Fluoroelastomer Compatibility with Bioalcohol Fuels

9-1-99 Fluoroelastomer Compatibility with Bioalcohol Fuels Eric W. Thomas DuPont Performance Elastomers L.L.C. Copyright 9 SAE International ABSTRACT Global acceptance and use of biofuels is growing rapidly in the transportation sector. Diminishing reserves of limited and costly fossil fuel resources and a growing realization that world peak oil production will most likely occur within the next decade is driving significant investment in sustainable biofuels. Legislative, regulatory and market forces are driving developments which seek to reduce vehicle emissions, improve fuel efficiency, er environmental greenhouse gases and strengthen the economy. The use of alternate, sustainable, renewable fuels, preferably of domestic origin, is fostering considerable investment in new technologies. One promising technology is the addition of aliphatic alcohols to gasoline and diesel fuels. The compatibility of seal and hose materials commonly used in automotive fuel systems with conventional hydrocarbon fuels is well known. Over the past forty-five years fluorohydrocarbon elastomers have been successfully used in passenger car and truck and offway gasoline and petrodiesel fuel delivery and metering systems. More recently, biofuels such as ethanol have become technically and economically attractive blending constituents for gasoline and diesel fuels. These biomass fuels present their own set of material compatibility challenges to automotive fuel storage, delivery, and metering system component hardware. In this presentation the compatibility of selected fluoroelastomers with ethanol and butanol, and their respective alcohol-based blend fuels, will be reviewed. Fluoroelastomers that have historically been used for petroleum based fuels will be compared to several new types that display improved compatibility and sealing functionality with alcohol-based fuels. Accelerated long term testing results will be presented that characterize physical properties and property retention in bioalcohols through 18 hours at C. Properties relevant to sealing applications and hose will also be discussed and best in class elastomers will be lighted, so that, in gaining a better understanding of the respective capabilities of performance fluorohydrocarbon elastomers, the engineer may design more robust sealing systems for bioalcohol service. INTRODUCTION For nearly half a century fluorohydrocarbon elastomers (FKM) have demonstrated excellent performance in aerospace and automotive sealing applications. FKM s have demonstrated broad compatibility with petroleumbased fuels since their adoption in the automotive sector in 1961 (fuel metering components for carburetion) and are preferred elastomers for today s sophisticated common rail fuel injection systems. Market forces are driving spark ignition (SI) and compression ignition (CI) engine developments to decrease vehicle emissions, improve fuel efficiency and reduce environmental greenhouse gases (primarily CO ). The use of alternate, sustainable, renewable fuels, preferably of domestic origin, is driving significant investment in new technologies. Biofuels, such as ethanol and butanol, are technically and economically attractive blending constituents for gasoline. Ethyl alcohol, or ethanol, is a common, oxygencontaining, pollution control additive used in automotive gasoline, typically at the 6-1% level. Ethanol usage has been largely limited to the mid-western part of North America, due in part to supply and price constraints. As new manufacturing technology makes this alcohol more plentiful and government incentives continue to The Engineering Meetings Board has approved this paper for publication. It has successfully completed SAE s peer review process under the supervision of the session organizer. This process requires a minimum of three (3) reviews by industry experts. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. ISSN 18-7191 Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. SAE Customer Service: Tel: 877-66-733 (inside USA and Canada) Tel: 7-776-97 (outside USA) Fax: 7-776-79 Email: CustomerService@sae.org SAE Web Address: http://www.sae.org Printed in USA

encourage further production and use, there is a drive to use ethanol in automotive fuel at er percentages, for example at 8%, in E8 flex fuel. In 7 it is estimated that 6. billion gallons of ethanol was produced in North America for the transportation sector [1]. Ethanol offers performance in SI engines due to its inherently research octane number (RON). This property, together with its latent heat of vaporization, enables internal combustion engines running on this fuel to achieve thermal efficiency and performance levels. Engines optimized to exploit these beneficial properties are prevalent in today s world of performance motorsports. In 7 the IndyCar Series moved to 1% ethanol and in 9 the V8 Supercar Championship will move to E8. Further confidence in the performance attributes of ethanol is evident with Lotus Engineering s noteworthy efforts to optimize engine performance with E8 in flexfuel vehicles [, 3]. The rapidly increasing use of ethanol derived from corn has contributed to a growing controversy and debate about food vs. fuel []. Technology breakthroughs have been recently announced by DuPont and Genencor to develop and commercialize -cost cellulosic ethanol utilizing proprietary enzyme technologies and production platforms enabling biomass-to-sugar conversion rates []. Equally significant, General Motors announced a partnership with Coskata to use the company s technology to affordably and efficiently produce ethanol from a wide variety of renewable sources, garbage, old tires and plant waste [6]. These new technologies hold great potential in bringing cost ethanol to the market, where the fuel will not be derived from, or compete with, grain-based food sources. In 6 DuPont and BP announced the creation of a partnership to develop, produce and market biobutanol to meet the increasing global demand for renewable transportation fuels [7]. EXPERIMENTAL The materials evaluated in this study were prepared using commercial grades of fluoroelastomers. ASTM D118 designations are used to describe each material. FLUOROLEASTOMERS EVALUATED Typically, fluoroelastomers are differentiated by their monomer composition and the cure system employed. Traditional copolymer and terpolymer compositions, based on vinylidene fluoride (VF ), hexafluoropropylene (HFP) and tetrafluoroethylene (TFE), are considered standard products within the rubber industry. Other products offering some special aspect of functionality not provided for by the standard products are considered specialty types. For example, FKM- and FKM- contain the monomer perfluoromethylvinylether (PMVE) to provide enhanced temperature flexibility. Recently, advanced polymerization and curing technology designed to manufacture specialty fluoroelastomers with end use properties that meet evolving market needs was developed and commercialized [8,9]. These polymers, called Viton fluoroelastomer made with Advanced Polymer Architecture (APA Polymers), were designed to provide an excellent balance of processing and physical properties [1, 11]. These improvements are due to the modified molecular weight distributions, improved cure site monomers and unique polymer end group chemistry that were applied to the manufacture of these polymers. While the basic building blocks (monomers) are the same, polymer architecture has advanced considerably. The APA polymers have superior properties and processing characteristics when compared to their predecessors. These advances in polymer design have been necessary to keep pace with automotive seal fabricators quest to optimize part quality and manufacturing yields, in order to maintain competitive position, without sacrificing end-use properties. Table 1 lists those fluoroelastomers evaluated in this study and describes some of their attributes. Table 1 - Fluoroelastomers Evaluated Product General Description FKM-A61C Viton A-61C, VF /HFP, 66.% fluorine, bisphenol cure FKM-B61C Viton B-61C, VF /HFP/TFE, 68.% fluorine, bisphenol cure FKM- VTR99 VTR-99, VF /HFP/TFE, 69.% fluorine, bisphenol cure FKM-F6C Viton F-6C, VF /HFP/TFE, 7.% fluorine, bisphenol cure FKM- Viton GLT-6S, VF /TFE/PMVE/CSM, 6.% fluorine, peroxide cure FKM- Viton GBLT-6S, VF /TFE/PMVE/CSM, 66.% fluorine, peroxide cure FKM- Viton GFLT-6S, VF /TFE/PMVE/CSM, 67.% fluorine, peroxide cure FKM- Viton GF-6S, VF /HFP/TFE/CSM, 7.% fluorine, peroxide cure Monomer Description: VF = vinylidene fluoride HFP = hexafluoropropylene TFE = tetrafluoroethylene PMVE = perfluoromethylvinylether CSM = proprietary cure-site monomer APA polymers are designated with an S suffix

FUELS EVALUATED Table lists those fuels evaluated in this study and describes their general composition. Table - Fuels Evaluated Fuel General Description Fuel C Blend of % isooctane and % toluene Ethanol Ethyl alcohol (denatured with % methanol) CE-1 1% ethanol blended with 9% Fuel C CE- % ethanol blended with 7% Fuel C CE- % ethanol blended with % Fuel C CE-8 8% ethanol blended with 1% Fuel C E-1 1% ethanol Butanol 1-Butyl alcohol (99.% A.C.S Reagent) UTG-91 Unleaded test gasoline 91 octane Fuel C / 8% Fuel C blended with % butanol Butanol UTG91 / Butanol 8% UTG-91 gasoline blended with % butanol Test Protocols - Standard Property and Aging Tests Sample Preparation: Compounding of the fluoroelastomers was carried out using standard rubber lab equipment (internal mixer and two-roll mill.) Standard processing techniques commonly practiced in the industry were employed in preparation of test specimens. Compound formulations, with the exception of FKM-VTR99, were typical of those recommended for sealing service. FKM-VTR99 has been designed for fuel hose service. All fluoroelastomers contained 3 pph MT carbon black (N-99) to produce a nominal 7-8 durometer hardness. Standard ASTM slabs were prepared by compression molding as fols: FKM-A61C FKM-B61C FKM-F6C Press Cure: 7 min/177 C Post Cure: 16 h/3 C in air circulating oven Compression set in accordance with ASTM D 39-1. Temperature retraction (TR-1) in accordance with ASTM D 139-88. Glass transition temperature (Tg) by Differential Scanning Calorimetry in accordance with ASTM D 76-8. Measurements were made with a TA Instruments Differential Scanning Calorimeter, Model Q, and run in sub-ambient, modulated mode, at a scanning rate of C/min. Fluid immersion in accordance with ASTM D 71-98 all fuel testing was conducted in sealed 1 liter 316 stainless steel Parr pressure vessels. Permeation testing in accordance with SAE J66 using the Thwing Albert cup weight loss method. Compression Stress Relaxation measurements in accordance with ASTM D 617-97 using an MTS Model 83 and modified Wykeham-Farrance evaluation jigs designed to apply a constant compressive strain to a molded pellet test specimen. Static O-ring Seal Testing - This test method utilizes a special test rig and protocol developed by DuPont. No established ASTM test method yet exists. This test rig has been previously described [1, 13] and has been routinely used to screen the temperature functionality of elastomeric sealing materials. The apparatus consists of a 316 stainless steel block and plugs as shown in Figure 1, along with associated fasteners, valves, fittings and tubing. This apparatus was adapted, with major modifications, from SAE Aerospace Material Specification 773B [1]. The gland design conforms to SAE Aerospace Recommended Practices 131 and 13A [1, 16]. A radial squeeze of 19% is provided within three separate cells of the test block. The surface finishes are of industry standard (.1m / in). Apparatus cooling is provided by a temperature environmental chamber capable of achieving -7C. Figure 1 FKM- FKM- FKM- FKM- Press Cure: 7 min/177 C Post Cure: 16 h/3 C in air circulating oven FKM-VTR99 Press Cure: 3 min/16 C (air) Foling post cure, the FKM test specimens used for physical property and aging tests were die-cut from the molded slabs and tested per ASTM methods. Test Methods: The fluoroelastomers were characterized using the test methods described be: Stress-strain in accordance with ASTM D 1-98a, Die C. Hardness in accordance with ASTM D -.

A61C B61C VTR99 F6C A61C B61C VTR99 F6C The test procedure consists of selecting three AS68-1 O-rings [17] of excellent visual surface quality (parting line quality is critical) being lightly lubricated with a temperature Krytox perfluorinated polyether grease prior to installation in the test plug gland. The plug is fastened into the block with socket head screws until finger tight. A spring-energized Teflon PTFE face seal provides secondary backup directly above the leak port. The apparatus is placed into the environmental chamber, internally pressurized with dry air to. MPa (6 psi), then cycled down to -7 C at ~. C per minute until leakage is detected via IGLS f meters. The temperature is monitored continuously by RTD platinum thermocouples. The test reports the median temperature of the three cells at 1. Cubic Centimeter per Minute (CCM) leakage. All data is captured via a computerized data acquisition system. The rig then auto-cycles back to room temperature and the procedure is repeated with another set of O-rings. The time to run the test varies, depending upon the Tg of the material. With the FKM polymers, 3- hours is typical. In our study denatured ethanol and respective blends with Fuel C were evaluated. Also, 1-butanol and respective blends with Fuel C and UTG-91 unleaded gasoline were evaluated. 1-butanol was selected as a surrogate for BP butanol, since commercial quantities of biobutanol are not yet available. Testing was conducted in Parr pressure vessels at C for a period through 18 hours (6 weeks) with fuel refreshed weekly. Fuel C, ethanol, and four different blend ratios of Fuel C with ethanol (CE-1, CE-, CE- and CE-8) were evaluated. Three different butanol fuels were tested ranging from neat butanol to blends of 8% Fuel C / % butanol and 8% unleaded gasoline / % butanol. % blends of butanol with fuel were chosen for our study based on the maximum amount of butanol permitted by law currently, 1% in European Union (EU 8) and 11.% in the USA. EPA regulations require fuel manufacturers to produce fuels that are substantially similar ( sub-sim ) to fuels used for vehicle certification. RESULTS AND DISCUSSION Compatibility of elastomeric components with service fluids is typically ascertained by accelerated aging under laboratory conditions evaluating volume swell and physical property retention after exposure. The testing performed in this study employed methods typically used to characterize elastomeric sealing performance. Hardness changes, retention of stress-strain properties (tensile strength and elongation), volume swell, compression set, compressive stress relaxation, and temperature tests were performed. Vulcanizates of the FKM materials were subjected to immersion in the various alcohol-fuel blends. Hardness Change: Typically, one of the effects of fluid swell in elastomers is a corresponding loss of hardness (softening). This generally holds true for the materials reported in this study. Figures - indicate a strong relationship exists between hardness loss and the volume swell values shown later in this presentation. Hardness Change (pts) - -1-1 - - Figure Hardness Change in Ethanol Fuels - 18 hours at C A61C B61C VTR99 F6C Fuel C CE-1 CE- CE- CE-8 Ethanol Some softening (loss of hardness) was also observed as a function of predicted swell; however the correlation was not as strong as with swell. Noteworthy was the softening observed with FKM-VTR99 (a fluorine content FKM developed for fuel hose). Here we observe swell, but softening (Figures -). Since FKM- VTR99 was not postcured in this study, its crosslink network is not as tight as the other postcured FKM polymers evaluated, and shows a er loss is hardness ( points) after fuel immersion that is neither unexpected nor surprising. Hardness Change (pts) - - -6-8 -1-1 -1-16 -18 - Figure 3 Hardness Change in Fuel C / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Historically, elastomer compatibility testing has shown that FKM has generally been judged to be resistant to ethanol and methanol fuels based on accelerated fluid aging [18, 19].

A61C B61C VTR99 F6C A61C B61C VTR99 F6C A61C B61C VTR99 F6C A61C B61C VTR99 F6C A61C B61C VTR99 F6C All of the materials evaluated soften in the bioalcohols evaluated in this study. The best retention of hardness was observed with the PMVE-containing FKM's. Even after six weeks exposure, all were within -6 points of their original hardness. FKM- and FKM- displayed the least hardness change. The est softening was observed with blends of Fuel C and ethanol. This is consistent with historical testing performed with this aromatic-containing reference fuel. The other trend observed is that blends of Fuel C and ethanol result in more softening than either 1% Fuel C or 1% ethanol. The largest change appears to occur with the CE-1 and CE- fuel blends. Figure Change in Tensile Strength (%) -1 - -3 - - -6 Figure 6 Tensile Change in Fuel C / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Hardness Change (pts) - - -6-8 -1-1 -1-16 -18 - Hardness Change in UTG91 / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Tensile Retention: Figures -7 illustrate the trends observed in tensile changes after fuel immersion through 18 hours at C. Figure indicates that the er fluorine polymers show a er degree of tensile change than do the er fluorine FKMs. The second trend seen is that tensile loss is greater in fuel-ethanol blends than in either 1% Fuel C, 1% ethanol, or blends with Fuel C and butanol or Fuel C and gasoline (UTG91). Another trend seen is that VTR-99, which was not postcured, tends to exhibit a er degree of tensile loss. Figure Tensile Change in Ethanol Fuels - 18 hours at C Change in Tensile Strength (%) -1 - -3 - - -6 Figure 7 Tensile Change in UTG91 / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Elongation Retention: Retention of elongation after aging is a valid measure of fluoroelastomer compatibility to alcohols []. Elongation at break, and the change of this property during end-use service, is of particular importance to the function of static and dynamic seals. Therefore, the elongation change after aging in the respective alcohols evaluated will be reviewed. Figure 8 Elongation Change in Ethanol Fuels - 18 hours at C Change in Tensile Strength (%) -1 - -3 - - -6-7 -8 A61C B61C VTR99 F6C Fuel C CE-1 CE- CE- CE-8 Ethanol Change in Elongation at Break (%) 3 1-1 - -3 - - A61C B61C VTR99 F6C Fuel C CE-1 CE- CE- CE-8 Ethanol

A61C B61C VTR99 F6C Figures 8 and 9 depict the changes in elongation after fuel immersion. The trends for changes in elongation are similar to those observed for change in tensile, as shown in Figures -7. The three temperature FKM polymers show the est degree of loss in elongation with er fluorine FKM- showing the largest loss in elongation. With these three FKM polymers, the same trend is seen for changes in elongation that was exhibited by changes in tensile, i.e.: where straight 1% Fuel C and 1% ethanol do not have as large an effect as do blends of fuel and ethanol. In the case of the three standard FKM polymers, er loss in elongation is observed and two of the polymers, FKM-F6C and FKM-VTR99, actually increase in elongation after fuel immersion. While very little change is seen with the FKM-, the FKM-6C and FKM-VTR99 actually exhibit greater increase in elongation when the fuel is a blend of fuel and ethanol, than when these polymers are immersed in either 1% Fuel C or 1% ethanol. Change in Elongation at Break (%) 1-1 - -3 - - Figure 9 Elongation Change in Fuel C / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Volume Swell: Historically, volume swell, as tested in the laboratory under accelerated conditions, has been used as an indicator of elastomer chemical resistance to a specific fluid. Although it should not be used as the sole criterion for material selection, volume swell is of value in predicting seal compatibility. The amount of volume swell in a given fluid (fuel, oil, grease) can affect sealing performance, as it relates to available O-ring groove volume for example, or the abrasion and wear of dynamic sealing devices like radial lip seals. Volume swell can also be an indication of chemical attack when the volume swell of an FKM in a fluid with a minor amount of additive is significantly different than it is in the unmodified fluid. Volume Swell in Ethanol Fuels: Figure 1 illustrates the swell values for each FKM material when immersed in Fuel C through 18 hours at C. The degree of swell is essentially proportional to the fluorine weight percent for the polymers. FKM- (6.% fluorine) displayed the est volume swell (1%), foled by FKM-A61C (66% fluorine), FKM- (66% fluorine), FKM- (66.% fluorine) and FKM- B61C (68% fluorine). Those polymers containing 69-7% fluorine (FKM-VTR99, FKM-F6C and FKM- ) displayed the est swell values in Fuel C (- 9%). This trend will be observed throughout the balance of this study and closely mirrors those results previously reported by Stevens [1]. 16 1 1 1 8 6 A61C Figure 1 Volume Swell in Fuel C at C B61C VTR99 F6C hours 18 hours Figures 11-16 illustrate the swell values for the various fluoroelastomers when immersed in each ethanol based fuel (CE-1, CE-, CE-, CE-8 and E-1) over 18 hours at C. Blends of Fuel C and ethanol swelled all the FKM polymers more than did either 1% Fuel C or 1% ethanol. 1 1 A61C Figure 11 Volume Swell in CE-1 at C B61C VTR99 F6C hours 18 hours

Figure 1 shows the maximum FKM swell for all FKM types was observed with the CE- fuel blend. FKM- swelled 3%, foled by FKM-A61C (%), FKM- (16%), FKM- (1%) and FKM-B- 61C (16%). The fluorine-containing FKM s, FKM- VTR99 (69.% fluorine), FKM-6C (7.% fluorine) and FKM- (7.% fluorine), exhibited the est volume swell (11%) in all fuels tested after 18 hours. 1 1 A61C Figure 1 Volume Swell in CE- at C B61C VTR99 F6C hours 18 hours Figure 13 illustrates the swell values in CE- and Figure 1 in CE-8. Here we observe that the er fluorine FKM s swell 16-%; the value progressively diminishing to 8-1%% for the fluorine-containing types. 3 1 1 A61C Figure 13 Volume Swell in CE- at C B61C VTR99 F6C hours 18 hours 18 16 1 1 1 8 6 A61C Figure 1 Volume Swell in CE-8 at C B61C VTR99 F6C hours 18 hours Figure 1 provides the swell values in E-1 (1% denatured ethanol). Here we observe that the er fluorine FKM s swell 1-13%; the value progressively diminishing to less than % for the fluorinecontaining types. 1 1 1 8 6 A61C Figure 1 Volume Swell in Ethanol at C B61C VTR99 F6C hours 18 hours Volume Swell in Butanol Fuels: We observed very swell in butanol (<%) for all the FKM s tested (Figure 16) and er swell values with either the RFC or UTG- 91 fuel blends (Figures 17 & 18). 3 3 1 1 A61C Figure 16 Volume Swell in Butanol at C B61C VTR99 F6C hours 67 hours 18 hours

Figures 17 and 18 illustrates er fluorine-containing FKM- has a maximum swell up to 17% after 18 hours exposure at C in RFC / butanol (8/) and somewhat less with unleaded gasoline and butanol fuel blends. These test results support the general rule of thumb that fuel swell is primarily governed by polymer wt % fluorine. 18 16 1 1 1 8 6 1 1 1 8 6 Figure 17 Volume Swell in Fuel C / Butanol (8/) at C A61C B61C Figure 18 VTR99 F6C hours 67 hours 18 hours Volume Swell in UTG91 / Butanol (8/) at C A61C B61C VTR99 F6C hours 67 hours 18 hours Figure 19 provides a compiled view of volume swell for all the FKM materials tested after 18 hours at C in butanol, ethanol, Fuel C, 8/ blend of Fuel C / butanol, and a 7/ blend of Fuel C / ethanol. Maximum swell observed for all FKM materials was observed with the 7/ blend of Fuel C and ethanol (CE-). 1 1 Figure 19 Volume Swell in Bioalcohol Fuels - 18 hours at C A61C B61C VTR99 Butanol Ethanol Fuel C Fuel C/Butanol (8/) Fuel C/Ethanol (7/) F6C Effect of Temperature on Swell: Figure illustrates the effect of temperature on swell. Testing was conducted at,, 6 and 8 C in CE-1 for 168 hours. We note that increasing temperature amplifies volume swell. The increase can be appreciable. For example, even fluorine content FKM- exhibited a swell increase from % when tested at C, increasing proportionally to 18% at 8 C. Whenever possible, the temperature of our bench testing should reflect known field conditions in order optimize material selection for desired service life. 3 3 1 1 A61C Figure Effect of Temperature on Swell Volume Swell in CE-1 B61C VTR99 F6C @ C @ C @ 6 C @ 8 C

Permeation Resistance: Permeation resistance was measured on the same polymers using the same six fuel-ethanol blends and fuel-butanol blends. Permeation testing was conducted per SAE J66, using the cup weight loss method, which has been described in detail in a previous SAE papers [1, ]. Testing was conduct for 18 hours at C with the results being expressed in permeation rate units of grams-mm/m /day. The results are shown in Figures 1 and. Figure 1 Perm rate (g-mm/m²/day) 1 1 1 8 6 Permeation of Fuel C / Ethanol Blends at C A61C B61C VTR99 F6C Fuel C CE-1 CE- CE- CE-8 E-1 Much the same trend was observed as in the previously reported volume swell results [18]. The er fluorine content FKM- exhibited the est permeation rate ranging from ~7 g-mm/m /day in Fuel C to a peak of ~11 g-mm/m /day in CE- fuel and ~7 g- mm/m /day in neat ethanol. A strong trend was observed for permeation rates where levels of ethanol in Fuel C appear to exhibit er permeation rates. Specifically, CE- fuel, with % ethanol content, consistently produced the est permeation rates, regardless of the FKM polymer being tested. FKM- exhibited a marked improvement over FKM- when comparing the temperature FKM polymers, and FKM- had the est permeation rate of any of the polymers evaluated in this study. Figure provides a representation of the permeation rates of the FKM s when exposed to butanol, Fuel C / butanol blend (8/) and UTG-91 / butanol blend (8/). Here we observe significantly er permeation rates (approximately 7% er) with 1% butanol vs. denatured ethanol. Perm rate (g-mm/m²/day) 1 1 A61C Figure Permeation of Fuel C / UTG91 / Butanol Blends at C B61C Butanol Fuel C/Butanol (8/) UTG91/Butanol (8/) VTR99 F6C Sealing under Compression: When elastomers are used as sealing devices, they are compressed or strained to some degree. When that strain is removed after a period of time, the material will not fully recover its original shape. Creep, stress relaxation and compression set are undesirable related phenomena, which occur in all elastomeric articles and reflect the inherent viscoelastic nature of an elastomer and the limited stability of vulcanizate crosslinks. Creep is a time dependent increase in deformation under conditions of constant stress [3]. Continuous stress relaxation is decay in stress, as a function of time, under conditions of constant strain [3]. It is of great importance to sealing devices such as O-rings and gaskets. Both creep and stress relaxation is significant because they often play a role in the failure of rubber components. Set, often referred to as "permanent set", or irrecoverable creep, is the permanent deformation which remains when a material is released from the strain imposed and is measured in tension, or more commonly, in compression [3]. Both stress relaxation and creep are the result of physical and chemical relaxation processes. The physical process is due to the viscoelastic nature of rubber and usually decreases linearly with the logarithm of time. It is associated with reorientation of the elastomer molecular network under strain, with disengagement and rearrangement of chain entanglements. The chemical process is generally caused by chain scission or isomerization of crosslinks and usually occurs linearly as a function of linear time. The time dependence of the chemical process is a function of the order of the chemical reaction, the temperature and the imposed stress. Both processes occur simultaneously. Physical relaxation predominates at short times, while chemical effects are more significant at longer duration. When an elastomeric component is subjected to a static load, the load will cause a progressive increase in deformation as a function of time. Biaxial stress relaxation on compression in rubber consists of both

physical creep and chemical creep (due to molecular chain breaking). When a constant strain is imposed on an elastomer, the force necessary to maintain that strain is not constant but decays exponentially with time from the initial maximum to an eventual equilibrium state. This phenomenon of force decay is called stress relaxation and is of great importance in rubber sealing devices such as O-rings, packings and gaskets. Stress relaxation can be the dominant factor that limits the effective service life of the sealing device. Compression Set: In this method the test specimen is typically compressed %, held under compression for a specified time at temperature, and subsequently released. Measurements are then made to determine how a material recovers from this deformation. Testing can be performed in gas (typically air) or immersed in a fluid. In this study compression set was measured per ASTM D 39, Method B, on AS 68-1 O-rings. Figure 3 illustrates the compression set results obtained when tested in 1 C air for 168, 336, and 18 hours. Shorter duration testing has been routinely specified in automotive sealing documents; however as the industry moves towards extended powertrain warranties, longer term testing is becoming increasingly necessary. The best long-term (18 hour) compression set properties observed were obtained with FKM-A61C, which employs a thermally robust cure system. Compression Set (%) 3 3 1 1 Figure 3 Compression Set in Air at 1 C, O-rings A61C B61C F6C hours 18 hours Stress Relaxation: The stress relaxation data reported was measured using an MTS 83 instrument and modified Wykeham-Farrance fixtures designed to apply a constant compressive strain to a molded pellet test specimen. Other styles of stress relaxation apparatus have been documented and each has value, as well as limitations. What we seek is a useful and repeatable tool to predict long-term elastomeric sealing performance. In our study we tested for 18 hours (six weeks) in air at 1 C, as well as 18 hours in CE-1 at C. Figure shows the percentage of compressive force retained as a function of time in air. Here we can observe how the compressive force decays over time with the fluoroelastomers. FKM- and FKM- exhibited the best performance, retaining 6% of their original compressive force. This was foled by FKM- FKM- (63%), FKM-A61C (6%), FKM-GF- S (6%), FKM-B61C (7%) and FKM-6C (9%) after 18 hours exposure. Retained Sealing Force 1% 9% 8% 7% 6% % % 3% % 1% % Figure Stress Relaxation in Air at 1 C 8 168 336 67 8 18 Time (hours) A61C B61C F6C Figure illustrates the percentage of compressive force retained as a function of time in CE-1 after 18 hours at C. The benefit of fuel swell partially negates the progressive decay of compressive force. With the exception of FKM-6C, all of the fluoroelastomers evaluated retained better than 8% of their original force values when tested in CE-1 fuel at C. Retained Sealing Force 11% 1% 9% 8% 7% 6% % % Figure Stress Relaxation in CE-1 at C 8 168 336 67 8 18 Time (hours) A61C B61C F6C Low Temperature Properties: At temperatures the modulus of an elastomeric sealing element increases as the temperature decreases. As the sealing material approaches its glass transition temperature (Tg), hardness increases, and it ultimately loses its ability to remain flexible []. Sealing force decays very rapidly. Figure 6 illustrates the temperature characteristics of the various fluoroelastomers tested under subambient conditions. Here we compare the results obtained from differential scanning calorimetry (Tg),

B61C F6C GBLT- S temperature retraction (TR-1) and static O-ring sealing characterizing temperature properties. FKM- A61C has a Tg of -17 C and a temperature retraction value of -16 C. This requirement (TR-1) is reflected in the major FKM aerospace sealing documents currently in use (AMS776, AMS79, AMS316, AMS318). FKM-B61C has a Tg of -1 C and FKM-6C of -8 C. Note that FKM- has the least desirable temperature properties, the expected result of its VF content. The best temperature properties were observed with the PMVE-containing fluoroelastomers: FKM-, FKM- and FKM-. These polymers have Tg s of -31 C, -7 C and - C, respectively, and are increasingly specified and used for critical sealing applications found in sophisticated fuel injection systems. Temperature, C A61C -1 - -3 - - -6 Figure 6 Low Temperature Properties of Fluoroelastomers (HFP-containing types) A61C B61C F6C TR-1 Tg O-ring Sealing (PMVE-containing types) Comparison with temperature retraction (TR-1) and glass transition (Tg) illustrate good correlation. From the static O-ring seal testing we see about a 1 C correlation between temperature sealing values and those obtained from TR-1 or Tg methods. The ability of O-rings to seal at temperature correlates closely with glass transition and temperature retraction. More importantly, the ability of fluoroelastomer O-rings to seal at temperature, as determined by this method, is predominately dependent upon polymer composition. CONCLUSION Tests of physical properties of eight different fluoroelastomer polymers, of varying fluorine content, in a standard MT black formulation, after immersion in different Fuel C, ethanol, butanol, UTG-91 and their blends indicated the foling trends: Retention of physical properties is primarily a function of polymer fluorine content. Higher fluorine content results in better physical property retention (hardness, tensile strength, elongation at break), er volume swell and er permeation in all fuels tested. When immersed in blends of Fuel C, ethanol and butanol, most of the FKM polymers tested exhibited greater tensile and elongation changes than they did in either 1% Fuel C, 1% ethanol or 1% butanol. Blends of Fuel C and ethanol and Fuel C, unleaded gasoline and butanol, swelled all the FKM polymers to a greater extent than either 1% ethanol or 1% butanol. Blends of Fuel C and ethanol are the most detrimental to the FKM polymers evaluated. The worst case fuel for volume swell of all the FKM polymers evaluated was CE- (% ethanol / 7% Fuel C). High fluorine content, peroxide cured FKM-, FKM-, and FKM- exhibited er hardness changes than bisphenol cured FKM-F6C and FKM- VTR99. FKM-VTR99, which was not postcured, had good retention of tensile and elongation properties after fuel immersion. FKM-VTR99 had a er loss of hardness after fuel immersion. Trends in volume swell appear to fol the trends seen in hardness change, with the exception of FKM-F6C, and FKM-VTR99 which was not postcured. High fluorine content FKM- exhibited the est volume swell and est permeation in all fuels tested. The temperature properties of the fluoroelastomers evaluated is predominately a function of polymer composition. Of the specialty temperature types tested, FKM-, FKM- and FKM- exhibited the best temperature properties. The selection of fluoroelastomer for bioalcohol service is typically a function of the temperature requirements of the specific end-use application. ACKNOWLEDGMENTS The author would like to thank John Leonhard and Russ Morrison, Laboratory Technologists, DuPont Performance Elastomers - Akron Lab, for their dedicated work in generating the data presented in this paper. The author is also appreciative for the valued consultation provided by Ronald Stevens, Scientist, DuPont Performance Elastomers, Stow, Ohio. REFERENCES 1. Renewable Fuels Association www.ethanolrfa.org. Turner, J.W.G., R.J. Pearson, B. Holland, and A. Peck, Alcohol-Based Fuels in High-Performance Engines, presented at SAE Fuels and Emissions Conference, Cape Town, South Africa, 7-1- 6, January 7. 3. Turner, J.W.G., A. Peck and R.J. Pearson, Flex- Fuel Vehicle Development to Promote Synthetic

Alcohols as the Basis of a Potential Negative-CO Energy Economy, presented at SAE 1 th Asia Pacific Automotive Engineering Conference, Hollywood, California, 7-1-3618, August -8, 7.. Walter, Sarah R., Corn Ethanol as an Alternative Fuel: Technical and Economic Issues, prepared as part of the WISE program, 6-1-179, 6.. www.dupontdanisco.com 6. www.coskata.com 7. http://www.dupont.com/biofuels/en_us/ 8. Stevens, Ronald D. and Lyons, Donald F., Paper #9, New, Improved Processing HFP-Peroxide Cured Types of Viton, 16 th meeting of the Rubber Division, American Chemical Society, Cleveland, Ohio, October 1. 9. Lyons, Donald F. and Stevens, Ronald D., Paper #3, New, Improved Processing PMVE-Peroxide Cured Types of Viton, presented at the 16 th meeting of the Rubber Division, American Chemical Society, Cleveland, Ohio, October 1. 1. Bowers, Stephen, New Specialty Viton Products for Automotive Sealing Applications, presented at the 17 th International Conference on Fluid Sealing, York, United Kingdom, April 8-1, 3. 11. Thomas, Eric, New Fluoroelastomer Developments for Aerospace Sealing Applications, presented at the 163 rd meeting of the Rubber Division, American Chemical Society, San Francisco, California, April 8-3, 3. 1. Stevens, Ronald D. and Thomas. Eric. W., Low Temperature Sealing Capabilities of Fluoroelastomers, presented at SAE World Congress, 919, February 6-March, 199. 13. Thomas, Eric W., Fluoroelastomer and Perfluoroelastomer Compatibility with Advanced Gas Turbine Lubricants, presented at SAE Aerospace Congress, 3-1-39, September 8-1, 3. 1. Rings, Sealing, Fluorosilicone Rubber, High Temperature Fuel and Oil Resistant, SAE AMS 773C, 199. 1. Gland Design, Elastomeric O-Ring Seals, General Considerations, SAE ARP 131, 1973. 16. Gland Design, Elastomeric O-ring Seals, Static Radial, SAE ARP 13A, 1977. 17. Aerospace Size Standard For O-Rings, SAE AS 68A, 198. 18. Stevens, Ronald D. and Thomas, Eric W., Fluoroelastomer Developments for Automotive Fuel Systems, presented at SAE World Congress, 88, February 9-March, 1988. 19. Stevens, Ronald D. and Stahl, William M., Fuel- Alcohol Permeation Rates of Fluoroelastomers, Fluoroplastics, and other Fuel Resistant Materials, presented at SAE World Congress, 9163, February -8, 199.. Dobel, Theresa M. and Bauerle, John B., New FKM Developments for Automotive Powertrain Applications, presented at SAE World Congress, -1-7, March 6-9,. 1. Stevens, Ronald D., Fuel and Permeation Resistance of Fluoroelastomers to Ethanol Blends, presented at the 17 th meeting of the Rubber Division, American Chemical Society, Cincinnati, Ohio, October 1-1, 6.. Stevens, Ronald D., Permeation and Stress Relaxation Resistance of Elastomeric Fuel Seal Materials, presented at SAE World Congress, 1-1-117, March -8, 1. 3. Gent, A.N., Engineering with Rubber, nd Edition, HanserGardner Publications, Cincinnati, OH, 1, pp. 18-183.. Designing With Elastomers for Use at Low Temperatures, Near or Be Glass Temperature, SAE AIR 1387B, 199. CONTACT Eric Thomas is a Senior Technology Engineer with DuPont Performance Elastomers located in Stow, Ohio. He may be reached at 33-99-691 or e-mail at eric.w.thomas@dupontelastomers.com. APPENDIX Fuel C available from: Phillips Chemical Co. Phone: 1-918-661-991 Ethyl alcohol, 9%, denatured, available from: EMD Chemicals Inc. Phone: 1-86-3-63 UTG-91 unleaded test gasoline 91 octane available from Chevron Phillips Chemical Co. 1-Butanol (99.% Reagent) available from Sigma Aldrich Corp. Abbreviations: CI Compression ignition FKM Fluorohydrocarbon elastomer RON Research octane number SI Spark ignition SPECIAL NOTE: The information set forth herein is furnished free of charge and is based on technical data that DuPont Performance Elastomers believes to be reliable. It is intended for use by persons having technical skill, at their own discretion and risk. Handling precaution information is given with the understanding that those using it will satisfy themselves that their particular conditions of use present no health or safety hazards. Since conditions of product use and disposal are outside our control, we make no warranties, express or implied, and assume no liability in connection with any use of this information. As with any material, evaluation of any compound under end-use conditions prior to specification is essential. Nothing herein is to be taken as a license to operate or a recommendation to infringe on patents. Caution: Do not use in medical applications involving permanent implantation in the human body. For other medical applications, discuss with your DuPont Performance Elastomers customer service representative and read Medical Caution Statement H-6937. Viton fluoroelastomer is a registered trademark of DuPont Performance Elastomers LLC. Krytox performance lubricants and Teflon fluorocarbon resins are registered trademarks of the DuPont Co.