EPDM profile extrusion including sponge production, which is also frequently used by Keltan customers, are available. Altogether, this brings Keltan to a unique position today, not just as a leader in the EPDM market, but also as an EPDM player with a truly global presence combined with a very strong, local customer support.

Sustainable innovation

Vanadium-based Ziegler Natta polymerization:

In order to obtain highly homogeneous EPDM products with good elastic properties, the classical vanadium-based ZN catalysts have been used from the start of the Keltan EPDM production. These catalysts are able to incorporate high amounts of propylene and ENB co-monomers in an homogeneous manner. An aluminum alkyl co-catalyst is required in excess in order to alkylate and activate the vanadium pre-catalyst with vanadium in oxidation state of 3, 4, or 5 (Fig. 8). The low activity of the vanadium catalyst is attributed to, amongst others, the very low content of active sites. This is mainly caused by the reduction of the catalytically active vanadium(III) species by the co-catalyst to inactive, low-valent vanadium(II) species. Especially at temperatures above 60°C, the productivity and Mooney capability of these catalyst systems is too low. The addition of organochloro compounds, so-called promotors, to the vanadium catalyst increases the productivity by up to ten times as well as further enhances the single-site character of the catalyst system, resulting in EPDM polymers with very narrow chemical composition distribution (CCD) and very narrow molecular weight distribution (MWD). These promotors are able to re-oxidize vanadium(II) species to chlorinated vanadium(III) species, which subsequently are re-alkylated by the aluminum alkyl.
Still, the productivity of the catalyst/co-catalyst/promotor system is relatively low and, in general, requires a catalyst removal step in the EPDM (solution) process. Also, the detailed structure of the active catalyst has not been resolved, despite numerous efforts by the scientific world. In the 1980s a new class of promotors has been developed by Keltan [6,7,8]. A special class of halogenated esters was found to be extremely suitable for the production of EPDM. They allowed the fine-tuning of the EPDM molecular structure for optimum elastic performance and the development of controlled long chain branching (see below), which significantly improves the mixing and processing behavior of the rubber.

Fig. 9: Volume resistivity on a logarithmic scale as a measure for carbon black dispersion versus mixing time for i) narrow, linear, ii) very broad, very highly branched and iii) narrow, controlled long chain branched EPDMs with similar Mooney viscosity [9].

Fig. 9: Volume resistivity on a logarithmic scale as a measure for carbon black dispersion versus mixing time for i) narrow, linear, ii) very broad, very highly branched and iii) narrow, controlled long chain branched EPDMs with similar Mooney viscosity [9].

Long-chain branching:

Typically, increasing the molecular weight of the EPDM polymer is beneficial for the ultimate properties and the elasticity of the crosslinked EPDM rubber products, but the corresponding high viscosity limits the mixing and processing of the rubber compounds. In the past, the mixing and processing of EPDM rubber was improved by increasing the MWD for a given Mooney viscosity. This can be achieved by modifying the catalyst and the polymerization conditions or exploiting multiple reactor technology. The disadvantage of (very) broad EPDM products is that they have a relatively low number-averaged molecular weight, which results in inferior solid state performance, i.e. lower strength and elasticity. To by-pass this issue, long-chain branching (LCB) was introduced as a way to improve the balance of mixing and processing vs. solid state properties. For a given polymer Mooney viscosity, increasing LCB results in a decreased compound Mooney viscosity, which is beneficial for extrusion and injection moulding. For EPDM, LCB can be achieved via various routes. Cationic branching occurs for ENB-EPDM, catalyzed by acidic catalyst species. Alternatively, LCB can be achieved via polymerization of unsaturated polymer end groups (macromers) or the residual unsaturation of the third monomer. In the past, Keltan had quite a number of DCPD-EPDMs (particularly suitable for peroxide vulcanization) and mixed ENB/DCPD-EPDMs with (very) high LCB levels in its product portfolio for improved mixing and processing.

Unfortunately, branching of EPDM via DCPD incorporation in combination with a ZN catalyst suitable for DCPD-EPDM is accompanied by a strong broadening of MWD and, thus, somewhat inferior properties. In addition, DCPD is a C10 diene with a higher boiling point compared to the C9 diene ENB. As a result, the amount of residual DCPD in recovered EPDM products is relatively high. In the late 1990’s controlled long-chain branching (CLCB) with only limited MWD broadening was developed. As a result, ENB-EPDM with CLCB truly provides the best balance between mixing and processing vs. solid state performance. Fig. 9 shows the volume resistivity as a measure for the carbon black dispersion for a given EPDM compound formulation for i) a narrow, linear, ii) a very broad, highly branched and iii) a narrow CLCB EPDM polymer, all with similar Mooney viscosities [9]. The very broad, highly LCB polymer is characterized by a very slow carbon black incorporation, although eventually it facilitates the highest level of black dispersion, though at economically unacceptable mixing times. The narrow, linear polymer enables faster mixing, albeit to a much lower level of carbon black dispersion. Finally, from a practical point of view the CLCB polymer gives the best results, i.e. a relatively fast mixing and a high carbon black dispersion. Over the years many Keltan EPDM products in the Geleen plants have been converted to CLCB grades. All new Keltan ACE grades (see below) have CLCB.

Fig. 10: Phase angle δ versus frequency measured at 125°C for EPDMs with increasing degree of long chain branching and concurrent decrease of ∆δ.

Fig. 10: Phase angle δ versus frequency measured at 125°C for EPDMs with increasing degree of long chain branching and concurrent decrease of ∆δ.

Next to the development of LCB and CLCB EPDM products, Keltan has developed new tools to assess the degree of LCB. A more traditional way of quantifying LCB is via the so-called Mooney Stress Relaxation (MSR) slope, which is used as an in-house, process control parameter. First, the Mooney viscosity of an EP(D)M rubber is measured under standard conditions [ML (1+4 @ 125°C], i.e. with 1 min. of conditioning time for the cold rubber sample to heat up in the warmed chamber, followed by 4 min. of measurement time with a rotor speed of 2 rpm. The rotor is then stopped and the relaxation of the torque is followed as a function of time. The logarithm of the torque is finally plotted versus the logarithm of the time, yielding a straight line with a slope representative of the MSR value. Note that MSR decreases with increasing LCB and is not only affected by LCB but also by the molecular weight. A more sophisticated method, applying the so-called Δδ parameter, is used as an in-house, product quality control parameter for CLCB. Δδ is defined as the difference between the phase angles (δ) at frequencies of 10-1 and 102 rad/s, as determined by a frequency sweep with Dynamic Mechanical Spectrometry at 125°C [10]. The Δδ value decreases with an increasing degree of LCB (Fig. 10) and, in contrast to MSR, is independent of the molecular weight. Both the MSR slope and Δδ are single-value parameters for describing LCB. The so-called dilution rheology method as developed by McLeish cs. [11] has been modified and implemented for EP(D)M rubber to provide a better description of LCB [12]. The dilution rheology results can either be converted into a single, convenient parameter, the so-called dilution slope, or alternatively, by combining the dilution rheology results with gel permeation chromatography data, the volume fraction of branched polymer, the averaged molecular weight between branching points and the branching density can be calculated.

Fig. 11: Generic structure of Keltan ACE
titanium κ1-amidinate cyclopentadienyl
catalysts (X = Cl, Me; R = H, alkyl, aryl).

Fig. 11: Generic structure of Keltan ACE titanium κ1-amidinate cyclopentadienyl
catalysts (X = Cl, Me; R = H, alkyl, aryl).

Post-metallocene catalysis:

Classical, vanadium-based ZN catalysts have been the work horse for the production of EPDM for many years. The major drawback of ZN catalysis is the low productivity, requiring de-ashing of the product and the necessity of relatively low reactor temperatures (below 60°C) in order to reach the high molecular weights needed for thermoset rubber applications. Low reactor temperatures require costly deep cooling of the monomer/solvent feed (Fig. 4) as well as the use of more energy to remove unreacted monomers and solvent from the product. The addition of so-called promotors increases the productivity by up to ten times as well as enhances the single-site character of the catalyst system, resulting in EPDM polymers with very narrow CCD and MWD. Nevertheless, the relatively low productivity requires a catalyst removal step within the EPDM production process (Fig. 4). In order to overcome the low reactivity and limited molecular weight capability of conventional and promoted vanadium catalysts at elevated reactor temperatures, metallocene catalysts (metal center sandwiched by two cyclopentadienyl [Cp] ligands) and post-metallocene catalysts (metal center with no or just one Cp ‚ligand) have been developed for polyolefin and EPDM polymerization since the early 1990’s by many groups in both academia and industry. Eventually, (post-)metallocene catalysis was successfully implemented in the industry for the production of polyolefins including EPDM. A decade ago, the development of the so-called Keltan Advanced Catalyst Elastomer technology [13], abbreviated to Keltan ACE, was started, triggered by the monocyclopentadienyl ketimide and phosphinimide Group 4 catalyst technology platform developed by Nova Chemicals, targeting the production of linear, low-density PE (LLDPE). Keltan has introduced κ1-amidinate ligands tethered on titanium Cp complexes (Fig. 11), which were shown to be excellent catalysts with an outstanding, catalytic performance in the copolymerization of ethylene and α-olefins.

The Keltan ACE process is more sustainable compared to conventional ZN technology, because of the higher achievable reactor temperatures (less deep cooling of the monomers is required, resulting in considerable energy savings) and the higher affinity for propylene and the diene (less energy required for recycling of unreacted monomers). Thanks to the high ACE catalyst efficiency, the catalyst residue in the final EPDM product is extremely low, which is beneficial for heat ageing and electrical applications. In addition, the process does not require catalyst extraction as for conventional ZN catalysts, and, thus, produces no catalyst waste. Finally, the production capability in terms of gel content has strongly improved, i. e. virtually no gel. The end result is that the Keltan ACE technology is cleaner and has a reduced carbon footprint compared to ZN technology of ~10%.

With Keltan ACE catalyst technology a complete range of EPDM products with polymer characteristics similar to their ZN counterparts can be produced, including products with very high ENB content, very high molecular weight, CLCB and very high oil extension [14]. These Keltan ACE EPDM products were designed in such a way that they behave as perform-alikes (PAL), i.e. drop-ins for traditional ZN EPDM products. As well as the production of conventional ENB-EPDMs, the Keltan ACE catalyst was also explored for EPDM product diversification, such as the production of EPDMs with very high amounts of DCPD or VNB without excessive gela­ tion and reactor fouling, i.e. products that cannot or are extremely difficult to obtain via classical ZN catalysis (see above).

Reactive extrusion:

EP(D)M is an apolar polymer with a low level of unsaturation, which is not compatible in combinations with polar compounds and has hardly any chemical reactivity. In addition, wet recovery of EP(D)M is limited at low molecular weights, because of undesired flow and stickiness of the crumb in the final crumb drying step before baling. Keltan has developed reactive extrusion (REX) processes to overcome these issues. One of the world’s largest, twin-screw, co-rotating extruders designed for REX has been in operation on the Triunfo (Brazil) site since 2008. With an appropriate screw concept and extruder conditions, EPM rubber can be down-sheared in the REX extruder to lower molecular weight products without any sacrifice to MWD and color. These very low molecular weight EPM products are applied as so-called viscosity index improvers (VII) for engine oils used for the lubrication of internal combustion engines, and facilitate the operation of the engine not only during a cold start in the winter, but also at high temperatures when the engine is running for prolonged times. VIIs are dissolved in the lubrication oil to decrease the temperature dependency of the oil viscosity and thus, enable effective lubrication over a wide temperature range with one single oil product (no change of winter vs. summer oils needed) [15]. EPM rubber is the polymer of choice, because of its high thickening power combined with its high thermal stability and resistance against oxidation. Alternatively, EPM can be grafted with maleic anhydride in the REX extruder, yielding maleated EPM. Such an EPM product with a reactive, cyclic anhydride moiety grafted along the polymer backbone allows interaction between apolar EPM and more polar substances and thus, can be used as a compatibilizer and adhesive for combinations of EP(D)M with polyamides, polyesters, metals, minerals and glass. In addition, the maleated EPM can be further reacted with suit­able aromatic amines, such as N-phenyl-p-phenylenediamine, yielding EPM products with aromatic amine/imide functionalities which are applied as VIIs with dispersion index credits. The latter products combine VII performance with dispersing soot and metal particles in the engine oil, reducing wear of the engine [15].

New EPDM products and applications

EPDM is a hydrocarbon rubber with a very flexible, fully saturated polymer backbone and a low level of unsatura­tion in the side groups. As a result, EPDM can be vulcanised with sulfur curatives and shows high peroxide curing efficiency. EPDM combines fair mechanical strength with excellent weathering resistance, since it is intrinsically resistant to heat, oxygen, ozone and UV irradiation. The apolar character enables extension and compounding with large amounts of oil plasticizer and results in a very good resistance to aqueous systems and polar solvents. Finally, EPDM accepts very high levels of oil plasticizer and (reinforcing) filler(s), facilitating economic compounding. Consequently, EPDM is the rubber material of choice for outdoor and elevated temperature applications. EPDM is typically applied in automotive (sealing systems, radiator hoses, window wipers), building & construction (window gaskets, roof sheeting, waste and potable water seals), plastic modification (impact modification of PP, TPVs), wire and cable insulation and domestic appliances. Keltan has a track record of developing new EPDM products for enhanced performance in both existing and new EPDM applications. Keltan’s catalysis and technology position has resulted in a very broad EPDM product portfolio with a particularly strong position in products combining (very) high molecular weight (read: (very) high polymer Mooney viscosity), (very) high oil-extension, (very) high degree of branching and/or (very) high ENB content. Typically, a very high molecular weight is beneficial for the ultimate physical properties and elasticity of the crosslinked rubber products, but limits the mixing and processing of the rubber compounds. Long-chain branching (LCB: see above) and/or oil extension are ways to improve the balance of melt processing and solid state performance. It should be noted that oil extension of very high molecular weight EPDM is a technical necessity, since it not only enables the recovery of the very high molecular weight, highly viscous polymer in the EPDM plant, but also the mixing at the rubber compounder.

Fig. 12: Improved surface aesthetics of EPDM sponge based on K6951C EPDM compared to competitor grade(s):
left: Pyxargus’ surface defect analysis; right: cross-sections of sponges [16].

Fig. 12: Improved surface aesthetics of EPDM sponge based on K6951C EPDM compared to competitor grade(s):
left: Pyxargus’ surface defect analysis; right: cross-sections of sponges [16].

The characteristics and applications of (very) high molecular weight and oil-extended Keltan EPDM products will be highlighted here. Note that the four digits in the Keltan product code represent the Mooney viscosity at 125°C (ML), the ENB content, the ethylene content and the extender oil content, respectively. For example, Keltan 5469 corresponds to an EPDM product with a ML of 52, an ENB content of 4.5%, an ethylene content of 63% and an oil content of 100 phr. The optional letter after the four digit code indicates other details, such as C for Keltan ACE technology, D for DCPD as diene, Q for produced in Orange slurry plant and R for REX product.

EPDM rubber sponge is applied in automotive sealing systems and its production requires a delicate balance be­tween vulcanization and blowing reactions. High ENB levels are required for fast sulfur vulcanization, whereas a high molecular weight is needed to give sufficient melt strength for optimum cell formation. Keltan supplies various dedicated sponge EPDM grades, such as K6950C, the even higher molecular weight K9950C and the 15 phr oil-extended version of the latter, K6951C. These polymers combine very high molecular weight with an ENB content of 9 wt% and are renowned for their superior sponge performance. The low ethylene content of 44 wt% corresponds to a fully amorphous product without any crystallinity and thus, guarantees the best elastic recovery at low temperatures. The rather high degree of branching (Δδ ranges between 10 and 20) facilitates high melt strength. The latter K6951C product combines the best mixing for excellent filler dispersion with fast extrusion speeds, a high level of physical properties and a very smooth extruded sponge surface (Fig. 12) [16]. K7752C completes this sub-portfolio of Keltan EPDMs for sponge applications with even higher molecular weight and, thus, also somewhat higher oil content (25 phr). K4465 is a very high molecular weight EPDM with a high degree of branching (Δδ around 17), a very broad MWD of around 5 and 50 phr of extender oil. It provides excellent collapse resistance, combined with very good mixing and processing and is applied for hoses, seals and belts. K5465Q is another very high molecular weight EPDM with 50 phr oil. With 64 wt% ethylene it has a semi-crystalline nature, resulting in more strength making it suitable for soft, high strength, non-black compounds, and is used for bull eye applications in washing machines.

More recently, K9565Q was developed, which is yet another very high molecular weight EPDM with 50 phr oil [17]. This semi-crystalline product with 62 wt% ethylene can be used as the basis for automotive applications requiring good dynamic performance, such as (engine) mounts, bushings, flexible couplings and tie bars. Engine mount compounds with rather low levels of reinforcing filler and extra plasticizer oil (typically 50 and 10 phr, respectively) based on K9565Q provide the best balance of strength and re­silience. In addition, the incorporation of small amounts of NR in the K9565Q compound has resulted in dynamic fatigue behavior outperforming the traditional NR-based engine mounts. Obviously, the heat ageing resistance of sulfur-vulcanized K9565Q engine mount compounds is superior to that of the traditional NR products, which start to deteriorate upon ageing at 100°C and above. The continuously increasing under-the-hood temperatures of cars with combustion engines will drive the replacement of NR by K9565Q in dynamic applications.

TPVs are heterogeneous blends of PP and EPDM, consisting of a crosslinked EPDM dispersion in a PP matrix, produced via dynamic vulcanization of the EPDM phase during melt mixing with PP in a batch mixer or extruder [18]. Increasing the EPDM molecular weight results in the best phase inversion, the finest rubber particle dispersion and eventually, in the best solid state performance of the final TPV product. K5469C is an ultra-high molecular weight EPDM with 100 phr oil extension and thus, the best-in-class EPDM product for TPV applications. The ethylene content of 58 wt% results in a semi-crystalline behaviour, which provides an optimum balance between strength and room-temperature elasticity of the TPV product. An alternative EPDM product for TPV applications is K4577 with 75 phr oil and a somewhat lower molecular weight compared to K5469C.

A final example is K4869C, which combines ultra-high molecular weight with 100 phr oil-extension, a very narrow MWD (close to 2) and a very high ENB content of 9 wt%. This EPDM product has the best elastic performance of all EPDM products commercially available and, is therefore the polymer of choice for low-hardness corner mouldings, exhaust mounts, grommets and washing machine seals. It is finally noted that Keltan uses a high boiling point white oil in all of its oil-extended EPDM products, except those produced in Orange.

Green EPDM compounding

One of the greatest challenges facing our industries today is the need to design and develop sustainable technical solutions to address the mega-trends in society, such as mobility and urbanization. Key issues that have to be dealt with are emissions and fuel efficiency. Another challenge is the need to offer solutions to support a lower dependency on fossil fuels. To this end, Keltan has made a pioneering move towards exploring a future based on renewable resources by developing the world’s first bio-based EPDM rubber, commercialized under the tradename Keltan Eco [19]. Keltan Eco is produced in the EPDM plant in Triunfo (Brazil) by means of a solution polymerization process using ZN catalyst technology and uses bio-based ethylene supplied by Braskem, which originates from sugar cane. The sugar from sugar cane is converted to ethanol, which is then dehydrated to ethylene by Braskem in its Triunfo plant. This bio-based ethylene is transported via a pipeline to the neighboring Keltan EPDM polymerization plant. Depending on the ethylene content of the particular grade, the bio-based content of Keltan Eco EPDM rubber ranges between 50 and 70 wt%. The sustainability of this product has been validated by a Life Cycle Assessment performed by PE International (now Thinkstep) and the bio-based content can be measured and traced back by the ASTM D6866 C14 test. From a technical perspective, Keltan Eco is a drop in for fossil-fuel-based EPDM without any compromise on quality.

Translating this to final articles produced from mixed EPDM compounds, a bio-based content of 15-20 wt% can be achieved if Keltan Eco is the only bio-based ingredient of the rubber compound. In further efforts to increase the bio-based content of Keltan Eco based products we explored so-called green rubber compounding and also de­veloped green TPVs. By careful screening of the other rubber compounding ingredients, recycled carbon black from the pyrolysis of waste tires, silica from rice husk ash and micro-cellulose from wood were selected as (reinforcing) fillers, while hydrogenated coconut oil and especially sugar-based squalane were found to be useful plasticizers. In the end, a dynamic seal compound based on the amorphous K8550 Eco EPDM, and a static seal compound based on the crystalline K5470 Eco EPDM, were developed with a sustainable content of 86 and 90%, respectively [20]. In another approach TPVs were developed based on K5470 Eco EPDM combined with a bio-based LLDPE ex Braskem and again squalane as the plasticizer, yielding a TPV product with 86% bio-based ingredients and a technical performance similar to the fossil-fuel-based reference [21]. Essentially, these Keltan Eco based TPVs combine a renewable origin with recycling via melt processing.

Future innovation

In summary, the technical innovations highlighted above show the continuous drive for sustainable innovation at Keltan. Today on the one hand, a large effort is made to develop a 2nd generation Keltan ACE catalyst, aiming at an EPDM manufacturing process with enhanced sustainability and EPDM products with further improved performance. High-throughput experimentation (HTE) is exploited for the fast screening of new ligands and catalysts in a parallel fashion. Experimental results are interpreted via a Quantitative Structure Activity Relationship (QSAR) approach with structural characteristics of the catalyst calculated via molecular modeling. On the other hand, improved and new EPDM products and/or new EPDM applications are being developed to meet not only our customers’ future needs, but also the challenges of today’s society, such as those related to future e-mobility, clean water, agriculture, air quality, scarcity of raw materials and global warming.


Acknowledgements: The authors wish to acknowledge all former and current Keltan colleagues, who have contributed to the technical achievements presented in this overview and to the success of Keltan EPDM today.

[1] “Research tussen vetkool en zoetstof”, H. Lintsen (Ed.), Stichting Historie der Techniek, Uitgeversmaatschappij Walburg Pers, Zutphen (2000).
[2] A single ply, EPDM-based roof sheet was implemented by van Beek EPDM B.V. in Born (the Netherlands) in 1968; in “50 jaar EPDM Prak­tijkervaring op het dak: 50 jaar
Herlatan”, Carlisle CM Europe (2016).
[3] G.L.M. Vroomen, S.S. Lievens and J.P. Maes, “Influence of engine coolant composition on the electrochemical degradation behavior of EPDM radiator hoses”, in “Engine
Coolant Testing”, R.E. Beal (Ed.), 4th volume, ASTM STP, 1335 (1999) 183.
[4] “Test method for evaluating the electrochemical resistance of coolant system hoses and materials”, J1684 test by Society of Automotive Engineers (SAE).
[5] M. van Duin and P. Hough, RFP Rubber Fibers Plastics 9 (2014) 230.
[6] G.G. Evens, E.M.J. Pijpers and R.H.M. Seevens, European Patent 0044119B1 to DSM Netherlands B.V. (1985).
[7] G.G. Evens, E.M. Pijpers and R.H.M. Seevens, in “Transition Metal Catalyzed Polymerizations: Ziegler-Natta and Metathesis Polymerizations“, R.P. Quirk and R.E. Hoff (Eds.),
Cambridge University Press (1988) 782.
[8] L. d‘Agnillo, J.B.P. Soares and G.H.J. van Doremaele, Macromol. Materials Engin. 290 (2005) 256.
[9] H.J.H. Beelen, Kautsch. Gummi Kunstst. 52 (1996) 406.
[10] H.C. Booij, Kautsch. Gummi. Kunstst. 44 (1991) 128.
[11] B.J. Crosby, M. Mangnus, W. de Groot, R. Daniels and T.C.B. McLeish, J. Rheol. 46 (2002) 401.
[12] K. Dullaert, G.H.J. van Doremaele, M. van Duin and H. Dikland, Rubber Chem. Technol. 86 (2013) 503.
[13] G.H.J. van Doremaele, M. van Duin, M. Valla and A. Berthoud, J. Polym. Sci., Polym. Chem. 55 (2017) 2877.
[14] M. Alvarez Grima, M. van Boggelen, M. Dees, G.H.J. van Doremaele, M. van Duin and P. Henricks-Knape, Rubber World 250 Jan. (2014) 37.
[15] “Chemistry and Technology of Lubricants”, R.M. Mortier, M.F. Fox and S. Orszulik (Eds.), Springer, Dordrecht (2010).
[16] G. Choonoo, M. Dees and P. Hough, “Keltan® 6251A, a versatile sponge grade combining easy processing with excellent physical properties”, presented at the Fall 180th
Technical Meeting of the ACS Rubber Division, Cleveland (USA) (2011).[17] J. Beelen, N. van der Aar, M. van Duin, P. Spanos and C. Gögelein, RFP 11 (2016) 176.
[18] A.Y. Coran and R.P. Patel, “Thermoplastic Elastomers based on Dynamically Vulcanised Elastomer/Thermoplastic Blends”, in “Thermoplastic Elastomers”, 2nd ed. N.R.
Holden et al. (Eds.), Hanser Publishers, Munich (1996) ch. 7.
[19] M. Alvarez Grima, P. Hough, D. Taylor and M. van Urk, Eur. Rubber J. (Mar-Apr 2013) 28.
[20] M. van Duin and P. Hough, to be published in Kautsch. Gummi Kunstst. 71 (2018). January issue.
[21] N. van der Aar, M. van Duin, P. Hough, H. van Doormalen and G. Rademakers, “Development of high-quality EPDM products with enhanced sustainability”, presented
during International Rubber Conference (2016) in Kitakyushu (Japan) and Thermoplastic Elastomers World Summit (2017) in Philadelphia (USA).

About the authors

Prof. dr. ir. Martin van Duin

ARLANXEO Performance Elastomers, Keltan Global R&D, Principal Scientist

Dr. Ir. Niels van der Aar

Arlanxeo High Performance Elastomers, Head of Technical Service and Application Development Keltan© EPDM, Arlanxeo Elastomers B.V.