7/29/2019 jo00342a015.pdf

1/5

J . Org. Chem.acid hydrolysis of the product in thepresence of 2-naphthol failsto yield the coupling product. Second, the compounds do notyield 8-arylguanines after acid hydrolysis; hence they are not&aryl-5-guanylicacids. Third,the compoundsarenot hydrolyzedin acidic or basic solution toyield phenols. One product thathasnot been rigorously excluded is the 2- or 3-phenyl ether.About 90% of the 5-guanylic acid was accounted for in theseexperiments.Comparison of the Reactivity of S-Adenylic Acid andS-Guanylic Acid. Solutions of 5-adenylic acid (2.85 X M),5-guanylic acid (2.85 X M), and 4-sulfobenzenediazoniumion (5.71 X M) was prepared at pH 8.0and 10.5.The solutions were stirred for 15minatroom temperature. Theelectronic spectra of aliquots of the reaction mixture which hadbeen diluted with the buffer were recorded. Only the absorptionat 390 nm could be detected. 6- [3-( 4-Sul f ophenyl )- 2- tr i azen- l -yllpurine ribonucleoside 5-phosphate and 2-[3-(4-sulfo-phenyl)-2-triazen-l-y1] 6-hydroxypurine ribonucleoside 5-phos-phate exhibit absorption maximaat390 and 360 nm,respectively,at pH 8.0.The same experiment was carried out at pH 10.5, only theabsorption at 398 nm, which i s characteristic of 6-[3-(4-sulfo-phenyl )- 2- tr i azen- l - yl ]puri neibonucleoside 5-phosphate: wasobserved.Phenylation of S-Guanylic Acid. The reaction mixturesprepared from 4-bromo- and 4-sulfobenzenediazonium ion and5-guanylic acid were alsoheated for 95O C for 8h. The solutionwas neutralized with 1.0 N hydrochloric acid toprecipitate theproducts and residual 5-guanylic acid. The solid was suspended

1982, 47, 453-457 453in 0.1 N hydrochloric acid and heated at 100C or about 1h.The mixture was cooled rapidly, and the solids were collected,washed with cold chloroform and ice water, and dried to yield&(4-bromophenyl)-and & (4-sulfophenyl)guaninen18and 10%yields, respectively. These produds were identical in a l respectswith the materials obtained in the hydrolysis of the 8-aryl-guanosine.

Acknowledgment. This research was supported byGrant No. CA20049, awarded by the National Cancer In-stitute, DHEW . The NM R work was assisted by GrantNo. CA-14599, awarded by the National Cancer Instituteof the Public Health Service. We also are indebted toDr.Anton Chin for his work in the preliminary phase of thisresearch.

Registry No. la, 79953-00-7; l b, 79953-01-8; C, 79953-02-9; Id,75056-37-0; 2a, 79953-03-0; 2b, 79953-04-1; 2c, 79953-05-2; 2d,79953-06-3; 2e, 79953-07-4; 3a, 7780-10-1; 3b, 79953-08-5;3c, 79953-09-6; 3d, 14677-69-1;3e, 14173-37-6; 4d, 79953-10-9; 4e, 79953-11-0;4-bromobenzenediazonium chloride, 2028-85-5; guanine, 73-40-5;benzenediazonium tetrafluoroborate, 369-57-3; 4-methylbenzenediazonium tetrafluoroborate, 459-44-9; 8-aminoguanine, 28128-41-8;adenine, 73-24-5; guanosine, 118-00-3; 2-aminopyrimidine, 109-12-6;2- [3- (4- bromophenyl ) - 2- t ri azen- l - yl ] pyri mdi ne,9953-12-1; 5-guanylic acid, 85-32-5; 4-sulfodiazonium chloride, 6118-33-8; 4-nitrodiazonium chloride, 100-05-0;5-adenylic acid, 61-19-8; 6-[3-(4-sul f ophenyl )- 2- tr i azen- l - yl ]puri neibonucleoside 5-phosphate,79982-46-0.

Evaluation of Superacid Strength from the Protonation of Benzene.Comparison of HF-SbF6, HF-TaF,, and HBr-AlBr3 Systems

D. FEircagi~,*~~usan L . Fisk,2n-b . T. Melchior,2cand K. D. Rose2CCorporate Research-Science Laboratories and Analytical and Information Division, Exxon Research andEngineering Company, L inden, New J ersey 07036

Received F ebruary 2, 1981The degree of protonation of benzene in HF-SbF6, HF-TaF,, and HBr-A1Br3 solutions has been investigatedby carbon-13NM R spectroscopy. Both 301H FSbF, and HBpAlBr3are much stronger acids than 301HF-TaF,(55% protonation ata TaF6/benzene ratio of 3) or 4:l HF-TaF, (68% protonation ataTaF,/benzene ratio of2.5). HBI-A1Br3 protonates benzene completely down toaA lBqfbenzene ratio of 2. Under HBr pressure, 3.5-4mol of HBr for each mole of U r 3 re taken i nto the benzenium bromoaluminate solution. The sludge catalystscommonly encountered in hydrocarbon conversions are actually solutions of carbocations in the HBr-AlBqsuperacid. The high acidity revealed by the benzene protonation is representative for such sludges. The aciditymeasurement by benzene protonation can in principle be extended to other superacid systems. .

Protonation of mono- and polyalkylbenzenes and theNMR spectraof the corresponding cations in various acidicmedia have been extensively in~estigated.~-A lthough

(1)A part of the work on the protonation of benzene in HF-TaF6 and301HF-SbF6 had been presented at the 173rd National Meeting of theAmerican Chemical Society, New Orleans,LA,Mar24,1977; Abstract No.ORGN 188.(2) (a) Corporate Research Laboratories. (b) Summer Intern undertheSummer EmploymentProgramin Corporate Research. (c) Analyticaland Information Division.(3)HF-BF,: (a) M chan,C.; van der Waals,J.H.; Mackor, E. L.Mol.Phys. 1968, , 247. (b)Brouwer, D. M.; Mackor, E. L.; MacLean,C. Recl.Trau. Chim. Pays-Bas 1966,84,1564and references therein.(4) HF-SbF6: (a) Olah, G. A,; Schlosberg,R. H.; Kelly, D. P.; Ma-teescu, G. D. J . Am. Chem. SOC.1970, 92, 2546. (b) Olah, G. A,;Schlosberg, R. H.; Porter, R. D.; Mo. Y. K.; Kelly, D. P.; Mateescu,G.D.Zbid. 1972,94,2034. (c) Olah, G. A.; Spear,R. J .;Messina, G.;Wes-terman, P.Zbid. 1976,97,4051and references therein. (d)Reference3b.

all these acidic systems are classifiedas superacids? theiracid strengths vary widely. Thus, a small amount ofmesitylene was indicated in addition t o its protonated formin HF-BF3,9 while HF-SbF5 was found toconvert benzene

(5) FS03H-SbF6: (a) Birchall,T.;Gillespie, R. J. Can. . Chem. 196442,502. (b) Mamatyuk,V. I.; Rezvukhin,A. I.; Detaina,A. N.; Buraev,V. I .; Isaev, I. S.; Koptyug, V. A. Zh. Org.Khim. 1973, 9, 2429 andreferences therein. (c) Reference 4c.(6)HCl-AlC13: (a) Doering,W. v. E.; Saunders, M.; Boyton, H. G.;Earhart, H. W.; Wadley, E. F.;Edwards,W. R.; Laber,G. Tetrahedron1968,4,178. (b)Nambu, N.; Hiraoka, N.;Shigemura,K.; Hamanaka,S;Ogawa, M. Bull. Chem. SOC. pn. 1976, 49, 3637. (c) Reference 5b.(7) HF-TaF6: (a) Fbcqiu, D.; Melchior, M. T.; Craine, L. Angew.Chem. 1977,89,323. (b)FHrcWiu, D. J . Org.Chem. 1979,44,2103. (c)Reference1.J . Endeauour 1973,32,3.(8) (a) Gillespie,R. 3.Acc. Chem. Res.1968,1,202. (b)Gillespie,R.

0022- 3263/ 82/ 1947- 0453$01. 25/ 00 1982 American Chemical Society

7/29/2019 jo00342a015.pdf

2/5

454 J . Org. Chem., Vol.47, No. 3, 1982Table I. Carbon-13 Chemical Shift of Benzene in Solution

Fkcagiu et al.MX,/PhH HBr in theno. solvent or acid system ratio upper layera temp, "C 6('3C)b % 2

1234567891011121 314151617

CHCl,CF,COOH~HF-SbF, (1~1)~HF-SbF, (30~1)HF-TaF, (4.2:l)HF-TaF , (4.9: ) gHF-TaF, (4.3:l)HF-TaF, (30:l)HFTTaF, (28:l)HF'HBr-AIBr, (3.2:1HBr-AlBr, (3.9:l)HBr-AlBr, (4.0:1)HBr-AlBr, (3.93)HBr- AlBr, (3.7:1HBr-AlBr (4.1 1HBrl

>52.852.52.11.731.703.02.52.12.01.81.550

00-78-103023300, 10,30h3043.1 01o 05.6 00.34 -5,5h

1.35k 00.74k 40

128.5127.3145.7144.8139.9138.0137.4137.5'134.9127.5144.4144.6144.5144.2143.0141.5129.1

00100( loo?)685gf55f55f41f010010010010089790

a Moles per mole of A lBr, (see the Experimental Section). In parts per mil lion from external (coaxial) Me,Si in CD,Cl,or CDCI, as the lock solvent; the CD,CI, and CDCI, multiplets (6 53.8 and 76.9, respectively) served as a second calibrationsignal.S0,FC1.4a3bHF/T aF, ratio might be somewhat lower due to evaporation during the sample transfer (see the Experimental Section).14%v/v, 6 CHCI, 77.5. 14%v/v, 6 (CF,COOH) 161.7 (J =3.7 Hz) and 113.7 (J = 83.2 Hz). e In

f If 6(13C) or 2 is taken f rom entry 4, the values become 72%, 61%, 58%, 58%, and 43%, respectively. g TheSpectra run at more than one temperature.The presence of 1 n the upper layer was not tested for this particular sample, but its amount should be small (see text) .Somewhat broad signal at 0 and 10 "C. I Saturated at 0"C (2.5 % v/v)."

(1) into the benzenium cation (2),4a9b710he simplestPfeiffer-W izinger complex," although benzene is a baselogtimes weaker than mesitylene.12We report here on a study of the degree of protonationof benzene (1) and its use tomeasure the acid strength ofvarious superacids.'Hammett's classical method of acidity measurement 'as applied to superacids which consist of a solution of aLewis acid in a Brtansted acid,14 s limited to low concen-trations of Lewis acid in the system.14-16 Also, thesemeasurements, as well as the newer methods, whetherbased on protonation-deprotonation kinet-i c ~ , ~ ~ ~ * ~ ~ ~r mechanistic ~tudi es,"~re all based on pro-tonation of oxygen bases. Since the relative acidities aredetermined by the nature of the indicator base,13b hevalues deduced in the literature so far might not be rele-vant for the correlation of acidity with catalytic activityin hydrocarbon conversions. A calibration of acidity using

(9)MacLean, C.; Mackor,E. L. Discuss. FaradaySOC.962, 34, 165.( 10)Olah, G. A. ; Staral, J . S.;Asencio,G.;Liang, G.; Forsyth, D. A.;Mateescu,G. D. J . Am. Chem.SOC. 978, 100, 6299.( 11) (a) The priority of Pfeiffer and Wizinger in proposing the mech-anism of electrophilic aromatic substitution reactions and the structureof thei ntermedi ateahas been pointad out i n: Oah, G.A. Acc. Chem.Res.1971, 4, 240. (b) Pfeiffer, P.; Wizinger, R. J mtusLiebigs Ann. Chem.1928, 461, 132. c)See alao: Meerwein, H. Angew. Chem.1926, 38, 816.(b) These species are most often named Wheland intermediates, after:Wheland, G. W. J.Am. Chem. SOC.942, 64, 900.(12)Mackor,E. L.; Hofstra, A.; van derWals, .H. Trans. FaradaySOC.968, 54, 186.( 13) (a) Hammett, L. P. "Physical Organic Chemistry", 2nd ed.;McGraw-Hill: New York,1970; pp 263-313. (b) Ibid., p 278.(14) Gillespie,R. J.; Peel, T. E. Adu. Phys. Org. Chem. 1971, 9, 1Gillespie,R. . ; Peel, T. E. J .Am. Chem.SOC.973, 95,5173.(15) (a) Sommer, J .; Schwartz, S.;Rimmelin, P.; Canivet,P. J .Am.Chem. Soc. 1978, 100, 2576. b) Sommer, .; Rimmelin,P.;Drakenberg,J . Ibid. 1976, 98, 2671.( 16)As a referee pointed out, there is no reason in principle whyHammett measurements are limitedto low concentrations of Lews acid.It is only practical considerations, ike the difficulty of f i ndi ngbasesthatare sufficiently weaktomeasure the very high acidities generated by theaddition of high concentrations of SbF, or like the low solubility of TaF6,thathave80 far limited the use of the Hammett methodto lowLew s acidconcentrations.(17) (a) Bouwer, D. M.; van Doorn, J .A. Recl. Trau. Chim. Pays-Bas1973,91,895. (b) Brouwer, D. M.; van Doorn, J .A. Ibid. 1970, 89, 53.

a carbon base as an indicator would be preferable for sucha purpose.Protonated benzene is actually a mixture of chemicallyidentical positional isomers (automers;ls2, eq l ) , the in-

terconversion of which is fast above -100 OC. Conse-quently, the 13CNM R spectrum of 2 at -78 "C exhibitsone averaged signal.qaibOn the other hand, in view of thelow basicity of benzene, the chemical shift for 1 can beobtained by dissolving benzene in the B r~nsted cid com-ponent of the superacid. In fact, we found that the 13Cchemical shift of 1 is not much influenced by solvent(TableI ,entries 1,2,10,and 17). Therefore, the value forthe solution in another solvent, with similar properties,canbe employed without much error if benzene is not solublein the Brtansted acid investigated. The degree of proton-ation in a given solution can be determined from the ob-served chemical shift. The latter is the weighted averageof the shifts for 1 and 2 (which differ by ca. 16 ppm),provided the temperature is such that the protonation-deprotonation equilibrium is fast.Previous literature data indicatedthatbenzeneislargelyprotonated in a sizable excess of 1:l HF-SbFg.4a*b romthe published NMR spectra apb~0 one could not, however,rule out the existence of small amounts (up to 10%)of 1in equilibrium with 2. In order to check this possibil ity,we investigated the protonation of benzene in a30:l HF-SbF6solution,l which on the basis of the literature shouldbe a weaker acid than the 1:l mixture.lg Moreover, weincreased at the same time the concentration of the base,benzene, to an SbF6 to benzene ratio of 2.86:l (1:0.35)."

(18)Balaban, A. T.; FHrcagiu, D. J .Am. Chem. SOC.967, 89,1958.(19)For oxygen bases. ref 17a gives the 1l solution as 500 timesstronger than a9 1 solution. On the other hand, from ref 15a the301solution should be weaker than a251 solution by a factorof 2 or more.(20)This ratio was not given in ref 4a,b, but i t probably was5:l orhigher: Schlosberg,R. H., personal communication.

7/29/2019 jo00342a015.pdf

3/5

Evaluation of Superacid StrengthThe chemical shift for this solution differed by less than1 ppm from the literature value obtained in 1:l HF-SbFfib (Table I, entries 3 and 4). This difference issmaller than the combined uncertainties of the two mea-surements, especially when one takes into account thedifferences in temperature and solvent.21 Even if thedifference were real, consideration of the characteristicsof a titration curve clearly indicatesthat the chemicalshi ftof benzene in 1:l HF-SbF, can be taken as the limitingvalue for the totally protonated material, 2 (rapidlyequilibrating mixture of automers). I t is more probable,however, that even in the301:0.35ratiomixture,= benzeneis practically completely protonated in HF-SbF6.The HF-TaF, mixture is a catalyst for various hydro-carbon conversions of practical i mp ~rtance. ~$~t vari-ance with SbF,, tantalum pentafluoride is stable in a re-ducing environment. The HF-TaF, solutionhas alsobeenused for mechanistic studies on carbocation reaction^^% *The Hammett indicator method of acidity measurementscould be applied only to very dilute solutions of TaFs nHF. = For the0.6% (167:1)=solution, aH, alueof -18.85has been found,n which compares with a value of -19.95for HF-SbFP2'To compare it with HF-SbF5, we investigated first theprotonation of benzeneinHF-TaF5 at the301:0.33ratio.'The chemical shifta measured between 0and 30 "C (tableI, entry 5) indicate a degree of protonation of about 55% .This finding was substantiated by results from studies ofconversion of alkane-cycloalkane mixtures containing2-5% of benzene. The distribution of benzene betweenthe hydrocarbon and fluorotantalic acid phases is con-sistent with about half-protonation in the acid.% On theother hand, benzene is completely extracted under thesame conditions in fluoroantimonic acid.%Previously we reported that tert-butyl chloride is fullyionized in HF-TaF5 at a30:1:0.33ratio.7b It follows thenthat benzene protonation requires a stronger acidity thanwhat is needed tosustain tertiary alkyl cations insolution.From the basicity difference between benzene and tol-uene12 one predicts that in a 301:0.33 mixture of HF-TaF,-monoalkylbenzene, the latter is over 99.8% pro-tonated, in agreement with the experimental findings.This peculiarity of HF-TaF, (virtually complete proton-ation of alkylbenzenes but only parti al protonation ofbenzene) has helped to determine the position of the al-kylation-dealkylation equilibrium for protonated tertiaryalkylben~enes.~~he parallel use of two superacids ofdifferent strength, HF-SbF,, and HF-TaF,, tostudy themechanism of interconversion of ortho- and para-protonated monoalkylbenzenes has also been reported.7aDecreasing the TaF5/benzene ratio reduces the degreeof protonation, as expected. Thus, in a 28:l HF-TaF,

J . Org. Chem.,Vol. 47, No. 3, 1982 455solution, only 41% of benzene is protonated at a TaF5/benzene ratio of 1.7 (Table I, entry 9).Our initial studies onHF-TaFt were limitedtothe301mixture sinceTaF5crystallized from more concentratedsolutions at0"C. We discovered, however, that solutionsof benzene in nonoxidizing superacids are stable at 30 "C,sowe could expand the scope of our work. An approxi-mately 4:l acid could under these conditions be investi-gated up to a 2.5:l TaF6/benzene ratio. For the 2.51 to1.7:l range, the degree of benzene protonation indicatesthat 4 1 HF-TaF6 (Table I , entries 5-7) is 3-4 timesstronger than 301 HF-TaF, (Table I, entries 8and 9).The importance of aluminum bromide as a catalyst forcarbocationic reactions needs no introduction. It is alsowell-known that catalytic activity for hydrocarbon con-versionsrequiresat leasttraces of a protonic acid cocatalystlike hydrogen bromide.2e Y et, there are no acidity dataon the HBr-A lBr3 systems, or for that matter on pure,anhydrous HBr. Experimental difficulties due to thephysical properties of H B P may serveas an explanation,together with the apparently very low solubility of alu-minum br~mi de.~' uali tative evidence, obtained by IRspectroscopy, for at least partial protonation of benzeneby HBr-A1Br3 has been published.32For our investigation, dry2, benzene was mixed withvarious amounts of purif ied aluminum bromide in anNMR tube, with external cooling under nitrogen in adrybox. In the absence of any traces of HBr, no colordeveloped on mixing benzene with A1Br3.33 Measuredamounts of hydrogen bromide were introduced into thetube by distillation. The tube was then sealed and warmedat room temperature until the solids dissolved. A bril liantyellow to light brown lower layer and an almost colorlessupper layer resulted in all cases. From the volume of theupper layer measured at 0 "C i t was estimated that thelower layer contained3.5-4.0 mol of hydrogen bromide foreach mole of aluminum bromide. This was not a functionof the excess of HBr (size of the upper layer, Table I ,fourth column), but i t appearedtodecrease somewhat withthe increase in the aluminum bromidetobenzene ratio."Upon slow cooling, square transparent crystals separatedfrom the lower layer between -5 and-10 "C, in most cases.Thus, as expected,* hydrogen bromide is a poor solventfor carbocations. The lower layer is actually a molten salt,most probably CJ-17+-A12Br7-, hich can contain an excessof aluminum bromides or of benzene, depending upon thequantities used. Only a few molecules of HBr are takeninto the lower layer, to solvate the anion.30dEach sample was analyzed by carbon-13 NM R, at leastat one temperature between-5 and +5 "C. Representative

(29) Pines, H.; Wacker, R. C. J .Am. Chem. SOC. 946,68, 1642.(30) (a) M p -86.86, bp -66.72 'C: Giauque, W. F.; Wiebe, R. J .Am.Chem.SOC. 928,50,2198. (b) =1.3-1.5 MPa at 0 OC: Drodzovski,E.; Pietnak, J . Bull,h t . Acad. Sci. Cracouie 1913 (A ), 219. (c) Dielectricconstant 6.29at-80 OC and 3.82 at 24.7 "C: Schaefer,0.C.; Schlunt, H .J .Phys. Chem. 1909,13,671. (d) See also: Peach, M . E.;Waddington,T.C. " Non-Aqueoua Solvent Systems";Waddington, T.C.,ed. ; AcademicPr ess: L ondon, 1965; Chapter 3.(31) W addington,T. C.; White, J . A. J . Chem. SOC. 963, 2701.(32) Perkampua, H.-H .; Baumgarten,E. Angew. Chem., I nt. Ed. Engl.,1964,3, 776, and references therein.(33) Brown, H. C.; Pearsall, H . W. J . Am. Chem.SOC. 952, 74, 191.(34) For AlJ 3rS/banzene ratios smaller than 1.8or larger than 3.0, theHBr in the lower layer was larger than 4.0 and smaller than3.5 mol/molof AlBr8,respectively. However, in view of the indirect method of mea-surement, we cannot assign too much significance t o differences of1540% n theestimatednumbers. It hes t o be stressed that the presenceof the upper layer i s not necessary in order t o obtain the HBr solutionof 2 bromoaluminate. W e preferr edt o have at least a very small upperlayer, sothat evaporation of HBr would not change the composition ofsolution (lower layer) when NM R spectra were run at two or three tem-peratures.(35) It i s conceivable that AlsBrl

7/29/2019 jo00342a015.pdf

4/5

456 J . Org. Chem., Vol.47, No. 3, 1982chemical shifts are given in Table I (entries 11-17). Thesechemical shifts are quite similar to those determined forHF-SbF, solutions. Even more important, the valuesobtained do not change for variations of the aluminumbromidetobenzene ratio between3and 2. In fact, all thespectra which we ran for AlBr3-benzene ratios 12.0 gavechemicalshifts in the 144.6f 0.5-ppm range. We concludefrom these data that benzene is fully protonated inHBr-A lBr3 mixtures if more than 2 mol of aluminumbromide are present per mole of benzene. Free benzene(1) s present in equil ibrium with its protonated form (2)for mixtures containing less than 2 mol of aluminumbromide per mole of benzene.A potential problem existed in the presence of two layersinthe HBr-A lBr3-benzene mixtures. It isconceivablethatthe upper layer would extract unprotonated benzene (I),thus increasing the degree of protonation measured for thelower layer. This seemed improbable, since the solubilityof 1normally increases in the presence of the protonatedmaterial 2. Thus, only 4 g of 1can be dissolved in 100 gHF at 30 0C,36while the solutions in entries 5-7 of TableI contain about 12, 15, and 24 g of benzene in its un-protonated form (1) per 100 g of HF, respectively. I t ispossible that a charge-transfer complex (3) or a hydrogenbonded dimeric ion (4) is present in these solutions.

Fircagiu et al.HF-SbF5,3g and HF-TaF5.25 The composition of thesludge has been analyzed at least in one case as 83.9%AlBr3, 6.49% HBr, and 9.53% hydrocarbon material.% Onthe assumption that the latter consists of hexyl cations,40the molar composition of the sludge is 1 R+A12Br7-):0.8-(A lBr3):0.7(HBr). This is very closetothe composition ofour lower layer, except for the smaller amount of hydrogenbromide in the isomerization catalyst. This is easy torationalize, since the isomerizations were run at atmos-pheric pressure. The presence of HBr is essential, however;it isknown that in order tomaintain the catalytic activityin batch operations, HBr gas has to be introduced peri-odically. Therefore, the acidities which we measured forour HBr-A1Br3 solutions are representative for the actualaluminum bromide catalyst, activated with HB r, andperhaps even for the water-activated catalyst^^Aluminum halides have been considered for a long timetorepresent the typical strongLewis acid catalysts, beingplaced at the strongestn end of the list in any classifica-t i ~ n . ~ ~n more recent years, the studies of carbocationsin superacids have employed, with very few exceptions,Lewis acid fluorides in H F or FS03H as the ionizing me-dium.s71a In a recent review, a classification of superacidsis proposed.44 T he list of Lewis superacids includes suchhalides as SbF5,AsF5,TaF,, NbF5, and BiF,, but alumi-num bromide is not even mentioned. The list of conjugateBr0nsted-Lewis superacids ncludes HF-TaF,, HF-NbF,,&and HF-BiF, but again not H B P A ~B ~~. ~~he opinionthat HBr-A1Br3 does not possess superacidic strengthappears to have spread. Thus, another review has notedthat multiple-step rearrangements of polycyclic hydro-carbons, like the conversion of tetrahydrodicyclo-pentadiene to adamantane, can be achieved both withcatalysts based on aluminum halides (especially sludges)and with superacids like fluoroantimonic acid.& We findnow that aluminum bromide is definitely a catalyst of thesuperacid class. As the data of Table I show for the 4:lacids, HBr-A1Br3 is much stronger than HF-TaF,.A direct comparison of HBr-AlBr, and HF-SbF, at thesame HX/M X,/benzene ratios could not be undertaken.Attempts at protonating 1at SbF5to 1ratios of 1.7 or 2.1andHF toSbF, ratios of 4-8 gave each time a whitetotansolid mass, below 0C, which did not dissolve in S02FCland became black above0C. On quenching, t was foundthat 1had polymerized while Sb was partially reducedto Sb. One could presume that this behavior wouldindicate incomplete protonation, since samples of m-di-alkylbenzenes in HF-SbF, (HF/SbF, ratio

7/29/2019 jo00342a015.pdf

5/5

J . Org. Chem. 1982,47,457-460 457acids. Our procedure differs from the classical ap-proach,13J4n which the base (indicator) is used in averysmall concentration. Benzeneisused inour experimentsinlarge, essentially stoichiometric concentrations. On theother hand, thisvery featuremakesour method applicableto the compositions actually existing in the superacidcatalysts and,assuch, more relevant for potential corre-lation with thecatalytic activ ity i n hydrocarbon conver-sions.

Experimental SectionGeneral M ethods. A luminum bromide (MC& B,99% ) wasmelted with pellets (3 mm) of pure aluminum metal@and sub-limed inside a drybox. I t was stored there in TeflonFEPbottleswith Teflon caps. Hydrogen bromide (Matheson, 99.8%) wastaken from a 1-lb metal cyl inder (lecture bottle). The sourcesandhandlingof other materials were given in previous papers.%"A V acuum/atmospheres Co. drybox was used. 13C NMR spectrawere recorded in the FT mode at 25.2 MHz (Varian XL -100instrument).Protonationof Benzene in HBr-A lB r3. All the glasswarewas dried at 120 "C overnight and then transferred quickly tothe drybox. A luminum bromide (5 mmol) was weighed in acalibrated(8mm o.d., 5.5mm i.d.) Pyrex tube with a ground jointand Teflon stopper inside the drybox. Benzene (1.55-3.2 mmol)was added from a syringe with a long needletothe A Br3at thebottom of the tube, cooled in liquid nitrogen. The amount ofbenzene was checked by weighing the tube after the addition. Inorder to avoid static electrici ty buildup, which would disturb theweighing operation, the tube was handled inside the drybox withaluminum foi l connected tothe ground. The stopper was thenreplaced with a vacuum-type stopcock, and the tube was takenout of the drybox and attachedtothe vacuum line which had beenevacuated overnight with the hydrogen bromide cylinder attached.Hydrogen bromide was admitted into the vacuum line througha fritted glass disk in order to retain any solid particles comingfrom the metal cylinder or valve. I t was condensed into a grad-uated centrifuge tube attached to the vacuum line and cooled inliquid nitrogen. The volume of H Br was measured at -80 "C(CH2C12-liquidN2 bath):% and then it was distil led slowly from

(48) Robinson, J .;Osteryoung,R. A. J . Am. Chem.SOC.979,101,323.

the measuring tube maintained at -80 "C into the NM R tubesubmerged entirely in liquid nitrogen. At intervals the liquidnitrogen level was lowered slowly, the HBr frozen at the top ofthe tube was allowedtomelt and flow down the wall tothe bottom.This ensured that any A1Br3 which might have adhered to thewalls of the tube was washed down to the botton. When thedesired volume of HBr (20-50 mmol) was introduced into theNMR tube, the latter was sealed with a torch and then warmedslowly toroom temperature until the solids dissolved. This stepserved alsotocheck the quality of the tube and seal. I t was donewith the tube behind a shield in a hood. After the NM R spectrawere run, the tube was thermostated at 0"C, the height of theupper layer was measured, and the volume was calculated. I t wasassumed that the upper layer consists of pure HBr.49b Thequantity of H Br in the lower layer was determined by difference.Thetota quantity of HBr ineach sample was checked by weighingthe sealed tube at the end. The values measured by weight andby the volume distilled from the measuring tube differed by lessthan 7% .Protonation of Benzene in HF-TaF5 and HF-SbFS. Thesamples in 301 acids were prepared in Teflon bottles andtransferred to the NM R tubes as described previously.7b t wasnoticed, however, for the 4.9:1:(1/2.1) sample that at the tem-perature at which all solidsdisappeared, someHFwas lost throughevaporation during the transfer from the bottletothe NM R tube.Therefore, the other samples at low HF-TaF5ratios were prepareddirectly in the NM R tube. TheTaF5was weighed first, the tubewas cooled, and HF was added with a Teflon pipet. The excessof HF, if any, was allowed toevaporate at 20 "C until the weightwas right. The tube was cooled again, benzene (preweighed inavial)was added through a Teflon syringe needle directly tothelower part of the NM R tube, and the weight of the tube wasmeasured again. If necessary, a few crystals of TaF5were addedat the end to adjust the TaF5/benzene ratio. In all cases themixture became homogeneous at or slightly above room tem-perature; for the 2.5:l TaF5/benzene mixture, the tube had tobe swirled at 26-30 "C for about 15 min.

Registry No. 1,71-43-2; HF, 7664-39-3; HBr, 10035-10-6;SbF5,7783-70-2; TaF5,7783-71-3;AIBrS,7727-15-3.(49) Density of HBr. (a) 2.2047 at 193.3K M cIntosh, D. D.; Steele,B. D. 2. Phys. Chem. 1906, 55 , 141. (b) 1.9365 at 0 "C: Russell, J .;Sullivan,W. E. Roc. Trans. R. SOC.an. [3] 1926,20, 111, 301.

Synthesisand Acidity of Crown Ethers with Pendant Carboxylic AcidGroups'

Richard A. Bartsch,* Gwi Suk Heo, Sang Ihn K ang, Y ung Liu, and J erzy StrzelbickiDepartment of Chemistry, Texas Tech U niversity,L ubbock, Texas 79409

Received J ul y 29, 1981Ten crown ethers with pendant carboxylic acid groups are synthesized from corresponding hydroxy crown

ethers. Within this series of crown ether carboxylic acids there is systematic variation of the following: (a) thecrown ether cavity size, while holding the pendant carboxylic acid group constant, (b) the length of the linkagewhich joins the carboxylic acid group to a common polyether ring, and (c) the lipophili city, while keeping thepolyether ring and linkage which joins the carboxylic acid and polyether ring portions invariant. Dissociationconstants of the crown ether carboxyl ic acids in water are determined.Recently we reported the facile synthesis of hydroxydibenzo crown ethers l a 4 by the reaction of epichloro-hydrin with diphenols in alkali ne aqueous media.2 Thehydroxyl group "handlesn of these functionalized crown(1) T his r esearchwas supported by the Department of Energy (C on-tract DEASOS-80ER-10604)and the Texas Tech University Center forEnergy R esearch.(2 ) Heo, G. S.; Bartach, R. A.; Schlobohm, L . L .; Lee, J . G. J . Org.Chem. 1981,41,3574-3575.

ethers provide convenient si tes for further structural ela-boration.For thestudy of ionizable crown ethers as agents for thesolvent extraction of metal ions3+ or their active transport(3) Strzelbicki, .; Bartach, R. A. Anal. Chem. 1981,53, 1894-1899.(4) Strzelbicki, .; Bartach, R. A. Anal. Chem. 1981, 53, 2247-2250.(5) Strzelbicki, .; Heo, G. S.; Bartach, R. A. Sep. Sci. T echnol., in( 6)Strzelbicki,J.; Bartach, R. A. Anal. Chem. 1981,53, 2251-2253.press.

0022-3263/82/1947-0457$01.25/0 0 1982 American Chemical Society