Energy & Fuels 2009, 23, 857–861
857
Interactions of Jet Fuels with Nitrile O-Rings: Petroleum-Derived versus Synthetic Fuels Robert J. Gormley,*,† Dirk D. Link,† John P. Baltrus,† and Paul H. Zandhuis‡ United States Department of Energy, National Energy Technology Laboratory, P.O. Box 10940, Pittsburgh, PennsylVania 15236-0940 and RDS-Parsons, P.O. Box 618, South Park, PennsylVania 15129 ReceiVed September 22, 2008. ReVised Manuscript ReceiVed NoVember 21, 2008
A transition from petroleum-derived jet fuels to blends with Fischer-Tropsch (F-T) fuels, and ultimately fully synthetic hydro-isomerized F-T fuels has raised concern about the fate of plasticizers in nitrile-butadiene rubber o-rings that are contacted by the fuels as this transition occurs. The partitioning of plasticizers and fuel molecules between nitrile o-rings and petroleum-derived, synthetic, and additized-synthetic jet fuels has been measured. Thermal desorption of o-rings soaked in the various jet fuels followed by gas chromatographic analysis with a mass spectrometric detector showed many of the plasticizer and stabilizer compounds were removed from the o-rings regardless of the contact fuel. Fuel molecules were observed to migrate into the o-rings for the petroleum-derived fuel as did both the fuel and additive for a synthetic F-T jet fuel additized with benzyl alcohol, but less for the unadditized synthetic fuel. The specific compounds or classes of compounds involved in the partitioning were identified and a semiquantitative comparison of relative partitioning of the compounds of interest was made. The results provide another step forward in improving the confidence level of using additized, fully synthetic jet fuel in the place of petroleum-derived fuel.
Introduction Over the past few years, the United States Department of Defense has made a concerted effort to incorporate FischerTropsch (F-T) fuels into its fuel portfolio in order to address national security concerns by decreasing its dependence on imported fuel. Substitution of synthetic fuels in military jets is an immediate goal because of the large volumes of the fuel currently used by the Air Force. Because the F-T fuel infrastructure in the U.S. is in its infancy, thus limiting the current availability of the fuel, the near-term goal has been to replace half of the petroleum-derived fuel and operate jet engines with a 50:50 blend.1-4 Also, operational concerns with using 100% synthetic fuel have led to the pursuance of the 50:50 blend. The synthetic fuel can be derived from a wide range of fossil fuel and biomass feedstocks by incorporating a gasification step prior to the F-T reaction. During the F-T process, the initial reaction products, which are long-chain hydrocarbons, are transferred to a second catalytic reactor where they undergo hydro-isomerization to produce mostly branched alkanes. This branching gives the fuel some of its desired properties, such as suitable density, low freezing point, and high cetane number. Because of the reaction conditions and catalyst used, the * To whom correspondence should be addressed. (412)
[email protected] † United States Department of Energy, National Energy Technology Laboratory. ‡ RDS-Parsons. (1) Ephron, D. Warplanes: The Air Force, B-52s and Mideast Oil; Newsweek, U.S. Edition, 2006. (2) Phillips, D. Military Looks At Synthetic Fuel for Bombers and Fighters; International Herald Tribune, 2007. (3) Upson, S., U.S. Air Force Synthetic Fuel Program in Limbo; IEEE Spectrum, 2008. (4) Welch, W. M., Military Tests First Synthetic Fuel for Jets; USA Today, 2006.
10.1021/ef8008037
resulting hydro-isomerized fuels are different from other synthetic jet fuels currently in use 5,6 in that they contain no aromatic compounds. The absence of aromatics and heteroatomcontaining organic compounds leads to a deficiency in swelling and lubricity characteristics in these synthetic fuels. While the use of a 50:50 synthetic:petroleum blend provides an immediate solution to the lack of aromatics in the synthetic fuel, the much longer-term goal of using a fully synthetic fuel will require that the swelling and lubricity deficiencies be addressed. Previous research has attempted to identify a suitable additive to improve the swelling of nitrile rubber o-rings exposed to synthetic fuels.7-9 Those efforts have identified benzyl alcohol as a potential additive to promote the swelling of nitrile rubber.8,10 While benzyl alcohol appears to be an excellent additive from a swelling standpoint, there are still concerns that must be addressed with respect to its effect on several other critical properties of the fuel. We have already begun to address one of those properties, the thermal stability of the fuel.10 During (5) Lamprecht, D. Fischer-Tropsch Fuel for Use by the U.S. Military as Battlefield-Use Fuel of the Future. Energy Fuels 2007, 21 (3), 1448– 1453. (6) Moses, C. A.; Wilson G.; Roets P. N. J. Evaluation of Sasol Synthetic Kerosene for Suitability as Jet Fuel. In U.K. AViation Fuels Committee Documentation, 2003. (7) Baltrus, J. P.; et al. Screening of Potential O-ring Swelling Additives for Ultra-Clean Transportation Fuels. Prep. Pap. -Am. Chem. Soc., DiV. Petrol. Chem. 2005, 50 (2), 764–766. (8) Baltrus, J. P., et al., Screening of Potential O-Ring Swelling Additives for Ultra-Clean Transportation Fuels. In Ultraclean Transportation Fuels; Ogunsola O. I., Gamwo I. K., Eds.; Oxford University Press: New York, 2007. (9) Graham, J. L.; et al. Swelling of Nitrile Rubber by Selected Aromatics Blended in a Synthetic Jet Fuel. Energy Fuels 2006, 20 (2), 759– 765. (10) Link, D. D.; et al. Potential Additives to Promote Seal Swell in Synthetic Fuels and Their Effect on Thermal Stability. Energy Fuels 2008, 22 (2), 1115–1120.
This article not subject to U.S. Copyright. Published 2009 by the American Chemical Society Published on Web 01/28/2009
858 Energy & Fuels, Vol. 23, 2009
Gormley et al.
recent discussions of our work, a concern was expressed about the fate of plasticizers in nitrile o-rings when they are exposed to synthetic and additized synthetic fuels. It is known that plasticizers are typically extracted when a conventional fuel contacts the O-ring, and this is considered to be normal. However, there is a concern that any change in the behavior of the plasticizers in the presence of synthetic fuels or additized synthetic fuels would be undesirable. The research described herein attempts to determine any differences in the fate of compounds present in nitrile o-rings on exposure to petroleum-derived jet fuel, synthetic jet fuel, and synthetic jet fuel additized with 1 volume % benzyl alcohol. It is intended to complement previously published work measuring the partition coefficients of various compounds between the fuel and nitrile rubber o-rings.9 Experimental Section Materials and Samples. Nitrile rubber o-rings are widely used in fuel applications, and thus were chosen for the current study. (To protect potentially sensitive manufacturer-specific composition data, the commercial source of the O-ring will not be identified herein.) Benzyl alcohol (BzOH) was purchased from Sigma-Aldrich and was used as received. The petroleum-derived fuel (JP-5) was obtained from colleagues at Wright-Patterson Air Force Base, while the synthetic fuel (S-5) was obtained from Syntroleum Corp, Tulsa, OK. It should be noted that the composition of all synthetic fuels produced by the Fischer-Tropsch process are not the same, as much depends on the catalysts and processes used, as well as postsynthesis treatment and upgrading procedures.5 The composition and properties of the S-5 used in the current study has been previously reported.10,11 Treatment of O-Rings. In order to be able to present semiquantitative results and because previous research 9,10 has shown that butadiene-nitrile o-rings change weight and volume after swelling or shrinking in various jet fuels, efforts were made to maintain a constant starting mass of O-ring. Thus, an O-ring piece with an initial weight of 0.0013 to 0.0014 g was cut from the O-ring with a razor blade. The O-ring piece was soaked in a given volume of fuel to maintain a constant ratio of volume of fuel to volume of O-ring of 114:1. This made current testing consistent with previously reported results for O-ring swelling tests.10 The O-ring pieces were soaked for a given period of time (hours to days) and then removed from the fuel and left exposed to air to allow residual fuel to evaporate. Various evaporation times (10, 30, and 120 min) were tested initially (data not shown) to determine an appropriate period. A shorter drying time resulted in more fuel being retained by the O-ring. A 30-min evaporation time in air at room temperature was used in order to have sufficient fuel available for detection by GC-MS and for convenience in allowing enough time to carry out measurements required before inserting the O-ring into the thermal desorption attachment. Analytical System. A heated temperature desorption system (TDS, Gerstel Corporation) was used to desorb and transfer O-ring constituents through a splitless inlet onto a cooled injection system (CIS). The materials collected on the CIS were then released onto the GC column through a 1:50 split injection. The temperature program used for the TDS was 40 to 350 at 60 °C/min, with a hold time of 3 min, while the CIS was heated from 0 to 350 at 12 °C/sec with a hold time of 3 min. The GC-MS system was an Agilent 6890 GC coupled to an Agilent 5973 mass spectrometric detector. The column was a Zebron ZB-1HT (Phenomenex, Torrance, CA) 30 m × 0.25 mm id, 0.25 µm film thickness. The oven temperature program was from 50 to 350 at 10 °C/min. The helium carrier was operated in constant flow mode, at 1.0 mL/min, which translated to a starting pressure of 7.7 psig. (11) Muzzell, P. A.; et al. Composition of Syntroleum S-5 and Conformance to JP-5 Specification. Prepr.-Am. Chem. Soc., DiV. Pet. Chem. 2004, 49 (4), 411–413.
Figure 1. Overlay of GC-MS chromatograms for o-rings soaked 1 day.
Chemstation control software (Agilent) was used for data acquisition and processing. Peaks in the various chromatograms were identified using ChemStation to perform NIST library searches and evaluate the quality of the match obtained.
Results and Discussion Baseline Case. The purpose of the first set of experiments in this study was to establish a baseline for what compounds could be released from acrylonitrile butadiene o-rings when exposed to various fuels. A solid desorption of the O-ring material into the GC, similar to that used in other studies,9 was used to determine the species that had been released from the O-ring after soaking in the liquids, after initial attempts using direct injection of liquid fuel samples proved unsuccessful because the target species were present in trace amounts and their detection was hampered by a high background from the fuel matrix. Attempts at using a previously developed method for extracting and concentrating these species prior to analysis 12,13 were also unsuccessful. It should be noted that the GC-MS methodology used in this study was not designed to provide truly quantitative data regarding the concentrations of the various species measured. This would have required significant calibration efforts using individual standards for unknown species to account for differences in ionization efficiencies and response factors. Instead, the peak areas obtained were meant to show broader trends in the species present in and released from as-received o-rings and those that have been soaked in various fuels for different lengths of time. The assignment of GC peaks can also be subject to considerable uncertainty, but for this work, the objective is more to follow the behavior of the class of compounds labeled “plasticizers” as a whole rather than focusing on specific compounds. Samples of dry, as-received, o-rings were analyzed according to the protocol described above for determining the characteristic species being desorbed from the O-ring when no fuels were present. The chromatogram of the dry O-ring is shown in Figures 1, 2, and 3 as a baseline for the dry O-ring tests as well as for comparison to illustrate changes that are taking place after treatment of the o-rings. There are a number of significant GC-MS peaks that appear when the O-ring undergoes thermal desorption. Retention times for these peaks, tentative identifica(12) Link, D. D.; Baltrus, J. P.; Zandhuis, P. Isolation and Identification of Nitrogen Species in Jet Fuel and Diesel Fuel. Energy Fuels 2007, 21 (3), 1575–1581. (13) Link, D. D.; et al. Extraction, Separation, and Identification of Polar Oxygen Species in Jet Fuel. Energy Fuels 2005, 19, 1693–1698.
Interactions of Jet Fuels with Nitrile O-Rings
Energy & Fuels, Vol. 23, 2009 859
Figure 2. Overlay of GC-MS chromatograms for o-rings soaked 7 days.
Figure 3. Overlay of GC-MS chromatograms for o-rings soaked 1 h. Table 1. Species Identified in the Thermal Desorption of Dry O-Ring Samples (as received) retention time (min)
index letter
peak identification
area countsa
8.9 12.3 13.1 16.1 17.8 18.2 19.8 20.1 24.0 24.8-25.3
C I F L M N O P Q R
2-(2-butoxyethoxy) ethanol trimethyl-dihydroquinoline butylated hydroxytoluene (BHT) tetradecanoic acid dibutyl phthalate hexadecanoic acid oleic acid octadecanoic acid dibutoxyethyl sebacate dibutoxyethoxyethyl adipate
352 71 68 26 49 191 287 176 631 12657
a
Area counts given in millions.
tion of the peaks based on library matching of the mass spectra, and area counts are all given in Table 1. Compounds are identified by an index letter in the Tables and Figures to make them more easily located in the chromatograms. As Table 1 shows, there are approximately 10 major peaks that appear in the chromatogram. Most of these peaks correspond to species that are typically used in the manufacture and/or processing of rubber materials. For example, 2-butoxyethoxy ethanol and BHT are added as stabilizers and processing aids in polymeric materials. Trimethyl dihydroquinoline is marketed as a commercial additive under such names as Flectol, Vulkanox, and NocracRW. Phthalates, sebacates, and adipates are commonly used plasticizers, typically to provide low temperature performance in nitrile rubber.14 The carboxylic acids are also commonly used as processing aids to the vulcanization process, as softeners, and as dispersing agents and filler/flow
modifiers, and are typically present in rubber compositions.15 The peaks listed in Table 1 were specifically monitored to assess the degree to which the fuels affect the composition of the O-ring. Further, the relatively flat regions between these major peaks were monitored to determine whether species from the fuel were taken up by the O-ring, then released during the desorption process. This, too, may have an effect on the performance of the O-ring. O-Rings Soaked for 1 Day. Figure 1 shows an overlay of the chromatograms of the original dry O-ring along with chromatograms for o-rings soaked for 1 day in S-5, JP-5, and S-5 spiked with 1% (vol) benzyl alcohol. This overlay was used to identify and evaluate the removal of certain species from the O-ring by the fuel, as well as the appearance of new peaks that can be attributed to components of the fuel being absorbed by the O-ring. Table 2 lists the peaks that were identified in the dry O-ring and their corresponding peak areas after soaking for 1 day in various fuel test mixtures, as well as new peaks (shown in bold) that appeared as a result of components from the fuel being released by the O-ring. The most obvious change is the appearance of the broad peak from 7.0 to 7.8 min in the S5+BzOH-soaked sample. This peak appears due to the release of a large amount of benzyl alcohol from the O-ring. Other differences of note for the o-rings soaked in the S5+BzOH compared to the dry O-ring are decreases in the peaks for 2-(2-butoxyethoxy) ethanol, BHT, dibutyl phthalate, and all of the carboxylic acid species. The peak at 24.0 (dibutoxyethyl sebacate) disappears almost completely, and the peak at 24.8-25.3 (dibutoxyethoxyethyl adipate) has a significant loss. The disappearance of a peak indicates that it is being solubilized and removed from the O-ring by the fuel. Previous experimental and computational modeling work explains why benzyl alcohol and similar compounds, such as phenol and naphthols, interact strongly with the isobutyronitrile monomer.8 The overall interaction is the result of both attraction between the phenolic hydrogen and the nitrogen atom of the isobutyronitrile monomer plus an interaction of the π-system of the aromatic moiety with the hydrogen atoms of the isobutyronitrile methyl groups. That previous work also examined the interactions between various other pure species and nitrile rubber on a more fundamental level. However, it is difficult to make specific assumptions when pure species are included as components of complex mixtures. For example, the work by Larchet shows that single component migrations may not be truly applicable, as shown by the inhibition of benzene uptake by the presence of n-heptane.16,17 For the S5-soaked o-rings, we see fewer changes in the chromatogram. Similar to the S5+BzOH, peak decreases are observed for the 2-(2-butoxyethoxy) ethanol, BHT, and dibutyl phthalate. The dibutyl phthalate appears to be removed to a lesser, albeit still significant, degree compared to soaking in either of the other fuels. Similar trends among all fuels are noted for the acid peaks and the dibutoxyethyl sebacate peak. The dibutoxyethoxyethyl adipate appears to be slightly more labile in the S5 than in the S5+BzOH. (14) Klingender, R. C., Handbook of Specialty Polymers; CRC Press: Boca Raton, FL; 2008, 572. (15) Hofmann, W., Rubber Technology Handbook; Oxford University Press: New York, 1989: 674. (16) Larchet, C.; Brun, J. P.; Guillou, M. Separation of Benzene-nHeptane Mixtures by Pervaporation with Elastomeric Membranes. I. Performance of Membranes. J. Membr. Sci. 1983, 15, 81–96. (17) Larchet, C.; Bulvestre, G.; Guillou, M. Separation of Benzenen-Heptane Mixtures by Pervaporation with Elastomeric Membranes. II. Contribution of Sorption to the Separation Mechanism. J. Membr. Sci. 1984, 17, 263–274.
860 Energy & Fuels, Vol. 23, 2009
Gormley et al.
Table 2. Peaks Originating from the Dry O-Ring Material along with Additional Peaks from Fuel Species after Soaking for 1 Day
a
retention time (min)
index letter
peak identification
area countsa (in S-5)
area countsa (in JP-5)
area countsa (in S-5+BzOH)
7.0-7.8 7.9 8.9 9.3 10.4 10.9 12.0 12.1 12.3 13.1 13.3 16.1 17.8 18.2 19.8 20.1 24.0 24.8-25.3
A B C D E F G H I J K L M N O P Q R
benzyl alcohol undecane 2-(2-butoxyethoxy) ethanol dodecane methylnaphthalene dimethyltetrahydro-naphthalene tetradecane dimethylnaphthalene trimethyl-dihydroquinoline butylated hydroxytoluene (BHT) pentadecane tetradecanoic acid dibutyl phthalate hexadecanoic acid oleic acid octadecanoic acid dibutoxyethyl sebacate dibutoxyethoxyethyl adipate
0 36 45 37 0 0 33 0 37 0 25 3 7 30 29 43 0.05 23
0 201 49 298 238 158 nd b 907 75 7 36 3 0 27 26 49 0.04 11
17540 15 56 20 0 0 34 0 18 0 25 4 3 37 22 45 0 98
Area counts given in millions. b Peak for tetradecane in JP-5 soaked O-ring is completely obscured by dimethylnaphthalene peak.
Table 3. Peaks Originating from the Dry O-Ring Material along with Additional Peaks from Fuel Species after Soaking for 7 Days
a
retention time (min)
index letter
peak identification
area countsa (in S-5)
area countsa (in JP-5)
area countsa (in S-5+BzOH)
7.0-7.8 7.9 8.9 9.3 10.4 10.9 12.0 12.1 12.3 13.1 13.3 16.1 17.8 18.2 19.8 20.1 24.0 24.8-25.3
A B C D E F G H I J K L M N O P Q R
benzyl alcohol undecane 2-(2-butoxyethoxy) ethanol dodecane methylnaphthalene dimethyltetrahydro-naphthalene tetradecane dimethylnaphthalene trimethyl-dihydroquinoline butylated hydroxytoluene (BHT) pentadecane tetradecanoic acid dibutyl phthalate hexadecanoic acid oleic acid octadecanoic acid dibutoxyethyl sebacate dibutoxyethoxyethyl adipate
0 6 54 10 0 0 27 0 28 0 28 2 0 15 22 40 0 54
0 42 25 91 136 146 nd b 334 49 5 25 3 0 21 22 42 0 25
13375 4 38 9 0 0 21 0 15 0 23 2 0 16 17 46 2 247
Area counts given in millions. b Peak for tetradecane in JP-5 soaked O-ring is completely obscured by dimethylnaphthalene peak.
When the o-rings are soaked in the JP-5, the trends in species removal are similar and follow closely those seen with the S5 and the S5+BzOH. It appears that the dibutoxyethyl sebacate peak is removed to an even greater extent in the JP-5. With the JP-5 fuel, we also see the appearance of several peaks that are not seen to any significant degree with the other mixtures. Namely, peaks in the range of 7.8 to 13.3 appear when the O-ring is soaked in JP-5. A number of these peaks (7.9, 9.3, 10.8, 12.0, 13.3) correspond to normal alkanes (C-11, C-12, C-13, C-14, and C-15, respectively). The most significant peaks that appear are for aromatic species, with large peaks at 10.4, 10.9, and 12.1 min that correspond to methylnaphthalene, dimethyltetrahydro-naphthalene, and dimethylnaphthalene, respectively. Further, the presence of aromatic species in JP-5 speaks to the phenomenon whereby aromatic species are often used as seal swelling enhancers owing to their enhanced interaction with the O-ring materials. It is believed that the appearance of the normal alkane peaks in our work has more to do with the cosolvent capabilities provided by the aromatic species being taken up by the O-ring rather than any inherent solubility of these alkane species in the O-ring itself, especially considering the pertinent solubility parameters for the alkanes.8,9,18 (18) Hansen, C. M., Hansen Solubility Parameters-A Users Handbook; CRC Press: New York, 2000.
O-Rings Soaked for 7-Days. The soaking tests for the o-rings were extended to 7 days in order to determine whether prolonged soaking would change the degree of partitioning of compounds between the o-rings and the fuels. The results for the 7-day tests are presented in Table 3 and Figure 2. Once again, the most prominent changes to the chromatograms appear in the range of 7.0 to 7.8 min for the S5+BzOHsoaked sample, where the benzyl alcohol elutes, as well as in the range from 24.0 to 25.3, where the peaks for dibutoxyethyl sebacate and dibutoxyethoxyethyl adipate show a dramatic decrease after soaking in the fuels for 7 days. The relative amounts of removal indicate that soaking for 7 days has a similar effect as soaking for 1 day. Namely, the peak at 24.0 min disappears completely in each of the fuel mixtures, while that at 24.8-25.3 is reduced by 98% in the S5+BzOH, and by over 99% in the S-5 and JP-5. Similar changes are observed for the reduction of the peaks for 2-(2-butoxyethoxy) ethanol, and BHT. In each fuel, there is a decrease in the levels of these species after soaking. Just like the 1-day soak, several peaks appear after soaking the O-ring in JP-5 fuel. Again, these peaks are mostly naphthalenes, as well as normal alkanes that have been coadsorbed by the O-ring. The most significant peaks that appear for aromatic species are at 10.4, 10.9, and 12.1 min, which
Interactions of Jet Fuels with Nitrile O-Rings
Energy & Fuels, Vol. 23, 2009 861
Table 4. Peaks Originating from the Dry O-Ring Material along with Additional Peaks from Fuel Species after Soaking for 1 h
a
retention time (min)
index letter
peak identification
area countsa (in S-5)
area countsa (in JP-5)
area countsa (in S-5+BzOH)
7.0-7.8 7.9 8.9 9.3 10.4 10.9 12.0 12.1 12.3 13.1 13.3 16.1 17.8 18.2 19.8 20.1 24.0 24.8-25.3
A B C D E F G H I J K L M N O P Q R
benzyl alcohol undecane 2-(2-butoxyethoxy) ethanol dodecane methylnaphthalene dimethyltetrahydro-naphthalene tetradecane dimethylnaphthalene trimethyl-dihydroquinoline butylated hydroxytoluene (bht) pentadecane tetradecanoic acid dibutyl phthalate hexadecanoic acid oleic acid octadecanoic acid dibutoxyethyl sebacate dibutoxyethoxyethyl adipate
0 8 72 12 0 0 21 0 23 14 12 4 1 33 32 37 72 5669
0 67 134 136 129 25 nd b 238 52 9 22 7 0 53 34 50 67 5828
12798 8 166 23 0 0 40 0 37 14 15 119 32 48 38 49 70 4521
Area counts given in millions. b Peak for tetradecane in JP-5 soaked O-ring is completely obscured by dimethylnaphthalene peak.
correspond to methylnaphthalene, dimethyltetrahydro-naphthalene, and dimethylnaphthalene, respectively. O-Rings Soaked for 1 h. Because the results for the 1-day and 7-day tests were similar, an additional set of tests was conducted with a soak time of 1 h to determine if differences could be seen. Differences would indicate how rapidly certain components are removed from or absorbed by the O-ring. The data is shown in Table 4, and the chromatograms are shown in Figure 3. In the 1-h soaks, there appear to be several areas of similarity with the 1-day and 7-day soaks. Again, a large peak was observed from 7.0 to 7.8 min in the S5+BzOH-soaked sample. It appears that the uptake of this species by the O-ring is rapid. For the JP-5-soaked sample, large peaks corresponding to the naphthalenic species (from 10.4-12.1 min) are observed, indicating that they are rapidly taken up by the O-ring. However, the uptake of dimethyltetrahydro-naphthalene is more moderate. Looking at compounds being released from the o-rings, there are differences in loss compared with the longer soaking times. For some species, such as the carboxylic acids eluting in the 18.0-20.1 min range, the losses are slightly less than the degree seen with longer soaks. However, the loss of these species is still substantial compared to their starting content in the dry O-ring. For the peak at 24.0 min (dibutoxyethyl sebacate), the loss of the species is not to the degree observed with longer soaking times. In this case, the species is being removed from the O-ring during the 1-h soak to a lesser degree than during the longer soaks; the degree of removal is consistent among all fuel mixtures. This is also the case for the adipate, in the 24.7-25.3 min range. A broad peak with large response still exists for all three fuels in this region. Similarly, the peaks for 2-(2-butoxyethoxy) ethanol, at 9.0 min, and for BHT, at 13.0
min, do not decrease as significantly during the initial hour. It is evident that the rate of release of components from the O-ring is dependent on the type of species being eluted. Slower migration of some species from the o-rings during the 1-h experiments would not be an issue during their long-term use in jet fuel systems. Conclusions In summary, benzyl alcohol, even at low concentrations of 1% (vol) in the synthetic jet fuel, preferentially partitions into the nitrile rubber O-ring. More research is required to determine the effect that such partitioning may have on the performance of o-rings in contact with a synthetic fuel additized with benzyl alcohol. On the basis of the semiquantitative analyses, the amount of fuel that is absorbed by the O-ring is greatest for the 1% benzyl alcohol in S-5, lesser for the JP-5, and least for the synthetic fuel. However, the amount and type of species extracted from the o-rings, most of which appear to be plasticizers or stabilizers, is roughly the same whether the O-ring is soaked in petroleum-derived fuel, synthetic jet fuel, or 1% benzyl alcohol in the synthetic jet fuel. Therefore, it does not appear that the presence of benzyl alcohol significantly alters the degree to which O-ring substituents are solubilized and removed from nitrile rubber o-rings. The important result shown is that the extent to which the components are removed from the o-rings is similar for various fuels and is not unique to alternative fuels. The extent to which the extracted components are replaced by the absorption of fuel and a fuel additive varies. The impact of this on O-ring performance is not known and it could be the subject of future work. EF8008037
