290
Ind. Eng. Chem. Prod. Res. Dev. 1985, 2 4 , 290-293
Liquid-Phase Oxidation of Undecanal by ferf -Butyl Hydroperoxide in Dodecane George W. Mushrush," Robert N. Hazleti, and Harold 0. Eaton Chemistry Divislon, Combustion and Fuels Branch, Code 6 180, Naval Research Laboratoty, Washington, DC 20375-5000
The fen-butyl hydroperoxide (f-BHP) oxidation of undecanal in a model fuel, dodecane, at 140 "C is klnetlcally complex. I t was possible, however, to relate the distribution of the major products to a few competing reactions. The product mix was determined for several reaction periods. The yield of products varied, but not the individual products. The major lower molecular weight products included decane and ten-butyl alcohol. Gaseous products included carbon monoxide and isobutylene as major products. Higher molecular weight products included the major product, undecanoic acid, along wkh its condensation products (ten-butyl and methyl esters), ethers, peroxides, and soiventderived products. Also, an analytical method of high reproductibility has been developed which may be applicable to the study of other t-BHP oxidative processes in the liquid state.
Introduction In general, jet fuels deteriorate in quality with time. One of the significant undesirable changes is the formation of insoluble material which can plug nozzles and filters and coat heat exchanger surfaces. Fuel degradation is observed to occur under long-term, low-temperature storage conditions (storage stability) as well as short-term high-temperature stress (thermal oxidative stability) (Hazlett, 1980, Taylor, 1974; Scott, 1965; Taylor and Wallace, 1967). The latter situation is encountered under flight conditions, where fuel is forced to serve as a coolant on ita path to the combustion chamber. Although slight thermal degradation of fuel is known to occur in nonoxidizing atmospheres, the presence of oxygen will greatly accelerate the deterioration of fuel properties as well as significantly lower the temperature at which undesirable products are formed. Thus, the stability of jet fuels is frequently dependent upon the nature of potential autoxidation pathways which can take place under aircraft operating conditions. Hydrocarbon autoxidation is fairly well understood and involves, in simple terms, the following sequential steps (Howard, 1973; Benson and Shaw, 1970; Mayo, 1968; Denny and Rosen, 1962; Hiatt and Strachan, 1963; Walling, 1957; Emanuel et al., 1967). initiation propagation
- + -+
+ In Re + 0 2 ROy + R-H R-H
termination 2R02.
-+
Re
InH
RO2.
--*
R-
Re
+
ROOR
2R.-*R-R
(b)
ROOH
alcohol + ketone
RO2.
(a)
+ O2
(4 (d) (e) (f)
The initiation step, reaction a, affords an alkyl free radical. This step can be surface catalyzed (Howard and Ingold, 1969; Emanuel, 1967; Denisov, 1973). The propagation steps b and c carry the chain to the relatively stable hydroperoxide product. Step c is usually rate controlling; however, at very low oxygen concentrations (ca. 1 ppm), step b can be rate controlling. Termination steps are also oxygen dependent. Thus, step d predominates at high oxygen concentrations and steps e and f a t low concentrations. The rates of the reactions in autoxidation schemes are dependent on hydrocarbon structures, oxygen concentra-
tion and temperature (Morse, 1957; Hiatt and Irwin, 1968; Howard et al., 1969; Denisov, 1974). Catalysts and freeradical initiators can materially alter both the rate of oxidation and the product mix (Benson, 1964; Richardson, 1965). If sufficient oxygen is present, the hydroperoxide concentration will reach a high level. If the available oxygen concentration is low, but the temperature is raised, the hydroperoxide concentration will be limited by free-radical decomposition processes. This regimen (low oxygen and increasing temperature) is similar to the environment found in an aircraft fuel system. In this situation, fuel degradation can be associated with hydroperoxide formation and/or hydroperoxide decomposition. Aldehydes, as well as carbon monoxide, are among the other products observed in the decomposition of hydroperoxides (Hazlett et al., 1977). The thermal decomposition of alkyl hydroperoxides, unlike that of dialkyl peroxides, is complex (Walling, 1957; Hiatt and Irwin,1968; Gray and Williams, 1959). In alkane solvent at 100 "C or less, tert-butyl hydroperoxide decomposes at a negligible rate (Walling and Wagner, 1964). At temperatures greater than 120 "C, tert-butyl hydroperoxide decomposes rapidly by an autoinitiated pathway (Hiatt, 1980; Mosher and Durham, 1960). The major reaction pathway in the 120 "C (or higher) decomposition of tert-butyl hydroperoxide involves attack by free radicals present in the solution. The detailed mechanism of hydroperoxide decomposition is complicated since free radicals are sensitive to structural, solvent, and stereoelectronic effects. Aldehydes are known to be present in turbine fuels. Cullis et al. (1981) have shown that in the liquid phase oxidation of decane, a hydroperoxide was the primary oxidation product while other oxygen species, such as aldehydes, were secondary oxidation products. Aldehydes are smoothly converted to the corresponding carboxylic acid by a number of oxidants, including alkaline hydrogen peroxide, peracids, and molecular oxygen (Baeyer and Villiger, 1899; Lemaire and Niclause, 1965; Kenley and Traylor, 1975). This paper is concerned with the reaction between a primary autoxidation product, a hydroperoxide, and a secondary product, an aldehyde. Specifically, we examine the tert-butyl hydroperoxide oxidation of undecanal in a deareated model fuel, dodecane, at 140 "C. The reaction was studied for time periods of 15,30,60,120, and 180 min.
This article not subject to US. Copyright. Published 1985 by the American Chemical Society
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 2, 1985
Additionally, we have developed reaction conditions and an analytical method of high reproducibility which may be applicable to the study of other hydroperoxideoxidative processes.
Experimental Section Reagents. tert-Butyl hydroperoxide (t-BHP) (90%) obtained from Aldrich Chemical Co. was distilled in vacuo to 99.9% purity. Undecanal was synthesized (Org Synth 1973) and purified likewise to 99+% purity. l,4-Cyclohexadiene was obtained from Aldrich Chemical Co. and used without further purification. The dodecane solvent used (Fisher certified) was refluxed and distilled from CaH,. Method. The reactions were carried out in sealed borosilicate glass tubes. The reagents (typically 0.4 mmol of t-BHP, and 0.8 mmol of undecanal in 3.6 mmol of dodecane solvent) were weighed into 6 in. long, lI4-in. 0.d. Pyrex tubes closed at one end and fitted at the other with a stainless steel valve via a Swagelok (Teflon ferrules) fitting. The tube was attached to a vacuum system, cooled to 77 K, and subjected to several freezepumpthaw cycles. The tube was then subsequently flame-sealed below the valve. The ullage volume (0.30 mL) was kept constant for all runs. The deaerated samples were warmed to room temperature and immersed in a Cole-Parmer fluidized sand bath. The temperature (140 "C) was controlled by a Leeds and Northrup Electromax I11 temperature controller. The total pressure during each run was estimated to be less than 1 atm since the solvents and reactants all boil above 140 "C. After the reaction period, the sealed tube was cooled to 77 K and opened. The tube was capped, warmed to room temperature, and the internal standards were added. The solution was transferred to a screw cap vial (Teflon capliner) and stored at 0 "C until analysis. Since a typical chromatogram required 60 min, two internal standards were utilized. One, ethylbenzene, afforded quantitation for the peaks with short retention times, and a second, 1-phenyldodecane,for the peaks with longer retention times. Samples were heated for time periods of 15, 30,60,120 or 180 min. All tubes were subjected to the same cleaning procedure. They were filled with toluene, cleaned with a nylon brush, rinsed with toluene twice, then with methylene chloride, and dried in air at 150 "C for 8 h. A search of the literature gives a few examples of catalytic behavior with glass systems (Benson, 1964; Kirk and Knox, 1960; Hiatt et al., 1968);however, when a glass tube was partially filled with crushed Pyrex, thus increasing the surface area, the results at 140 "C for the above time periods were not substantially altered. The reactions employing the radical scavenger 1,4cyclohexadiene were run as described above except the scavenger was added in 0.05-mmol increments from 0.05 to 0.15 mmol. The samples were analyzed by two techniques, both based on gas chromatography. Peak identification for both techniques was based on retention time matching with standards and mass spectrometry. In the first, a Varian gas chromatograph Model 3700 with flame ionization detector (F.I.D.) and equipped with a 50-m 0.21-mm i.d. wall-coated open tubular (OV-101) fused silica capillary column gave the necessary resolution to distinctly separate the individual components. A carrier gas flow of 1d / m i n was combined with an inlet split ratio of 60:l and a temperature program with an initial hold at 50 "C for 8 min, a ramp of 4O/min, and a final temperature of 260 "C. In the second technique, gases formed during the reaction were analyzed using a Perkin-Elmer Model Sigma 2
291
Table I. Mole % Conversion for the Reaction of Undecanal with tert-Butyl Hydroperoxide in Dodecane Solvent a t 140 OC reaction time, min 15
30
60
120 180
tert-butyl hydroperoxide derived acetone 1.0 2.3 2.4 2.4 2.1 37.9 50.6 53.5 71.4 78.3 tert-butyl alcohol di-tert-butyl peroxide 0.1 0.2 0.1 0.1 undecanal derived octane nonane 1-decene decane decanol undecanoic acid
0.1 0.1 0.4 0.3 0.7 1.0 1.6 2.2 0.3 0.4 0.4 0.5 0.3 30.6 34.3 41.8 48.2 52.3 0.3 0.3 0.4 0.9 42.0 42.6 44.7 45.1 40.4
condensation products methyl undecanoate tert-butyl undecanoate tert-butyl decyl ether tert-butyl decyl peroxide
0.1 0.1 0.5 1.2
0.1 0.1 0.9 1.6
0.2 0.2 1.0 1.3
0.2 0.2 1.1 0.9
0.1 0.1 0.6 0.4
solvent derived dodecanones dodecanols
0.1 0.1
0.3 0.3
0.3 0.7
0.4 1.4
0.9 1.9
gaseous products carbon monoxide isobutylene methane carbon dioxide
31.7 36.2 44.3 47.4 54.7 16.7 14.2 12.2 9.4 6.2 0.8 1.4 2.1 3.0 3.1 0.7 0.7 0.9 0.9 1.3
unreacted t-BHP undecanal
36.1 26.5 21.7 13.9 13.4 10.4 4.1 1.2
5.0 0.2
gas chromatograph equipped with a 6-ft 5A Molecular Sieve column (CO and CHI) or a 4-ft Porapak/S column (CO,). In the gas analysis, the column was operated at 55 "C. The chromatogram was recorded and integrated on a Hewlett-Packard Model 3390A reporting integrator. For this procedure, the valve was left on the reaction tube and after the appropriate reaction period the tube valve was connected directly to a GC gas sampling valve via a Swagelok connection. An external standard was used for calibration. A pressure gage measured the pressure in the sample loop at the time of analysis. A material balance was assessed for each compound. The principal peaks of the chromatogram account for approximately 913'% of the original compounds. The very small peaks account for another 7-8%. The product distribution was repeatable to 2-3% for each component.
Results and Discussion The results in Table I illustrate that the product distribution from the reaction of undecanal with t-BHP in deaerated dodecane solvent can be conveniently divided into lower and higher molecule weight products. The quantities in the table are based on percent conversion from the moles of reactants originally present. Products derived solely from undecanal are calculated on the basis of starting amount of undecanal. The t-BHP derived products (for example, tert-butyl alcohol) are similarly calculated based on the starting amount of t-BHP. Mixed condensation products (i.e., t-BHP + undecanal) are calculated on the basis of moles of undecanal. From undecanal, major products were decane, carbon monoxide, and undecanoic acid; minor products were alcohols, ketones, alkenes, and shorter chain alkanes. From t-BHP, major products were tert-butyl alcohol, isobutylene, and acetone; minor products were methane and di-tert-butyl
292
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24, No. 2, 1985
peroxide. Condensation products included esters of undecanoic acid, decyl tert-butyl peroxide and decyl tertbutyl ether. Solvent participation was indicated by the formation of dodecanones, dodecanols, and other dodecyl free radical termination products. Low Molecular Weight Products. t-BHP Products. The overall decomposition of t-BHP can be portrayed as 140 O C
t-BHP
self-initiation'
(CH,),CO.
+ *OH
hydrogen
t-BHP + (CH3)3CO. abstraction (CH,),COH
+ (CH3),CO0. (i)
In aliphatic hydrocarbon solvent at temperatures of 100 " C or less, @-scissionof the tert-butoxy radical is favored over hydrogen abstraction (Walling and Wagner, 1964). At the higher temperature of this study (140 "C), @-scission was found to be less important than hydrogen abstraction. This was confirmed by a comparison of the yields of acetone (1.0% at 15 min to 3.1% at 180 min) to the yield of tert-butyl alcohol (37.9% at 15 min increasing to 78.3% at 180 min). The termination product di-tert-butyl peroxide (0.1%) was present in low yield at all reaction periods. Of the radicals generated by the processes depicted in steps g-i the tert-butyl peroxy radical (i) was probably the least reactive (Howard et al., 1969). Thus, it would be expected to selectively form termination products. The decyl tert-butyl peroxide resulted from such a process. Products that appear to be formed by termination steps involving other radicals such as the alkoxy radical (i.e., decyl tert-butyl ether) were most likely formed as a result of an SH, type reaction. Undecanal Products. The major lower molecular weight product from undecanal oxidation was decane. Its yield varied from 30.6% a t 15 min to 52.3% at 180 min of reaction. Other products included 1-decene (0.345%) nonane (0.3-2.2%), octane (0.1-0.4%), and decanol (0.3-0.9%). Minor (less than 0.1%) products observed were decyl oxirane, isomeric decanols, branched methyl decanes, and branched eicosanes. Decane, the major alkane product, forms via decarbonylation of an initial acyl radical followed by hydrogen abstraction. The lower energy required for hydrogen abstraction from an aldehyde compared to a paraffin makes the aldehyde a favored reactant with both the alkoxy and the peroxy radicals generated in the system. The acyl radical once produced decarbonylates readily.
8 u
Unsubstituted alkyl aldehydes give the highest yield of acyl radicals (Kharasch et al., 1949). The acyl and decyl radicals once produced can react by several paths to yield molecular products. The decyl radical produces decane by hydrogen abstraction. The hydrogen abstraction could be from any of several molecules. For example
Rlo. + t-BHP
+
RloH + (CH,),COO-
(m)
would yield the tert-butyl peroxy radical. Another attractive pathway would be the reaction with the starting aldehyde to yield decane and an acyl radical or the elimination of hydrogen to yield decene. The low yield of decene (0.3-0.5%) indicates that this latter path was not an important reaction. The decyl radical can also undergo termination reactions with several other radicals in the system. Reaction with the hydroxyl radical afforded alcohols. However, alcohols could also be the result of an SH2type reaction with t-BHP. Reaction with the methyl radical yielded rz-undecane and isomeric branched methyl decanes. The Cll hydrocarbons were minor products (less than 0.1%) a t all reaction times. Gaseous Products. The gaseous products formed included carbon monoxide, isobutylene, methane, and carbon dioxide. Trace, but detectable, amounts of ethane and propane were also formed. No free oxygen was observed in any of the runs. As indicated in the table, CO increased from 31.7% to 54.7% at 180 min. Isobutylene was 16.7% initially and decreased to 6.2% at 180 min. Carbon dioxide increased slightly with time but attained only 1.3% conversion. Methane increased from 0.8% at 15 min to 3.1% at 180 min of reaction. Carbon monoxide was the major gaseous product formed. Its yield closely mirrored the yield of alkanes generated in the product mix. The decyl radical once formed has, as indicated, many viable reaction paths. Consequently, the decane yield was usually a little less than the carbon monoxide yield. The other major low molecular weight product was isobutylene. This product resulted from the acid-catalyzed dehydration of tert-butyl alcohol (Hiatt, 1980). The decreasing isobutylene yield (16.7-6.2% at 180 min) was not surprising in light of the many pathways open to a reactive olefin in a radical environment. The methane yield was similar to that of acetone. This was expected since they both form via @-scissionof the alkoxy radical (reaction h). The reactive methyl radical easily abstracts hydrogen to yield methane rather than reacting with other radicals present in the system. This accounted for the low yield of branched hydrocarbons and other methyl radical derived products. Carbon dioxide was the other gaseous product. Several mechanisms could account for its formation, but the most probable was a decarboxylation of the undecanoic acid. This would account for the observed decrease in undecanoic acid as reaction time increased. No free oxygen was observed at any reaction time. The lack of measured oxygen does not mean that it was not formed. It could form in low yield and be immediately consumed by any of several pathways. Among these are a reaction with CO, or more likely, a reaction with an acyl radical to generate undecanoic acid. High Molecular Weight Products. t -BHP/Undecanal Products. The higher molecular weight products formed included the major product, undecanoic acid (42.0% at 15 min gradually increasing to 45.1% at 120 min and then decreasing to 40.4% at 180 min), and its tertbutyl (0.1%) and methyl (0.1%) esters. Other condensation products included decyl tert-butyl ether (0.5% to 1.1% at 120 min then decreasing to 0.6% at 180 min) and decyl tert-butyl peroxide (increasing from 1.2% at 15 min to 1.6% at 30 min and then gradually decreasing to 0.4% at 180 min). Undecanoic acid can arise from several mechanisms. Among these are nucleophilic reaction of the aldehyde with t-BHP, the reaction of the undecanoyl radical with mo-
Ind. Eng. Chem. Prod. Res. Dev., Vol. 24,
lecular oxygen, the reaction of the undecanoyl radical with tert-butyl peroxy radical, or the reaction of an undecanoyl radical with t-BHP to yield the acid and the alkoxy radical. A series of 15-min reactions was run as described except that a radical scavenger (1,4-cyclohexadiene) was added in 0.05-mmol increments. Under these conditions, the yield of undecanoic acid decreased to a limiting value of 11.1% at a scavenger amount of 0.15 mmol. Thus a nonzero limiting yield for this acid suggests that the radical paths to the acid were responsible for most but not all of its yield. The decreasing yield of the acid at 180 min showed that the acid was undergoing secondary reactions at this long time period. The undecanoic acid once formed can react by several avenues. Among these are the formation of esters; both the tert-butyl and methyl esters were found. Further, the a hydrogen of an acid or ester is labile to attack by free radicals. The radical formed can add to an olefin, thus leading to branched CZl acids or esters as products. These substances were found in trace amounts at all reaction times. The tert-butyl peroxy radical was the least reactive and most plentiful radical present in the system. This would make a termination step involving this radical and a decyl radical a probable reactive pathway, thus accounting for the observed peroxide product. A like termination step involving the tert-butoxy radical was not expected since the reactivity of the alkoxy radical would preclude such a step. Consequently, the tert-butyl decyl ether was most likely the result of an SH2 type reaction. The decyl radical was observed to undergo dimerization to yield a trace amount of eicosane(s). The olefin, decene, was not only involved in lower molecular weight products, but can be implicated in the higher molecular weight products as well. Decene can react by the addition of an acyl radical. C,CH=CH,
li + c,,o
0
II
C,~HCH,CC,,
(n )
The resulting ketone (11-heneicosanone) was a minor product at all reaction times, but increased slightly as reaction time progressed. Solvent-Derived Products. The products that were observed from solvent participation were dodecanone(s) and dodecanol(s). These were minor products which increased with reaction time. The dodecanones and dodecanols are reported as totals of all isomers, because the dodecanones elute in an overlapping pattern (i.e., 6, 5,4 elute as one, and the 2 and 3 isomers elute separately) as do the dodecanols. Further complicating the analysis is the co-elution of dodecanones with the dodecanols. To analyze for these substances, the alcohols were silylated with Tri-Si1 (Sweeley et al., 1963). Alcohols react to form silyl ethers which have a higher volatility and thus a different retention time permitting a separation of alcohols and ketones.
Summary Aldehydes and hydroperoxides are reactive species which are known to be present in turbine fuels. The observed deterioration of fuels can manifest itself in many ways, including the formation of insoluble deposits both in storage and in the combustion chamber. Trace quantities
No. 2, 1985 293
of compounds such as carboxylic acids have also been implicated in deposit formation. This paper specifically examined the tert-butyl hydroperoxide oxidation of undecanal in a deaerated model fuel, dodecane, at 140 "C. The variation in product mix was studied over reaction times of 15 to 180 min. A common variety of products was observed for all reaction time periods. The yield of individual components, however, varied significantly with reaction time. Major lower molecular weight products included tertbutyl alcohol and isobutylene from t-BHP, decane, and carbon monoxide from undecanal. Other observed products were methane, carbon dioxide, acetone, alkanes, olefins, and alcohols. The methane yield closely follows that of acetone. Likewise, carbon monoxide mirrors the alkane yield. The major higher molecular weight product was undecanoic acid along with its methyl and tert-butyl esters. Condensation products included decyl tert-butyl ether and decyl tert-butyl peroxide. Other observed products were di-tert-butyl peroxide and trace amounts of eicosanes, 11-heneicosanone, and decyl oxirane. We suggest that the undecanoic acid was formed by two competing reaction pathways: by a nucleophilic reaction and by a free-radical reaction. Solvent participation was noted by the formation of dodecanones and dodecanols. Registry No. t-BHP, 75-91-2; undecanal, 112-44-7; decane, 124-18-5; tert-butyl alcohol, 75-65-0; undecanoic acid, 112-37-8; tert-butyl undecanoate, 95935-43-6;methyl undecanoate, 173186-8; isobutylene, 115-11-7; carbon monoxide, 630-08-0.
Literature Cited Baeyer, A,; Villiger, V. Chem. Ber. 1699, 32,3625. Benson, S.W. J. Chem. Phys. 1964, 4 0 , 1007. Benson, S. W.; Shaw, R. "Organic Peroxides"; Wiley-Interscience: New York, 1970;Chapter 2. Cullis, C. F.; Hirschler, M. M.; Rogers, R . L. Proc. R . SOC. Lond, Ser. A 1961, 375,543. Denisov, E. T. Usp. Khim. 1973, 42, 361. Denisov, E. T. "Liquid Phase Reaction Rate Constants": IFIIPlenum: New York, 1974. Denny, D. 6.;Rosen, J. D. Tetrahedron 1982, 20, 1137. Emanuel, N. M.; Denisov, E. T.; Maizus, 2 . K. "Liquid Phase Oxidation of Hydrocarbons"; Plenum Press: New York, 1967. Gray, P.; Williams, A. Chem. Rev. 1959, 59, 239. Hazlett, R. N. "Frontiers of Free Radical Chemistry"; Academic Press: New York, 1980. Hazlett, R. N.; Hall, J. M.; Matson, M. Ind. Eng. Chem. Prod. Res. Dev. 1977, 16, 171. Hiatt, R. R. "Frontiers of Free Radical Chemistry"; Academic Press: New York, 1980. Hiatt, R. R.; Irwin, K. C. J . Org. Chem. 1966, 33, 1436. Hiatt, R. R.; Mill, T.; Mayo, F. R. J . Org. Chem. 1966, 33, 1436. Hiatt, R. R.; Strachan, W. M. J. Org. Chem. 1983, 28, 1893. Howard, J. A. "Free Radicals"; Wiley-Interscience: New York, 1973. Howard, J. A.; Adamic, K.; Ingold, K. U. Can. J. Chem. 1969, 47, 3793. Howard, J. A.; Ingold, K. U. Can. J. Chem. 1969, 47,3797. Kenley. R. A,; Traylor, T. G. J. Am. Chem. SOC. 1975, 97,4700. Kharasch, M. S.;Urry, W. H.; Kuderna, 6.M. J. Org. Chem. 1947, 14,248. Kirk, A. D.; Knox, J. H. Trans. Faraday SOC. 1960, 5 6 , 1296. Lemaire, J.; Niclause, M. C . R . Acad. Sci. 1985, 260. Mayo, F. R. Acc. Chem. Res. 1988, 7 ,193. Morse, 0. K. J. Am. Chem. SOC. 1957, 79,3375. Mosher, H. S.;Durham, L. J. J . Am. Chem. SOC. 1960, 82,4537. "Organic Synthesis"; Wiley-Interscience: New York, 1973;Vol. 5, p 872. Richardson, W. H. J. Am. Chem. SOC. 1965, 87, 1096. Scott, G. "Autoxidation"; Elsevier: Amsterdam, 1965;Chapter 3. Sweeley, C. C.; Bentley, R.; Makits, M.; Wells, W. W. J. Am. Chem. SOC. 1963, 85,2497. Taylor, W. F. Ind. Eng. Chem. Prod. Res. Dev. 1974, 13, 133. Taylor, W. F.; Wallace, T. J. Ind. Eng. Chem. Prod. Res. Dev. 1967, 6 ,
258. Walling, C.; "Free Radicals in Solution"; Wiley-Interscience:
New York,
1957. Walling, C.; Wagner, P. J. J . Am. Chem. SOC. 1964, 86,3368.
Received for review October 3, 1984 Revised manuscript received February 5, 1985 Accepted February 25, 1985
