Strength of solid acids and acids in solution. Enhancement of acidity of centers on solid surfaces by anion stabilizing solvents and its consequence for catalysis

Article abstract of DOI:10.1021/ja9716158

A comparison of acidity of two solids, a poly(styrenesulfonic acid) (Amberlyst 15) and a perfluoroinated ion exchange polymer (Nafion-H, PFIEP) with the structurally related liquid acids methanesulfonic, sulfuric, and trifluoromethanesulfonic acid (TFMSA), was conducted with mesityl oxide as probe base (determination of the Δδ¹ parameter) and for the fluorinated materials also with hexamethylbenzene as the probe base. It was found that Nafion-H is similar in strength to 85% sulfuric acid, whereas Ambelyst 15 is much weaker than 80% methanesulfonic acid or 60% sulfuric acid. Thus, the solids are much weaker acids than their liquid structural analogs. This seems to be a general property, because the rigidity of the solids prevents the acid groups/sites from cooperating in the transfer of a hydron, an essential feature in the manifestation of superacidity. The postulation of superacidity for a number of solid acids appears to have no basis in fact. On the other hand, the acidity of the groups/sites on the surface can be increased by the interaction with a nonbasic solvent, capable of forming strong hydrogen bonds with the anion of the site (anion-stabilizing solvent). The anion-stabilizing solvent generates a new liquid phase around the acid site; for appropriate structures of the solid acid and solvent this phase can be superacidic. The acidity-enhancing effect of the anion-stabilizing solvent was found to have an important effect in boosting the catalytic activity of the solid for carbocationic reactions. A comparison of acidity of two solids, a poly(styrenesulfonic acid) (Amberlyst 15) and a perfluoroinated ion exchange polymer (Nafion-H, PFIEP) with the structurally related liquid acids methanesulfonic, sulfuric, and trifluromethanesulfonic acid (TFMSA), was conducted with mesityl oxide as probe base (determination of the Δδ¹ parameter) and for the fluorinated materials also with hexamethylbenzene as the probe base. It was found that Nafion-H is similar in strength to 85% sulfuric acid, whereas Amberlyst 15 is much weaker than 80% methanesulfonic acid or 60% sulfuric aicd. Thus, the solids are much weaker acids than their liquid structural analogs. This seems to be a general property, because the rigidity of the solids prevents the acid groups/sites from cooperating in the transfer of a hydron, an essential feature in the manifestation of superacidity. The postulation of superacidity for a number of solid acids appears to have no basis in fact. On the other hand, the acidity of the groups/sites on the surface can be increased by the interaction with a nonbasic solvent, capable of forming strong hydrogen bonds with the anion of the site (anion-stabilizing solvent). The anion-stabilizing solvent generates a new liquid phase around the acid site; for appropriate structures of the solid acid and solvent this phase can be superacidic. The acidity-enhancing effect of the anion-stabilizing solvent was found to have an important effect in boosting the catalytic activity of the solid for carbocationic reactions.

Full text of DOI:10.1021/ja9716158

11826
J. Am. Chem. Soc. 1997, 119, 11826-11831
Strength of Solid Acids and Acids in Solution. Enhancement of
Acidity of Centers on Solid Surfaces by Anion Stabilizing
Solvents and Its Consequence for Catalysis
Dan Faˇrcas¸iu,*,1a Anca Ghenciu,1a,b Gaye Marino,1c,d and Kenneth D. Rose1c
Contribution from the Department of Chemical and Petroleum Engineering, UniVersity of
Pittsburgh, Pittsburgh, PennsylVania 15261, and Corporate Research Laboratories,
Exxon Research and Engineering Company, Clinton Township, Route 22 East,
Annandale, New Jersey 08801
ReceiVed May 19, 1997. ReVised Manuscript ReceiVed September 9, 1997X
Abstract: A comparison of acidity of two solids, a poly(styrenesulfonic acid) (Amberlyst 15) and a perfluoroinated
ion exchange polymer (Nafion-H, PFIEP) with the structurally related liquid acids methanesulfonic, sulfuric, and
trifluoromethanesulfonic acid (TFMSA), was conducted with mesityl oxide as probe base (determination of the ∆δ1
parameter) and for the fluorinated materials also with hexamethylbenzene as the probe base. It was found that
Nafion-H is similar in strength to 85% sulfuric acid, whereas Ambelyst 15 is much weaker than 80% methanesulfonic
acid or 60% sulfuric acid. Thus, the solids are much weaker acids than their liquid structural analogs. This seems
to be a general property, because the rigidity of the solids prevents the acid groups/sites from cooperating in the
transfer of a hydron, an essential feature in the manifestation of superacidity. The postulation of superacidity for a
number of solid acids appears to have no basis in fact. On the other hand, the acidity of the groups/sites on the
surface can be increased by the interaction with a nonbasic solvent, capable of forming strong hydrogen bonds with
the anion of the site (anion-stabilizing solvent). The anion-stabilizing solvent generates a new liquid phase around
the acid site; for appropriate structures of the solid acid and solvent this phase can be superacidic. The acidity-
enhancing effect of the anion-stabilizing solvent was found to have an important effect in boosting the catalytic
activity of the solid for carbocationic reactions.
Introduction
reaction by cleavage of a C-H6 or a C-C bond.1c,d,7 This
matching of strength requires a good understanding of acid
strength of various catalysts.
In the acid-catalyzed conversion of an organic compound,
the acid strength of the catalyst has to be appropriately matched
with the basic strength of the substrate. For instance, carbo-
cationic reactions of alkenes can be induced in trifluoroacetic
acid solution (Ho -3), as shown in the formation of trifluoro-
acetates,2 or on slightly acidic solids. Addition of acetic acid,
however, requires the use of a small amount of stronger acid as
catalyst; otherwise, a molecular addition may occur (at higher
temperature).3 A further increase in acidity is not beneficial,
however, because it can lead to the conversion of the organic
acid to its cation, thus removing the nucleophile, and the
unwanted reaction of the carbocation with the olefin excess can
occur. This loss of olefin is the fastest at the acidity at which
the olefin is half-hydronated and leads to the formation of ill-
defined oligomeric cations with cyclopentenyl structure.4 At
an even higher acidity level, particularly in the superacid range
(assuming that the olefin can be dispersed fast enough into the
acid to avoid polymerization during mixing), the olefin is fully
converted to a persistent5 alkyl cation and no further reaction
is possible. On the other hand, a saturated hydrocarbon requires
superacidic strength of the catalyst, to initiate a carbocationic
Mechanistic representations for the main classes of reactions
in solutions have been secured for a long time.8 As the
chemistry of heterogeneous catalytic processes developed, the
same representations were applied, sometimes after a study of
the mechanism, but other times by simple analogy.9 For acid
catalysts, it was generally accepted that the carbocationic
mechanisms established in solution should intervene on the solid
surfaces as well.9 It was also accepted that the manifestation
and measurement of acidity should be the same for liquid and
solid acids. Consequently, the application of the Hammett
acidity function to solids was proposed and parameters Ho were
reported, based on the change of color of indicators adsorbed
on the surface,10 or on the change in position of the indicator
absorption band in the visible spectrum.11,12
(6) (a) Bloch, H. S.; Pines, H.; Schmerling, L. J. Am. Chem. Soc. 1946,
68, 153. (b) Nenitzescu, C. D.; Avram, M.; Sliam E. Bull. Soc. Chim. Fr.
1955, 1266. (c) Oelderik, J. M. Cited in Brouwer, D. M.; Mackor, E. L.
Proc. Chem. Soc. 1964, 147. (d) Oelderik, J. M.; Mackor, E. L.; Platteeuw,
J. C.; van der Wiel, A. British 981311 (appl. 1962); see also U.S. 3201494,
1965. (e) Brouwer, D. M.; Hogeveen, H. Prog. Phys. Org. Chem. 1972, 9,
179.
(7) (a) In HF-SbF5: Hogeveen, H.; Bickel, A. F. Chem. Commun. 1967,
635. Olah, G. A.; Lukas, J. J. Am. Chem. Soc. 1967, 89, 2227, 4739. (b) In
HF-TaF5: Faˇrcas¸iu, D.; Siskin, M.; Rhodes, R. P. J. Am. Chem. Soc. 1979,
101, 7671.
X Abstract published in AdVance ACS Abstracts, November 15, 1997.
(1) (a) University of Pittsburgh. (b) Current address: Union Carbide
Corp., Industrial Chemical Group R&D, Charleston, WV. (c) Exxon
Research and Engineering Company. (d) Current address: Merck Co.,
Rahway, NJ.
(2) Faˇrcas¸iu, D.; Marino, G.; Hsu, C. S. J. Org. Chem. 1994, 59, 163.
(3) (a) Cristol, S. J.; Morrill, T. C.; Sanchez, R. A. J. Org. Chem. 1966,
31, 2719, 2733. (b) Fahey, R. C. Top. Stereochem. 1969, 3, 253.
(4) Deno, N. C. In Carbonium Ions; Olah, G. A., Schleyer, P v. R., Eds.;
Wiley-Interscience: New York, 1970; Vol. 2, Chapter 18.
(5) Faˇrcas¸iu, D. In Nucleophilicity; Harris, J. M.; McManus, S. P., Eds.;
AdV. Chem. Ser. 215; American Chemical Society: Washington, DC, 1987;
Chapter 20.
(8) (a) Ingold, C. K. Structure and Mechanism in Organic Chemistry,
2nd ed.; Cornell University Press: Ithaca, NY, 1969. (b) Hammett, L. P.
Physical Organic Chemistry, 2nd ed.; McGraw Hill: New York, 1970.
(9) (a) Pines, H. The Chemistry of Catalytic Hydrocarbon ConVersions;
Academic Press: New York, 1981. (b) Gates, B. C.; Katzer, J. R.; Schuit,
G. C. A. Chemistry of Catalytic Processes; McGraw-Hill: New York, 1979.
(10) (a) Gayer, F. H. Ind. Eng. Chem. 1933, 25, 1933. (b) Walling, C.
J. Am. Chem. Soc. 1950, 72, 1164.
S0002-7863(97)01615-6 CCC: $14.00 © 1997 American Chemical Society

Strength of Solid Acids and Acids in Solution
J. Am. Chem. Soc., Vol. 119, No. 49, 1997 11827
with no liquid visible). The alkylation was conducted as above. At
the end of the experiment, the liquid product was decanted off, fresh
reactant mixture was addded, and the alkylation was repeated in the
same manner.
All samples were analyzed by GLC on a 3 m × 3 mm o.d. column
with 10% SP-1000 on Supelcoport. The progress of the reactions was
monitored by the disappearance of 1-hexene, through conversion to
products.
Isomerization of 1-Hexene.21 Nafion-H (0.877 mequiv acid groups/
g, 0.105 g) was placed in a 2-mL vial containing a Teflon-coated
magnetic stirring bar, then a screw-cap with a Teflon-lined septum was
tightened at the top of the vial, all inside a drybox. In the alternative
approach, the catalyst was soaked for 6 days in trifluoroacetic acid
(TFA) inside the drybox, then the liquid was removed by blowing
nitrogen over the solid and the vial was capped as before. The vial
was taken out of the drybox, 1-hexene (0.4 mL) was injected through
the septum and the mixture was stirred in a thermostated bath, at 26
°C. Samples were taken through the septum and analyzed by GLC on
a 30 m × 0.25 mm DB5 (95:5 methylphenyl-polysiloxane) capillary
column with standard coating thickness. The column was held at 36
°C for 5 min, then heated to 140 °C at 40 °C/min.
It was noted, however, that the Hammett approach was
developed for acids-as-solvents, and it is not generally applicable
even for acids in solutionn.13,14 It was shown that acidity
function studies are theoretically inapplicable to solids,15 which
explains the widely different Ho values obtained for the same
material in different studies (e.g. Nafion-H: e-12,16ab -10 to
-12,17 and -6.5;18 sulfated zirconia: <-1619 and -1211), as
well as the lack of correlation of catalytic activity with Ho.15b
Alternative methods of acidity evaluation have been reported,
but they are not adequately calibrated relative to known liquid
acids, as discussed previously.12 To elucidate the relationship
between solid acidity and acidity in solution, we conducted
experiments to determine the hydronating ability of structurally
similar solid and liquid acids, toward the same probe bases under
similar conditions, and we report our finding here. (In this study
we concern ourselves with Brønsted acid.)
Experimental Section
Materials. Purification and handling of the liquid acids, solvents,
and probe bases were described in a previous paper.14c Nafion-H (E.
I. DuPont PFIEP resin 511X) was received as the potassium salt. To
prepare the H form, the salt (25 g) was stirred with 10.5% HCl (133
mL) at 55-60 °C for 2 h, then the acid was decanted off and the
exchange was repeated twice with fresh acid. The solid was filtered
on a frit and washed with distilled water until the filtrate was neutral
(1 L of water was needed). After being dried in air, the material
contained 3.7% water (lost in the TGA below 150 °C). The residue
was dried in a vacuum oven, with the air replaced by nitrogen as residual
gas, for 11 h at 110-115 °C. Drying at higher temperature produces
darkening of the polymer. Titration with NaOH gave for various
batches of dry material an acidity content of 0.86-0.91 mequiv/g.
Amberlyst 15, purchased as the H form, was dried in the same way.
The catalytic substrates were used as purchased.
Results and Discussion
1. Comparison of Acid Strengths of Solid Acids with
Liquid Acid Analogs. If we choose two acids, I and II, which
have the same acid site (XH) tied to a radical and backbone,
respectively, of very similar structure, but one of them is a liquid
and the other a solid, it is fair to say that the comparison of I
and II gives a measure of the effect of “being a liquid” and
“being a solid” upon acidity. It is also fair to say that if the
nature and means of manifestation of acidity for liquids and
solids are the same, I and II should be at least of comparable
strengths.
It appeared immediately that the only solid acids that have
close liquid analogs are those in which the acid site is bonded
as a substituent to a backbone (Type A solid acids). Their
strength is determined by the nature of the acid group (e.g.
-SO3H > -COOH) and the electronic properties of the
backbone. An increase in the number of acid groups (reduction
in the distance between acid groups) increases the strength for
the first hydron transfer. With the exception of the acid group-
grafted22 layered metal phosphates,23 all acids in this class are
organic polymers carrying acid groups, such as sulfonic,22b,24,25
phosphonic,26 and carboxylic,22a,27,28 as substituents. The second
group of solid acids (Type B), in which the acid site is part of
the crystalline lattice, and the acidity is determined by the ability
of the lattice to provide a tridimensional delocalization of the
negative charge in the corresponding anion, like silica-alumina
or zeolites, do not have liquid analogs. Here, the reduction in
the distance between acid sites (e.g. by the increase in the Al/
Si ratio in zeolites) was said to decrease the acid strength.29,30
Analyses. NMR analyses of liquid and solid samples were
conducted as described previously,14 at 75.468 MHz and room
temperature for mesityl oxide samples and at 22.5 MHz and 50-55
°C for hexamethylbenzene samples. The spectra of the probe bases
on solid acids covered with solvents were acquired as for liquid samples.
Reaction of 1-Hexene with Toluene.20 A mixture of 1-hexene (0.25
g, 2.97 mmol), toluene (3.33 g, 36.2 mmol), and tridecane (0.206 g,
integration standard) was added to 1.02 g of Nafion-H (0.8625 mequiv
acid groups/g) in a 10 mL round-bottomed flask, also containing a
Teflon-coated stirring bar. A sample was taken and injected into the
GC as a time zero mixture, then the flask was quickly stoppered and
placed in a thermostated bath at 25.7 ( 0.3 °C. Samples were taken
and analyzed at intervals.
In other experiments, the catalyst was soaked for 4 days in a 75:25
CF3COOH-CHCl3 mixture, then the solvent was evaporated on a
vacuum line, until the solid remained only wet (10% weight increase,
(11) (a) Umanski, B. S.; Hall, W. K. J. Catal. 1990, 124, 97. (b) Umanski,
B. S.; Engelhardt, J.; Hall, W. K. J. Catal. 1991, 127, 128.
(12) Faˇrcas¸iu, D.; Ghenciu, A. Prog. NMR Spectrosc. 1996, 29, 129 and
references therein.
(13) Davis, M. M. Acid-Base BehaVior in Aprotic Organic SolVents;
NBS Monograph 105; National Bureau of Standards: Washington, DC,
1968.
(14) (a) Faˇrcas¸iu, D.; Marino, G.; Miller, G.; Kastrup, R. V. J. Am. Chem.
Soc. 1989, 111, 7210. (b) Faˇrcas¸iu, D.; Ghenciu, A.; Miller, G. J. Catal.
1992, 134, 118. (c) Faˇrcas¸iu, D.; Ghenciu, A.; Marino, G.; Kastrup, R. V.
J. Mol. Catal. 1997, 126, 141.
(15) (a) Faˇrcas¸iu, D., at the Symposium on Surface Science of Catalysis:
Strong Solid Acids; 206th National Meeting of the American Chemical
Society, Chicago, IL, Aug 26, 1993; Abstract COLL 211. (b) Faˇrcas¸iu, D.;
Ghenciu, A.; Li, J. Q. J. Catal. 1996, 158, 116.
(21) Faˇrcas¸iu, D. U.S. 4,672,147, 1987.
(22) (a) Dines, M. B.; DiGiacomo, P. M Inorg. Chem. 1981, 20, 92. (b)
DiGiacomo, P. M.; Dines, M. B. Polyhedron 1982, 1, 61.
(23) Alberti, G. Acc. Chem. Res. 1978, 11, 163.
(24) Thomas, G. G.; Davies, C. W. Nature 1927, 159, 373.
(25) Connolly, D. J.; Gresham, W. F. U.S. 3,282,875, 1966.
(26) Kapura, J. M.; Gates, B. C. Ind. Eng. Chem. Prod. Res. DeV. 1973,
12, 62.
(27) Bodamer, G.; Kunin, R. Ind. Eng. Chem. 1951, 43, 1082.
(28) Ukihashi, H.; Asawa, T.; Yamabe, M.; Miyake, H. Ger. Offen.
2,650.979, 1978.
(16) (a) Olah, G. A.; Iyer, P. S.; Surya Prakash, G. K. Synthesis 1986,
513. (b) Misono, M.; Okuhara, T. CHEMTECH 1993, 23, 23. (c) See the
many papers in the series “Catalysis by Solid Superacids” from the group
cited in ref 16a.
(17) Olah, G. A.; Surya Prakash, G. K. S.; Sommer, J. Superacids;
Wiley-Interscience: New York, 1985; p 58.
(18) Childs, R. F.; Mika-Gibala, A. J. Org. Chem. 1982, 47, 4204.
(19) Matsuhashi, H.; Hino, M.; Arata, K. Appl. Catal. 1990, 59, 205.
(20) Faˇrcas¸iu, D. U.S. 4,585,750, 1986.
(29) For example: Mortier, W. J. J. Catal. 1978, 55, 138.
(30) The existing data on the structure of acid sites of sulfated metal
oxides (Arata, K. Appl. Catal., A 1996, 146, 3) does not allow a secure
assignment of these catalysts. If the acid group is either a bisulfate molecule
bonded to the surface through one oxygen atom (possibly with secondary
coordination SdOfM) or a water molecule bonded to an adjacent Lewis
site, the material is a Type A acid. If the acidic proton is bonded to a lattice
oxygen and its strength is enhanced by an adjacent sulfate group, the material
should be considered a Type B solid acid.

11828 J. Am. Chem. Soc., Vol. 119, No. 49, 1997
Faˇrcas¸iu et al.
Table 1. Values of the ∆δ Parameter for 1 Dissolved in Liquid
An increase in the distance between acid sites beyond a certain
limit should in all cases eliminate their mutual interaction and
the acid should have the same strength in the n-th ionization as
in the first. Such independent, rather than consecutive ionization
of acid groups was observed even in solution, for polybasic
acids with very large molecules.31
Acids or Adsorbed Solid Acid Surfacesa
no.
1
acid
Ho
1/AHb
∆δb
50-51c,d
45.5
Nafion-H
1.0
1.5
2
3
4
5
5
5
5
Amberlyst15
0.368
1.000
1.070
0.728
1.000
1.038
0.643
0.982
1.000
0.540
1.000
1.033
0.523
1.000
1.020
0.522
1.000
1.010
0.548
1.000
1.036
33.6
32.4c,e
32.3
We selected for investigation two pairs of acids: methane-
sulfonic (MSA) and Amberlyst-15, a poly(styrenesulfonic acid)
cross-linked with p-divinylbenzene,32 and trifluoromethane-
sulfonic acid (TFMSA) and Nafion-H (also referred to as
perfluorinated ion exchange polymer, PFIEP), in which sulfonic
acid groups are connected to a perfuoropolymer backbone by
short perfluoroether chains. To widen the basis for comparison,
several concentrations of sulfuric acid were also examined.
Nafion-H has been repeatedly referred to as a solid superacid
and quantitative assessments of its acid strength were reported.16a,b
The numbers given did not carry, however, any reference to an
actual experiment, so it is likely that they were postulated on
the basis of the presumption of superacidity (see also the work
of Childs et al.18). The same presumption was accepted a priori
in the evaluation of acid strength of Nafion-H and other solid
acids from the broad-band NMR spectrum of adsorbed water
at 4 K,33 and from the IR spectra of adsorbed water or
acetonitrile.34 A subsequent study has shown, however, that
water is a particularly unfortunate choice of probe base.35 Also,
both equilibrium36a and rate measurements36b have shown that
there is no correlation between hydrogen bond donor ability
and acid strength, except for a series of closely related materials.
In the present work, we used probe bases for which the
positive charge in the conjugate acid is found mostly at carbon
atoms, which makes them reasonable models for the substrates
reacting with these acids in catalytic reactions.31 The probe
bases were used at near-stoichiometric ratio to the acid
molecules or sites, providing a measure of relatiVe hydronating
abilities (RHA), rather than acidity functions.14 The first series
of experiments used mesityl oxide as the probe base (1, eq 1).12
100% MSA
-7.60
-4.35
-9.00
-8.50
-7.50
-4.50
45.42
41.0c,e
40.44
40.86
38.11
38.0c,f
67.35
53.0c,e
52.39
65.72
51.0c,e
50.62
60.33
48.3c,e
47.78
44.64
39.5c,e
39.27
80.52% MSA
89.80% H2SO4
86.35% H2SO4
80.03% H2SO4
60.21% H2SO4
a For a collection of more extensive data, see ref 37. b Molar ratio
of 1 to acid or to the number of acid sites. c ∆δ1. d Broad signal.
e Interpolated value. f Extrapolated value.
with a poly(styrenesulfonic acid) was also found consistently
lower than the heat of neutralization with p-toluenesulfonic acid
in solution.39 These findings are understandable, because the
acid-base reaction is not represented by eq 2, but by eq 3,
instead.5,13,14c,40
AH + B a A- + BH+
nAH + A- a (AH)nA-
(2)
(3)
The clustering of the acid anion with excess acid was named
“homoconjugation”.41 Since this term has been used for years
to describe π-electron conjugation in systems with an interuption
of the σ skeleton (cf. homoaromaticity42), we have been using
“coperative effect” instead.5,43 Its importance for the hydron
transfer, particularly to carbon bases, has been discussed.14c,43
The rigidity of the backbone prevents the acid groups of
Nafion-H and Amberlyst15 from interacting in this manner.
Swelling in certain polar solvents, notably water, leads to chain
distortion and formation of clusters of sulfonic acid groups
imbedded in pools of solvent.44 The acidity of these pools,
however is leveled by the basicity of the solvent. It follows,
necessarily that acidity of solids is significantly lower than that
of structurally similar liquid acids.
2. Effect of Solvents on Solid Acidity. The comparison of
Nafion-H (PFIEP) with its structural analog, TFMSA, was also
conducted with hexamethylbenzene (3) as probe base (eq
4).5,14a,c It has been determined that 3 is converted to 4 in
CHCl3-SO2 solution to the extent of 5%, 16%, and 33%, for
acid to base ratios of 1.2, 2.4, and 3.6, respectively.5,14a,c By
contrast, the 13C NMR chemical shifts for the signals of 3
â
R
Me2CdCH-CO-Me + AH a
1
â
R
Me2C+-CHdC(OH)-Me + A- (1)
2
The chemical shift differences δC(â) - δC(R) (∆δ) for acid-
base ratios close to stoichiometric and the values, measured or
interpolated, for the base in the 1:1 mixture (∆δ1) for a few
acids are given in Table 1.37 It appears immediately that the
solid acids are much weaker than expected on the basis of the
strength of their liquid analogs. Thus, Amberlyst15 is much
weaker than neat MSA and even than 80% aqueous MSA or
the similarly strong 60.21% H2SO4. As for Nafion-H, far from
being a solid superacid as consistently claimed in the litera-
ture,16,17,38 it is similar in strength to 85% H2SO4. We should
point out that the heat of neutralization of amines and pyridines
(31) Faˇrcas¸iu, D.; Li, J. Q. J. Catal. 1995, 152, 198.
(32) Albright, R. L.; Jakovac, I. J. React. Polym. 1986, 4, 155.
(33) Heeribout, L.; Semmer, V.; Batamack, P.; Dore´mieux-Morin, C.;
Vincent, R.; Fraissard, J. Stud. Surf. Sci. Catal. 1996, 101, 831.
(34) Zecchina, A.; Geobaldo, F.; Spoto, G.; Bordiga, S.; Ricchiardi, G.;
Buzzoni, R.; Petrini, G. J. Phys. Chem. 1996, 100, 16584.
(35) Faˇrcas¸iu, D.; Haˆncu, D. J. Chem. Soc., Faraday Trans. 1997, 93,
2161.
(39) Arnet, E. M.; Haaksma, R. A.; Chawla, B.; Healy, M. H. J. Am.
Chem. Soc. 1986, 108, 4888.
(40) Meerwein, H. Ann. Chem. 1927, 455, 227.
(41) Kolthoff, I. M.; Chantooni, M. K. J. Am. Chem. Soc. 1965, 87, 1004.
(42) Winstein, S. J. Am. Chem. Soc. 1959, 81, 6524.
(36) (a) Faˇrcas¸iu, D.; Ghenciu, A. Catal. Lett. 1995, 31, 351. (b) Faˇrcas¸iu,
D.; Haˆncu, D. Paper in preparation.
(43) Faˇrcas¸iu, D., presented at the EUCHEM Conf. on Superacidic and
Superbasic Media, Cirencester, U.K., Sept. 13, 1984.
(37) Details are given in: Ghenciu A. PhD Thesis; University of
Pittsburgh: Pittsburgh, 1996.
(38) Akelah, A.; Sherrington, D. C. Chem. ReV. 1981, 81, 557.
(44) (a) Greenwood, A.; Block, M. Nature 1966, 212, 394. (b) Ceynowa,
J. Polymer 1978, 19, 73. (c) Sondheimer. S. J.; Bunce, N. J.; Fyfe, C. A.
J. Macromol. Sci. ReV. Macromol. Chem. Phys. 1986, C26, 353.

Strength of Solid Acids and Acids in Solution
J. Am. Chem. Soc., Vol. 119, No. 49, 1997 11829
Figure 1. 13C NMR spectra of hexamethylbenzene on Nafion-H
(PFIEP), with CDCl3 and TMS in the outer (coaxial) tube: (1) solvent,
CHCl3; (2) solvent, 75:25 TFA:CHCl3; (3) sample of spectrum 2, after
removal of the bulk solvent (see text).
deposited on Nafion-H in quantities as little as 1/4 molar to the
acid groups, or at the same ratio in a chloroform solution just
covering the solid (Figure 1, spectrum 1), were the same as on
silica gel and in the pure solvent, respectively. The total
inability of Nafion to hydronate 3, first reported years ago,43,45
is in line with the observations for mesityl oxide as base,
discussed above.
Figure 2. Variable-temperature 13C NMR spectra of hexamethylben-
zene in 75:25 TFA:CHCl3 on Nafion-H (PFIEP): (1) at 50 and (4) at
0 °C.
coalesce. The changes are reversed upon heating the sample
back to 50 °C. (Note that both spectra, 1 and 4, in Figure 2 are
expanded relative to those in Figure 1.)
AH + C6Me6 a A- + C6Me6H+
(4)
The superacid phase could be separated from the bulk liquid
phase by the following simple experiment: The tube with the
mixture (0.071 g of 3, 1.52 g of Nafion-H, 1.2 mL of mixed
solvent) was moved into the drybox, its content was poured
onto a frit, and the liquid was allowed to filter without suction.
The solid, which was still wet (the particles tended to lump
together, rather than flow freely), but with no liquid visible,
was immediately introduced into another NMR tube, capped
airtight, and its 13C NMR was run. As seen in Figure 1,
spectrum 3, the signals were weaker and somewhat broader than
in the sample before the elimination of the bulk liquid phase.
Of the aromatic carbon signals, only the low field peak (134.79
ppm) was present; of the two original solvents, the TFA signals
were still present, but the signal of chloroform was gone. Thus,
the superacid phase consists of a number of anion-stabilizing
solvent molecules, associated with the acid group on the surface,
containing dissolved, partially hydronated, base. It is not clear
whether the superacid phase is continuous and covers all the
surface or consists of isolated clusters (“droplets”) of solvent
molecules surrounding the individual acid groups.
3
4
Having established that the anion-stabilizing solvents displace
the acid-base equilibria in solution toward the ionized
species,14c,43 we checked then whether a similar effect would
be seen here. Addition of 3 (0.039 g) as a solution in a 75:25
CF3COOH-CHCl3 mixture (0.7 mL) to Nafion-H (1.01 g), had
a most remarkable consequence: each signal of 3 split in two.
This splitting is seen clearly in Figure 1, spectrum 2, for the
aromatic signal (132.66 and 135.80 ppm) and it is shown in
the resolution-enhanced spectrum for the methyl signal (15.60
and 15.85 ppm). In each pair, one of the signals has the same
chemical shift as for the system without TFA, whereas the other
is shifted to lower field, as appropriate for the rapidly exchang-
ing 3 a 4 system.
The observation of equilibrium hydronation of 3 is the
consequence of the interaction of TFA with the sulfonate anions,
as shown previously for the reaction in solution.14c This anion
stabilization by the solvent is described by eq 5. Different from
In the continuation of the experiment, the tube was again
moved into the dry-box and solvent of the same composition
was added. The 13C NMR spectrum of this mixture contained
all the signals of the original spectrum for solvents and base in
the two liquid phases (“bulk” and superacid) seen in Figure 1,
spectrum 2, except that the signals for the base were much
weaker because part of the base had been discarded with the
original bulk solution.
The phase separation experiment helped also establish that
the solvent effect observed was a surface phenomenon and was
not connected to swelling of the polymer which would have
increased the contact of the probe base with the acid groups
inside the solid. Note also that there is not much difference in
the degree of swelling by chloroform and by TFA (both are
small, only solvents which are both basic and hydroxylic, like
methanol, lead to a large degree of swelling) and there was no
difference between the results in the presence of chloroform or
with no solvent added. Only addition of TFA to the solvent
generates the superacid phase. Most important, however, if the
base had been brought inside the polymer by the solvent,
filtration could not have removed so quickly all the chloroform
and non-hydronated hexamethylbenzene (3); nor could simple
nSOH + A- a (SOH)nA-
(5)
solution is the existence of two signals, which means that there
are two types of molecules of free base 3: one type exchanges
rapidly with 4, the other does not. The two cannot coexist in
the same solution. Therefore, the anion-stabilizing solvent,
TFA, has induced a phase separation in the system, from two
phases (solid and liquid) to three phases, solid, bulk liquid, and
the third phase, which we can call the superacid phase, because
it is able to hydronate the hydrocarbon 3. This representation
of the process was also substantiated by the modifications
observed in the spectrum upon cooling the sample to 0 °C
(Figure 2, spectrum 4), when the low field signal for the
aromatic carbon disappeared, because the intramolecular hy-
drogen shift within the ion 4 has slowed down at this
temperature.14a Contrastingly, the upfield signal is somewhat
sharper at 0 °C, indicating that there is exchange between the
two signals, such that somewhere above 55 °C the signals would
(45) Faˇrcas¸iu, D.; Marino, G. M.; Kastrup. R. V.; Rose, K. D., presented
at the 185th National Meeting of the American Chemical Society, Seattle,
WA, 1983; Abstr. ORGN 160.

11830 J. Am. Chem. Soc., Vol. 119, No. 49, 1997
Faˇrcas¸iu et al.
Figure 3. 13C NMR spectra of hexamethylbenzene in 90:10 HFIP:
CHCl3 solution on Nafion-H (PFIEP), with CDCl3 and TMS in the
outer (coaxial) tube: (5) as prepared and (6) after addition of methanol.
addition of solvent afterwards reextract so quickly 3 into the
liquid phase.
When the base 3 was added as a solution in a 90:10 mixture
of hexafluoro-2-propanol (HFIP) and CHCl3 to Nafion-H, the
signals of 3 were also split, but the exchanging signal in each
pair was moved farther downfield than for the system with TFA.
Complete separation is achieved even for the methyl signals,
appearing at 15.01 and 16.47 ppm (Figure 3, spectrum 5; the
signal between those for the aromatic carbons at 133.87 and
140.69 ppm is a spike of the HFIP quartet). This behavior was
expected on the basis of our previous work, which showed that
HFIP is a better anion-stabilizing solvent than TFA and gives
a higher degree of hydronation by the acids in solution as well.14c
Addition of an excess of methanol, a much stronger base than
3, restores the spectrum of the non-hydronated 3 (Figure 3,
spectrum 6).43
3. Effect of Anion Stabilizing Solvents on the Catalytic
Activity of Solid Acids. The experiments described in the
previous section have shown that an organic base moves from
the bulk liquid to the superacid phase and back, slowly on the
NMR time scale. It remained to be seen whether this transfer
is fast enough on the chemical time scale to allow acid-catalyzed
reactions to be influenced by the increase in strength of the
acid sites by the anion-stabilizing solvent in the superacid phase.
For that purpose, we have examined two reactions, the alkylation
of toluene with 1-hexene (eq 6) and isomerization of the double
bond in 1-hexene (eq 7) on Nafion-H, both at room temperature
(26 °C).
Figure 4. Conversion of 1-hexene by reaction with excess toluene,
catalyzed by Nafion-H (PFIEP): (b) two-phase system (dry catalyst
and hydrocarbon mixture) and (O) three-phase system (catalyst wet
with HFIP and hydrocarbon mixture).
Table 2. Effect of the Anion-Stabilizing Solvent on the Alkylation
of Toluene with 1-Hexene Catalyzed by Nafion-Ha
unreacted 1-hexene, %
treated catalystb
no time h dry catalystb batch 1 batch 2 batch 3 batch 4
1
2
3
0.0
1.0
2.0
100
100
15.9d
4.7
86.7c
31.1
89.8c
59.3
87.9c
67.5e
4
5
6
7
2.25
3.0
3.5
4.25
5.5
89.2
18.2
13.3
11.9
2.7
37.6
79
8
9
19.0
2.5
10
11
12
13
20.8
28.1
48.6
74.0
41.2
34.3
25.1
2.0
4.3
2.1
14 117.9
15 140.3
16 189.3
16.2
10.0
6.6
Me-C6H5 + CH2dCH-C4H9 f Me-C6H4-C6H13
a The reaction conditions are described in the Experimental Section.
See also ref 20. b With TFA. c The initial composition includes some
material remaining from the previous batch. d At 1.05 h. e At 1.02 h.
(several isomers) (6)
CH2dCH-C4H9 f CH3CHdCH-C3H7
prepared by addition, with toluene and dry, untreated, Nafion-
H. No reaction took place.
(cis, trans) + C2H5CHdCH-C2H5 (cis, trans) (7)
The decrease in activity in Batches 1 to 4 was not caused by
loss of TFA, because addition of TFA to the catalyst did not
restore the activity. Instead, the catalyst was deactivated by
water, because the reagents were used as purchased, without
drying, and also the samples were taken by unstoppering the
reaction flask without protecting it from the atmospheric
moisture. To test this hypothesis, we added a small amount of
trifluoroacetic anhydride to the catalyst after Run 4 and reacted
another batch of hexene and toluene on it. The rate of
conversion of hexene in this experiment was the same as in
Batch 1.
Replacement of TFA with HFIP had similar results, only the
catalyst was even more active: the unreacted hexene dropped
below 5% in only 30 min. The comparison of the catalyst
treated with HFIP with the untreated catalyst in the alkylation
reaction is presented in Figure 4.
The progress of the first reaction, followed by GLC, is
expressed in Table 1 by the quantity of unreacted hexene
(limiting reactant) at a given time. In all cases, a small amount
of dialkylated product (dihexylbenzene isomers) was observed
as well. The conversion on the dry catalyst was much slower
than that on the catalyst covered with TFA. Even after three
batches of reactants had been converted with the same catalyst,
the solid acid treated with the anion-stabilizing solvent was much
better. Thus, in the reaction of Batch 4 on the latter, it took 3
h for the concentration of hexene to drop from 87.9% to 37.6%,
whereas it took 18.5 h (20.8 - 2.25) for a decrease from 89.2%
to 41.2% on the fresh, but untreated (“dry”), catalyst. In
separate experiments we determined that TFA deposited on silica
gel did not catalyze the alkylation reaction, but a slow addition
with formation of trifluoroacetate esters2 occurred. That the
trifluoroacetate esters were not intermediates in reaction was
tested by treating a mixture of 2- and 3-hexyl trifluoroacetates
When the two types of catalysts were compared in the
isomerization of 1-hexene, it was found that the dry catalyst

Strength of Solid Acids and Acids in Solution
J. Am. Chem. Soc., Vol. 119, No. 49, 1997 11831
gave 12.1% conversion (87.9% unreacted 1-hexene) after 4.35
h, whereas the TFA-covered catalyst gave 78.0% conversion
(22.0% unreacted 1-hexene) after 0.43 h and 88.0% conversion
(12% unreacted 1-hexene) after 0.8 h. In both cases only double
bond migration, with the formation of cis and trans isomers of
2- and 3-hexene, was observed, but no chain branching. Small
amounts of trifluoroacetate esters were formed as side products
on the treated catalyst.
and solvent this phase can be superacidic. Such liquid super-
acidic phases, containing carbocations in solutions, described
as ”sludges”, have been known to be formed at the surface of
catalysts like the aluminum halides and to be the locus of the
catalytic alkane conversion.46 The acidity-enhancing effect of
anion-stabilizing solvents was found to have an important effect
in boosting the catalytic activity of the solid for carbocationic
reactions. When such an assistance is not available or not
possible, an alternative mechanism may intervene in the
generation of the carbocationic or cationoidic intermediates from
the less reactive precursors like alkanes, for example one-
electron transfer,15,47 or reversible dehydrogenation by the action
of a noble metal cocatalyst.
Conclusions
The examination of two Type A solid acids together with
their structurally related liquid analogs has demonstrated that
the solids are much weaker acids than their liquid counterparts.
This seems to be a general property, because the rigidity of the
solids prevents the acid groups/sites from cooperating in the
transfer of a hydron according to eq 3, an essential feature in
the manifestation of superacidity. The postulation of super-
acidity for a number of solid acids appears to have no basis in
fact. On the other hand, the acidity of the groups/sites on the
surface can be increased by the interaction with a non-basic
solvent, capable of forming strong hydrogen bonds with the
anion of the site (anion-stabilizing solvent), according to eq 5.
The anion-stabilizing solvent generates a new liquid phase
around the acid site; for appropriate structures of the solid acid
Acknowledgment. This work was suported by a grant (CTS-
9528412) from the National Science Foundation. Support from
NSF was also obtained for the purchase of a DMX300 NMR
instrument (Grant No. CTS-9413698).
JA9716158
(46) (a) Schneider, A. U.S. 3,128,316, 1964. (b) Faˇrcas¸iu, D.; Fisk, S.
L.; Melchior, M. T.; Rose, K. D. J. Org. Chem. 1982, 47, 453 and references
therein. (c) Faˇrcas¸iu, D. Acc. Chem. Res. 1982, 15, 46. (d) Ghenciu, A.;
Faˇrcas¸iu, D. Catal. Lett. 1997, 44, 29.
(47) Ghenciu, A.; Faˇrcas¸iu, D. J. Mol. Catal. A. 1996, 109, 273.