Deprotonation of calixarenes in acetonitrile

Article abstract of DOI:10.1021/jo051312t

The pK_a values for calixarenes in MeCN have been determined by selective titration with bases using a spectroscopic method. These values are as follows: calix[4]arene pK_a(1) = 19.06 ± 0.22, pK _a(2) > 33; calix[6]arene pK

Full text of DOI:10.1021/jo051312t

Deprotonation of Calixarenes in Acetonitrile
Ian D. Cunningham* and Marc Woolfall
Chemistry, SBMS, University of Surrey, Guildford GU2 7XH, U.K.
i.cunningham@surrey.ac.uk
Received June 27, 2005
The pKa values for calixarenes in MeCN have been determined by selective titration with bases
using a spectroscopic method. These values are as follows: calix[4]arene pKa(1) ) 19.06 ( 0.22,
pKa(2) > 33; calix[6]arene pKa(1) ) 15.59 ( 0.06, pKa(2) ) 23.85 ( 0.35, pKa(3) > 33; calix[8]arene
pKa(1) ) 17.20 ( 0.20, pKa(2) ) 20.32 ( 0.31, pKa(3) > 33. The trends in acidity are rationalized
using structures generated by a DFT model. For mono-deprotonation, the degree and nature of
hydrogen bonding in the anion is the dominant factor; for di-deprotonation, spatial separation of
the anionic charges becomes important.
Introduction
sons. Selective lower-rim (i.e., hydroxy) modification is
increasingly important in modifying calixarene function,4
and this usually involves initial deprotonation; however,
much current design is empirical. Comprehensive pKa
data would allow improved rationalization of modification
methodology. In calixarene host-guest interactions,5
many guests are potential bases toward the calixarene;
pKa data would allow better assessment of the extent of
ionic (e.g., ion-pair) vs dipolar (e.g., hydrogen-bonded)
complexation. Calixarenes are increasingly being used
as selective hosts for neutral and cationic guests, and the
strength of binding is dependent on the protonation of
the calixarene;6 reliable pKa data will allow better design
of such systems. Calixarenes show a complex and varied
conformational chemistry;7 examination of pKa data can
provide information about conformational effects. Calix-
arene salts have been proposed as novel surfactants, and
Calixarenes (e.g., 1) continue to excite interest at the
forefront of chemistry.1 However, although the phenolic
hydroxy groups are responsible for many calixarene
properties, it is surprising to find no comprehensive,
systematic, and reliable pKa data for these intriguing
compounds. The most recent update of calixarene chem-
istry by Gutsche devotes only two pages to this topic (he
quotes “the accurate measurement of their pKa values
has posed some difficulties”);2 the original monograph in
1989 could manage only one page.3 Why are these pKa
data important? Apart from the academic incongruity of
science’s ignorance of a fundamental property of such
extensively studied compounds, there are further rea-
(1) (a) Alema´n, C.; den Otter, W. K.; Tolpekina, T. V.; Briels, W. J.
J. Org. Chem. 2004, 69, 951. (b) Lee, J. Y.; Kim, S. K.; Jung, J. H.;
Kim, J. S. J. Org. Chem. 2005, 70, 1463. (c) Eggert, J. P. W.;
Harrowfield, J.; Lu¨nig, U.; Skelton, B. W.; White, A. H.; Loffler, K.;
Konrad, S. Eur. J. Org. Chem. 2005, 1348. (d) Kim, S. K.; Lee, S. H.;
Lee, J. Y.; Bartsch, R. A.; Kim, J. S. J. Am. Chem. Soc. 2004, 126,
16499. (e) Notestein, J. H.; Iglesia, E.; Katz, A. J. Am. Chem. Soc. 2004,
126, 16478. (f) Asfari, Z.; Bo¨hmer, V.; Harrowfield, J.; Vicens, J.
Calixarenes 2001; Kluwer Academic: Dordrecht, The Netherlands,
2001. (g) Mandolini, L.; Ungaro, R. Calixarenes in Action; Imperial
College Press: London, 2000.
(2) Gutsche, C. D. In Calixarenes Revisited, Monographs in Su-
pramolecular Chemistry; Stoddart, J. F., Ed.; The Royal Society of
Chemistry: Cambridge, U.K., 1998.
(3) Gutsche, C. D. In Calixarenes, Monographs in Supramolecular
Chemistry; Stoddart, J. F., Ed.; The Royal Society of Chemistry:
Cambridge, U.K., 1989.
(4) (a) Webber, P. R. A.; Chen, G. Z.; Drew, M. G. B.; Beer, P. B.
Angew. Chem., Int. Ed. 2001, 40, 2265. (b) Tumcharern, G.; Tuntulani,
T.; Coles, S. J.; Hursthouse, M. B.; Kilburn, J. D. Org. Lett. 2003, 5,
4971.
(5) Brouwer, E. B.; Udachin, K. A.; Enright, G. D.; Ripmeester, J.
A. Chem. Commun. 2000, 1905.
(6) Danil de Namor, A. F.; Cleverley, R. M.; Zapata-Ormachea, M.
L. Chem. Rev. 1998, 98, 2495.
(7) (a) Gutsche, C. D. In Calixarenes Revisited, Monographs in
Supramolecular Chemistry; Stoddart, J. F., Ed.; The Royal Society of
Chemistry: Cambridge, U.K., 1998; Chapter 4. (b) Gutsche, C. D. In
Calixarenes, Monographs in Supramolecular Chemistry; Stoddart, J.
F., Ed.; The Royal Society of Chemistry: Cambridge, U.K., 1989;
Chapter 4.
10.1021/jo051312t CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/20/2005
9248
J. Org. Chem. 2005, 70, 9248-9256

Deprotonation of Calixarenes in Acetonitrile
CHART 1. Structures for Organic Bases 2-16
better pKa data would remove uncertainty over the
degree of neutralization.8 Finally, calixarenes are ideally
constructed to facilitate hydrogen bonding, both intramo-
lecularly and as a directing force in supramolecular
structural motifs.9 As one of the most important factors
in acidity is hydrogen-bonding stabilization of an anion,10
a study of calixarene acidity would give insight into the
wider aspects of hydrogen bonding.
Some pKa values have been determined for nitro or
sulfonic acid modified calixarenes.11 These groups facili-
tate pKa determination by increasing solubility and
reducing the pKa values, so for example, values of 4-6
are quoted for mono-nitrated calix[4]arenes.12 However,
data for the parent calixarenes are far more limited.
Shinkai has estimated pKa(1)water 13 values for 1a, 1b, and
1c as 4.11, 3.62, and 4.05, respectively, from titrations
with nitrophenoxide bases in THF,14 and potentiometric
titrations have yielded some values in benzonitrile
(PhCN).15 However, the Shinkai approach was indirect,
and the potentiometric method is limited to solvents in
which calixarene solubility is high (>1 mM), and so
excludes many of the most common solvents used in
organic chemistry, including water, MeCN, THF, MeOH,
EtOH, hexane, etc.
In this work, pKa values for tert-butylcalix[n]arenes (n
) 4, 6, 8) 1a, 1b, and 1c have been determined in MeCN
using a spectroscopic method. Organic bases 2-16 (see
Chart 1) of known pKa,16-18 many with accessible UV-
vis spectroscopic characteristics, were used to selectively
deprotonate the calixarenes. Selective deprotonation by
a base whose pKa is closely tuned to that of the acid in
question, e.g., H4A, H3A-, etc., is essential. Calixarenes
are multibasic acids, so if titrated with a strong base,
the various forms will exist together to an extent deter-
mined by the relative pKa(1), pKa(2), etc., values. Exami-
nation of the UV-vis spectra at different titration stages
might be expected to allow for calculation of the relative
pKa values.19 However, this approach depends very much
on there being clean, well-defined, and distinct UV-vis
spectra for the various acidic species, and this is unlikely
to be the case for the calixarenes, which tend not to show
clear trends in UV-vis properties.2,3 In summary, it is
not safe to rely on the spectroscopic changes to the
calixarene alone to provide reliable data for pKa calcula-
tion, hence the inclusion of UV-active bases. Also, the
spectroscopic method is the one of choice in situations,
as here, in which substrate solubility is very low and high
dilution is required to minimize the potential for ion-
(10) Meot-Ner, M. Chem. Rev. 2005, 105, 213.
(11) Lu¨nig, U. Adv. Phys. Org. Chem. 1995, 30, 63.
(12) Bo¨hmer, V.; Schade, E.; Vogt, W. Macromol. Rapid Commun.
1984, 5, 221.
(13) In this paper, the terms pKa(1), pKa(2), etc., refer to the constant
for first, second, etc., deprotonation in the specified solvent given as a
superscipt, e.g., water, PhCN, etc. Where no solvent is specified, the
term relates to pKa in general, or to the value in acetonitrile (MeCN).
(14) Araki, K.; Iwamoto, K.; Shinkai, S.; Matsuda, T. Bull. Chem.
Soc. Jpn. 1990, 63, 3480.
(15) (a) Danil de Namor, A. F.; Garrido Pardo, M. T.; Pacheco
Tanaka, D. A.; Sueros Velarde, F. J.; Ca´rdenas Garcia, J. D.; Cabaleiro,
M. C.; Al-Rawi, J. M. A. J. Chem. Soc., Faraday Trans. 1993, 89, 2727.
(b) Danil de Namor, A. F.; Wang, J.; Gomez Orellana, I.; Sueros
Velarde, F. J.; Pacheco Tanaka, D. A. J. Inclusion Phenom. Mol.
Recognit. Chem. 1994, 19, 371.
(16) Kaljurand, I.; Ku¨tt, A.; Soova´li, L.; Rodima, T.; Ma¨emets, V.;
Leito, I.; Koppel, I. A. J. Org. Chem. 2005, 70, 1019.
(17) Rodima, T.; Kaljurand, I.; Pihl, A.; Ma¨emets, V.; Leito, I.;
Koppel, I. A. J. Org. Chem. 2002, 67, 1873.
(8) Cunningham, I. D.; Danks, T. N.; Heyes, D. M.; Courtois, J.-P.;
Moreton, D. Colloids Surf., A 2003, 229, 137.
(9) (a) Enright, G. D.; Udachin, K. A.; Ripmeester, J. A. Chem.
Commun. 2004, 1360. (b) Udachin, K. A.; Enright, G. D.; Brown, P.
O.; Ripmeester, J. A. Chem. Commun. 2002, 2162. (c) Brouwer, E. B.;
Udachin, K. A.; Enright, G. D.; Ratcliffe, C. I.; Ripmeester, J. A. Chem.
Commun. 1998, 587.
(18) Kaljurand, I.; Rodima, T.; Leito, I.; Koppel, I. A.; Schwesinger,
R. J. Org. Chem. 2000, 65, 6202.
(19) For example, see the method used by: Leito, I.; Rodima, T.;
Koppel, I. A.; Schwesinger, R.; Vlasov, V. M. J. Org. Chem. 1997, 62,
8479.
J. Org. Chem, Vol. 70, No. 23, 2005 9249

Cunningham and Woolfall
TABLE 1. Basicity Scale of Bases in MeCN
a
base
pKa
base
pKa
2
3
4
5
6
7
8
9
33.53b
25.49
24.34
23.12
22.34
21.19
20.84
19.56
10
11
12
13
14
15
16
19.66
19.07
18.82
18.56
17.62
16.91
16.31
a Most values are taken from ref 16. These are slightly revised
values of those found in refs 17 and 18. b Value taken from ref 23.
pairing, ion-neutral molecule interaction, and calixarene
di- and trimerization.20 Furthermore, the dilute solutions
(µM) associated with the spectroscopic method allow for
the use of simple concentrations rather than activities.
FIGURE 1. UV-vis spectral change on addition of 1 equiv
of DBU 4 to calix[4]arene in MeCN.
Results
Bases. This work makes use of a large number of
organic bases 2-16 whose pKa values lie in the estimated
range of calixarene acidity (i.e.,15-33) in MeCN. These
bases have been chosen because a self-consistent basicity
scale in the 3.8-32 range for these materials (ca. 89
bases) in MeCN is available.16-18 Furthermore, many of
the bases give UV-vis spectra distinct from the calix-
arenes, and show significant spectral changes on proto-
nation to complement those of the calixarene. The bases
used in this work are illustrated, and pKa values (for
conjugate acids) are collected in Table 1. Slightly different
values are found throughout the literature for the same
bases.21-30 The main source (ref 16) is the most recent,
and the values are part of a self-consistent scale. Bases
2-4, 9, 10, 12, and 14-16 are commercially available,
whereas the remainder, 5-8, 11, and 13, were synthe-
sized according to literature procedures. The conjugate
acids were prepared and isolated as PF6-, BPh4-, or Cl-
salts, or in situ by addition of HCl to a solution of base
in MeCN.
FIGURE 2. Titration of calix[4]arene with DBU 4 at 315 nm.
Importantly, no further change occurred after the addi-
tion of the first equivalent, as shown in the titration plot
(at 315 nm) of Abs/cHA vs cB/cHA (Figure 2).
Further examination of Figure 1 reveals isosbestic
points at 282 and 265 nm, and because neither DBU 4
nor its conjugate acid absorbs at these wavelengths, these
mark the wavelengths where ꢀ(H4A) ) ꢀ(H3A-). Similar
plots were obtained for 2 and 3.
Changes due to a mono-deprotonation are also seen on
addition of a strong visible base 6. Figure 3 shows the
titration of calix[4]arene 1a by 6 at 265 nm (the isosbestic
point of H4A/H3A-). The lack of absorbance is due to
protonation of the added base by the calixarene (ꢀ(BH+)265
≈ 0), and absorbance at 265 nm is seen only after 1 equiv
(ꢀ(B)265 ) 17 000).
Titration with the bases 8-14 spanning the pKa range
20.84-17.62 located the pKa of calix[4]arene within this
range, and titration with bases 11, 12, and 13 yielded
solutions with measurable amounts of all components B,
BH+, H4A, and H3A- at equilibrium. The values of the
extinction coefficients for use in the calculations (see
Experimental Section) are readily determined for B, BH+,
and H4A. The important value for H3A- was determined
from titrations resulting in complete mono-deprotonation
using strong bases transparent in the UV-vis region of
interest (2, 3, and 4, along with their conjugate acids,
are transparent at 290 nm, where H3A- absorbs strongly;
see Figure 1). Alternatively, direct use of ꢀ(H3A-) was
Calix[4]arene. Addition of 1 equiv of the strong base
DBU 4, invisible above 265 nm, to a solution of calix[4]-
arene in MeCN at 25 °C led to the spectral change shown
in Figure 1.
This change is attributed to the conversion of the
calixarene to its anion, and the literature reports similar
changes for the addition of a base to p-allylcalix[4]arene.31
(20) Inokuchi, F.; Shinkai, S. J. Chem. Soc., Perkin Trans. 2 1996,
601.
(21) Rodima, T.; Ma¨emets, V.; Koppel, I. J. Chem. Soc., Perkin Trans.
1 2000, 2637.
(22) Leffek, K. T.; Pruszynski, P.; Thanapaalasingham, K. Can. J.
Chem. 1989, 67, 590.
(23) Kisanga, P.; Verkade, J. G.; Schwesinger, R. J. Org. Chem.
2000, 65, 5431.
(24) Schwesinger, R.; Willaredt, J.; Schlemper, H.; Keller, M.;
Schmitt, D.; Fritz, H. Chem. Ber. 1994, 127, 2435.
(25) Coetzee, J. F.; Padmanabhan, G. R. J. Am. Chem. Soc. 1965,
87, 5005.
(26) Pawlak, Z.; Zundel, G.; Fritsch, J.; Wawrzyno´w, A.; Kuna, S.;
Tusk, M. Electrochim. Acta 1984, 29, 391.
(27) Zvedina, E. A.; Zdanova, M. P.; Bren’, V. A.; Dorofeenko, G. Z.
Chem. Heterocycl. Compd. 1974, 11, 1461.
(28) Augustin-Nowacka, D.; Chmurzyn˜ski, L. Anal. Chim. Acta
1999, 381, 215.
(29) Pawlak, Z. J. Chem. Thermodyn. 1987, 19, 443.
(30) Pawlak, Z.; Urbanczyk, G. J. Mol. Struct. 1988, 177, 401.
(31) Gutsche, C. D.; Iqbal, M.; Alam, I. J. Am. Chem. Soc. 1987,
109, 4314.
9250 J. Org. Chem., Vol. 70, No. 23, 2005

Deprotonation of Calixarenes in Acetonitrile
FIGURE 5. UV-vis change on addition of excess triflic acid
to calix[8]arene in MeCN.
FIGURE 3. Titration of calix[4]arene with (phenylimino)-
tripyrrolidinophosphorane 6 at 265 nm.
Evidence for such an association, the formation of an
endo complex, was found by Gutsche for neutralization
of p-allylcalix[4]arene by the primary amine neopentyl-
amine in MeCN.31 However, we believe such a contribu-
tion here to be minimal; in particular, we point to the
fact that a similar pKa value was obtained using triethyl-
amine 12 and the highly hindered bases 11 and 13.
Although the K1 term (eq 1) is dependent only on the
relative acidities of HA and BH+, the K2 term would also
be dependent on the structure of the base, so an overall
K ) K1K2 that has a significant contribution from K2
would differ greatly for different bases, as would the pKa
value derived from it.
It is of interest that no evidence of further deprotona-
tion was found, even when the stronger (relative to
DBU) base 3 (pKa ) 25.49) or Verkade’s superbase 2
(pKa ) 33.53) was used. The pKa of the latter is close
to the accepted autoprotolysis constant of MeCN
(pKautoprotolysisMeCN > 33), so it appears that only a monoan-
ion of calix[4]arene is possible in this solvent.
Calix[8]arene. Dissolution of calix[8]arene 1c in
MeCN at 25 °C gave the spectrum shown in Figure 5. It
is noteworthy that there is a shoulder above 300 nm
similar to that seen for the mono-deprotonated calix[4]-
arene anion. A similar shoulder was also noted but not
explained by Gutsche for p-allylcalix[4]arene in EtOH or
MeCN/water.31
Titration with triflic acid (a strong acid compared to
the calixarenes, even in MeCN) led to loss of this
shoulder, and a spectrum of the form seen for calix[4]-
arene (Figure 1) and for calix[8]arene in chloroform or
dioxane.32 The shoulder at λ > 300 nm, therefore, we
attribute to some preexisting mono-deprotonated calix-
[8]arene in the base-free MeCN.33 This titration shows
the isosbestic points for H8A/H7A- to be at 291 and
257 nm.
FIGURE 4. Titration of calix[4]arene with 12 at 290 nm. The
upper and lower broken lines represent calculated plots for
complete mono-deprotonation and complete non-deprotonation,
respectively. The solid line represents the calculated plot based
on the measured value of K.
TABLE 2. Determination of pKa for
Mono-Deprotonation of Calix[4]arene 1a in MeCN
at 25 °C
base
pKa (BH+)
λ (nm)
K
pKa (H4A/H3A-)
11
12
13
19.07
18.82
18.56
290
290
265
282
290
0.53 ( 0.09
0.57 ( 0.14
0.53 ( 0.11
0.53 ( 0.06
0.78 ( 0.22
19.36 ( 0.08
19.09 ( 0.10
18.84 ( 0.09
18.84 ( 0.05
19.15 ( 0.09
19.06 ( 0.22
mean
avoided by selection of a wavelength corresponding to an
isosbestic point between H4A and H3A-. This approach
was most useful for those bases (i.e., 6, 8, 10, 11, 13, and
14) showing significant spectral changes on protonation
in the UV-vis region of interest. Titrations with bases
11, 12, and 13 (a typical plot is shown in Figure 4) yielded
the data of Table 2, and gave a pKa value for H4A of 19.06
( 0.22.
Since K is derived from the reaction between the bases
and calixarene, the possibility of an association between
the protonated base and deprotonated acid must be
considered, as per eq 1.
Titration of calix[8]arene with DBU 4 monitored at 305
nm gave a shallow increase in absorbance up to 1 equiv,
followed by a steeper increase up to 2 equiv, and then no
(32) Gutsche, C. D. In Calixarenes Revisited, Monographs in Su-
pramolecular Chemistry; Stoddart, J. F., Ed.; The Royal Society of
Chemistry: Cambridge, U.K., 1998; p 79.
(33) We are grateful to a reviewer for the suggestion that this species
might be an endo-calixarene zwitterion, wherein a H+ has left its
hydroxy group to reside within the cavity, stabilized by hydrogen
bonding to the remaining hydroxy groups.
K1
H4A + B y\
z H3A- + BH+ yK2
\
z BH+H3A-
(1)
J. Org. Chem, Vol. 70, No. 23, 2005 9251

Cunningham and Woolfall
TABLE 3. Determination of pKa for the Second
Deprotonation of Calix[8]arene 1c in MeCN at 25 °C
base
pKa (BH+) λ (nm)
K
pKa (H7A-/H6A2-
)
8
20.84
20.84
19.56
305
285
305
3.7 ( 2.0
1.9 ( 0.5
0.37 ( 0.03
20.36 ( 0.22
20.60 ( 0.15
19.99 ( 0.04
20.32 ( 0.31
9
mean
TABLE 4. Determination of pKa for the First
Deprotonation of Calix[8]arene 1c in MeCN at 25 °C
base
pKa (BH+)
λ (nm)
K
pKa (H8A/H7A-)
11
19.07
18.56
17.62
305/260
305/260
305/260
52 ( 19
39 ( 16
3.0 ( 1.5
17.4 ( 0.2
17.0 ( 0.2
17.2 ( 0.25
17.20 ( 0.20
13
14
mean
FIGURE 6. UV-vis change on addition of excess triflic acid
further change. These changes are attributed to mono-
followed by di-deprotonation of the calixarene. Since
neither the base 4 nor its conjugate acid absorbs above
260 nm, the experiments yield extinction coefficients for
H7A- and H6A2- (e.g., ꢀ305(H7A-) ) 9382, ꢀ305(H6A2-) )
17 964), and isosbestic points for H7A-/H6A2- are ob-
served at 265 and 284 nm, slightly different than those
for mono-deprotonation. The existence of two calixarene
pKa values well below that of DBU (24.34) is supported
by titrations with UV-active bases 5 (pKa ) 23.12) and 6
(pKa ) 22.34) monitored at the calixarene isosbestic point,
265 nm. These show no absorbance due to base until 2
equiv has been added.
Titration with the guanidine base 8 and pyrrolidine 9
gave a linear increase in absorbance at 305 nm up to 1
equiv of base, followed by a curved increase beyond this.
This reflects the presence of measurable amounts of
H7A-, H6A2-, B, and BH+, and allowed for calculation of
the second pKa as 20.11 ( 0.35 (Table 3). Guanidine 8
shows some absorbance at the 285 nm isosbestic point
for calixarene deprotonation, and a value determined at
this wavelength is included in Table 3.
to calix[6]arene in MeCN.
Titration at the H6A/H5A- isosbestic points of 262 and
283 nm with the large hindered base 7 (pKa ) 21.19)
showed complete mono-deprotonation of the calixarene
up to 1 equiv, followed by an increase in absorbance due
to the (strongly absorbing) base 7. Clearly, mono-depro-
tonation occurs with a pKa below 21.19. Titration with
triethylamine (12) (pKa ) 18.82) at 305 nm, where only
H5A- absorbs, also showed a change only up to 1 equiv,
with no further change after this. This yielded extinction
coefficients for H5A- (ꢀ305(H5A-) ) 5478), confirmed the
isosbestic points for H6A/H5A- at 262 and 283 nm, and
located the first pKa for calix[6]arene at well below 18.82.
Titration with the more compact base pyrrolidine (9) (pKa
) 19.56) gave a result broadly similar to that for
triethylamine (12), except that a slight linear rise in
absorbance was observed after the addition of 1 equiv;
furthermore, there was a slight shift in the isosbestic
point (ca. 2 nm). This may be indicative of some interac-
tion between the calix[6]arene monoanion and pyrrolidine
and/or its conjugate acid.
Determination of the first pKa of calix[8]arene is
complicated by the high level of preexisting mono-
deprotonated calixarene in the MeCN solution. Titration
with bases 11, 13, and 14 each showed a small further
increase in deprotonation, indicating that the first pKa
for calix[8]arene is close to those for these bases. The
equation for K is still valid, except that the simplification
of [A-] ) [BH+] no longer holds. This requires indepen-
dent measurement of [BH+] (or [B]) and [A-]. This is
possible for these three bases, since they have significant
absorption at 260 nm near the H8A/H7A- isosbestic point.
A value for the first pKa of calix[8]arene is found to be
17.20 ( 0.20 (Table 4).
Titration with the strong base 2 (pKa ) 33.53) showed
some curvature, but the final product, even after titration
with more than 3 equiv of base, was not different from
the dianion obtained using the much weaker base 4 (pKa
) 24.13). For calix[8]arene, a second deprotonation is the
maximum possible in MeCN.
As with calix[8]arene, precise determination of the first
pKa of calix[6]arene is complicated by the high level of
preexisting mono-deprotonated calixarene in the MeCN
solution. Titration at >300 nm with bases 15 and 16 each
showed a small curved increase in absorbance due to
further mono-deprotonation of calix[6]arene, indicating
that its first pKa is in this region. However, for these
titrations, although determination of [A-] was facile at
wavelengths >300 nm, independent measurement of
[BH+] (or [B]) was not possible because of a lack of
appropriate absorbance for 15 and 16 at the calixarene
isosbestic points. A modified equation for K in which
[BH+] was estimated was used,34 and pKa values for
H6A/H5A- are shown in Table 5.
(34) In this case eq 9 (see Experimental Section) for K is used. If
the interfering base is called X, then [BH+] ) [A-] - [XH+]. At cB ) 0,
the existing A- is from the deprotonation of HA by X, so KX
[XH+][A-]/{(cX - [XH+])[HA]}. However, if it is assumed that cX
)
.
[XH+] (i.e., that [X] ≈ cX), then [XH+] ) (KXcX)[HA]/[A-]. Since [XH+]
is likely to decrease as cB is added, this approximation will hold
throughout the titration. The term KXcX can be calculated from the cB
Calix[6]arene. Dissolution of calix[6]arene in MeCN
at 25 °C gave a spectrum showing, as with calix[8]arene,
a shoulder attributed to the presence of some preexisting
mono-deprotonated calix[6]arene.33 This disappeared on
titration with triflic acid (Figure 6), and isosbestic points
are observed at 262 and 283 nm.
) 0 reading where [XH+] ) [A-], and so KXcX ) [A-]2/[HA] or KXcX
)
[A-]2/(cHA - [A-]). Therefore, [XH+] can then be calculated at any
titration point using [XH+] ) (KXcX)[HA]/[A-], and from this, [BH+]
can be calculated using [BH+] ) [A-] - [XH+]. This requires a direct
measurement of [A-].
9252 J. Org. Chem., Vol. 70, No. 23, 2005

Deprotonation of Calixarenes in Acetonitrile
TABLE 5. Determination of pKa for the First
TABLE 7. pKa Values for Calixarenes in MeCN at 25 °C
Deprotonation of Calix[6]arene 1b in MeCN at 25 °C
and Calculated Gas-Phase Acidities
base
pKa (BH+)
λ (nm)
K
pKa (H8A/H7A-)
experimental pKa
∆Ea
Gb
calixarene
values
(kJ mol-1
)
(kJ mol-1
)
15
16
mean
16.91
16.31
305
310
19 ( 4
5.98 ( 0.75
15.64 ( 0.08
15.54 ( 0.05
15.59 ( 0.06
1b pKa(1)
1c pKa(1)
1a pKa(1)
1c pKa(2)
1b pKa(2)
1a pKa(2)
all pKa(3)
phenol
15.59 ( 0.06
17.20 ( 0.20
19.06 ( 0.22
20.32 ( 0.31
23.85 ( 0.35
>33
1363
1361
1396
1561
1621
1779
>1916
1524
1337
1335
1370
1535
1595
1753
>1890
1498
TABLE 6. Determination of pKa for the Second
Deprotonation of Calix[6]arene 1b in MeCN at 25 °C
.33
26.6
base
pKa (BH+)
λ (nm)
K
pKa (H8A/H7A-)
3
4
25.49
24.34
310/315
315
71 ( 34
32 ( 10
23.6 ( 0.2
24.10 ( 0.13
23.85 ( 0.35
a Difference in total energy between acid and conjugate base,
E(Hn-1A(m+1)-) - E(HnAm-); e.g., for the H-calix[6]arene pKa(2),
mean
∆E ) E(H4A2-) - E(H5A-) in kJ mol-1 b Gas-phase acidity here
.
is defined following Chipman (Chipman, D. M. J. Phys. Chem. A
ZPE
2002, 106, 7413), such that G ) GAB + GP, where GAB ) HAB
+
Measurement of a second pKa value for calix[6]arene
also proved difficult. Titration with 1 equiv of a strong
base such as 3 gave the UV-vis spectrum of the mono-
deprotonated calixarene H5A-. Further titration with 3
yielded a product with a λmax at 310 nm. This product is
assigned as the dianion, H4A2-, with a H5A-/H4A2- pKa
well above 21.19 (where titration with 7 showed only
mono-deprotonation) and just below 25.49 (the pKa of 3).
Isosbestic points at 271 and 285 nm were slightly shifted
from those for mono-deprotonation (262 and 283 nm).
Titration with 3 and the weaker base DBU 4 (pKa )
24.34) yielded a second pKa value for calix[6]arene, i.e.,
for H5A-/H4A2- (Table 6).
Titration with Verkade’s superbase 2 appeared to give
a curved increase in absorbance in the H5A-/H4A2- region
above 300 nm before leveling off after 1.5-2.0 equiv, and
then a further sharper increase after this. However, the
final spectrum, even after 4.5 equiv, was the same as that
corresponding to the dianion obtained using 3 and 4. It
is not clear to us why this base (alone) shows this
behavior, although given its pKa (33.53) there may be
some complication caused by deprotonation of the solvent.
Computational Studies. The calixarenes and their
anions were modeled using the B3LYP DFT method,
which is preferred to HF because of the significant
difference in electron distribution between the acid and
anionic forms. Given the sizes of the molecules, the tetra-
t-Bu-depleted H-calixarenes were used, as well as the
economical 6-31G* basis set. All calculated structures are
shown in the Supporting Information.
GAB + EABelec, with the three AB terms being the zero-point
nuclear vibrational energy, the thermal correction term, and the
internal electronic energy (∆E in Table 7), respectively, and GP is
the free energy of the proton at -6.3 kcal/mol (-26.3 kJ/mol)
(Hirschfelder, J. O.; Curtiss, C. F.; Bird, R. B. Molecular Theory
of Gases and Liquids; Wiley: New York, 1964). Chipman quotes
0T
ZPE
0T
the combined HAB
+ GAB as GABnuc, and it’s equal to -7
to -9 kcal/mol for a range of acids (H2O, MeOH, PhOH (-8.1
kcal/mol, -34 kJ/mol), HCO2H, MeCO2H, and PhCO2H) calculated
by HF/6-31+G* and B3LYP/6-31+G* models; the value for phenol
was used here.
1,2-alternate, or 1,3-alternate. However, it seems sensible
to assume a cone conformation, which maximizes internal
hydrogen bonding in a solvent such as MeCN, whose
ability to solvate anions is poor. The B3LYP/6-31G*
energy difference between H3A- and H4A is given as ∆E
in Table 7. The B3LYP/6-31G*-optimized cone-derived
dianion gave a much larger ∆E for H3A-/H2A2-, with that
for a 1,2-alternate dianion larger still.
For H-calix[6]arene, experimental evidence indicates
that three conformers are most common: the “pinched”
cone, the “winged” cone, and the 1,2,3-alternate.1f,3
Modeling at both MM39 and B3LYP/6-31G* levels for
H-calix[6]arene gave the pinched cone as the lowest-
energy structure, and results were similar for the monoan-
ion. Only one H-calix[6]arene dianion was chosen, which
derived from the winged-winged cone conformer, in
which the greatest cross-cone O-O separation is possible.
A calix[6]arene trianion was also based on the winged-
winged conformation, and was found to be at high energy.
Extensive conformational searching was not feasible
for H-calix[8]arene and its monoanion, so calculations
were made solely on the basis of the pleated loop
conformation.40 There is some experimental evidence that
a dianion might prefer the 1,2,3,4-alternate cone confor-
mation,41 and this is supported by MM calculations; so
this dianion was optimized at B3LYP/6-31G*. As for the
calix[6]arene, a trianion (based on a twisted loop) is much
higher in energy.
The B3LYP/6-31G* geometry-optimized structure for
H-calix[4]arene in the preferred cone conformation35,36
gives geometric parameters that compare favorably to
those found using an earlier model35 and to experimental
data.37 For the monoanion, semiempirical calculations by
Grootenhuis et al.38 indicated little difference in energies
in water for the various conformers, cone, partial cone,
(35) Bernadino, R. J.; Costa Cabral, B. J. J. Phys. Chem. A 1999,
103, 9080.
Discussion
(36) (a) Fischer, S.; Grootenhuis, P. D.; Groenen, L. C.; van Hoornm,
W. P.; van Veggel, F. C. J. M.; Reinhoudt, D. N.; Karplus, M. J. Am.
Chem. Soc. 1995, 117, 1611. (b) Harada, T.; Rudzinski, J. M.; Osawa,
E.; Shinkai, S. Tetrahedron 1993, 49, 5941. (c) Harada, T.; Oseto, F.;
Shinkai, S. Tetrahedron 1994, 50, 13377. (d) Harada, T.; Shinkai, S.
J. Chem. Soc., Perkin Trans. 2 1995, 2231. (e) Gutsche, C. D.; Bauer,
L. J. J. Am. Chem. Soc. 1985, 107, 6052.
The measured pKa values for the three calixarenes are
gathered in Table 7. The increased aqueous acidity, by
(39) Calculations at the MM level have also been carried out by:
Shinkai, S. J. Chem. Soc., Perkin Trans. 2 1995, 2231.
(40) Gutsche, C. D. In Calixarenes, Monographs in Supramolecular
Chemistry; Stoddart, J. F., Ed.; The Royal Society of Chemistry:
Cambridge, U.K., 1989; p 46.
(41) Clague, N. P.; Crane, J. D.; Sinn, E.; Young, N. A.; Clegg, W.;
Coles, S. J.; Teat, S. J.; Moreton, D. J. Chem. Commun. 1999, 379.
(37) Ungaro, R.; Pochini, A.; Andreetti, G. D.; Sangermano, V. J.
Chem. Soc., Perkin Trans. 2 1984, 1979.
(38) Grootenhuis, P. D. J.; Kollman, P. A.; Groenen, L. C.; Reinhoudt,
D. N.; van Hummel, G. J.; Ugozzoli, F.; Andreetti, G. D. J. Am. Chem.
Soc. 1990, 112, 4165.
J. Org. Chem, Vol. 70, No. 23, 2005 9253

Cunningham and Woolfall
water
TABLE 8. UV-Vis Spectrophotometric Data for
ca. 6 pKa units, of calixarenes relative to phenol (pKa
) 10.0)42 has been noted before,14 but in MeCN (phenol
Calixarenes in MeCN at 25 °C
MeCN
) 26.6)25,43 this difference is magnified to 8-11
calixarene
monoanion
dianion
pKa
λmax
λmax
λmax
(ꢀ)
calixarene
(ꢀ)
(ꢀ)
units. Clearly, in the anion-unfriendly MeCN, the im-
portance of internal hydrogen bonding is significant.
Acidity is also increased in acetonitrile relative to ben-
zonitrile solvent by 0.2-3 units.15 The computational
studies suggest a difference relative to phenol of 22-28
units in the gas phase. A semiqualitative assessment of
acidity in other solvents, perhaps suitable as a guide for
synthetic transformations, can be made by considering
these differences and the range across 1a-1c, e.g., for
mono-deprotonation in water, 0.5 unit (3.62-4.11);14
MeCN, 3.5 units (15.59-19.06) (this work); benzonitrile,
2.3 units;15 gas phase, 6 units (this work).
1a
279, 287
(10000, 8300)
280, 286
(12300, 12900)
283, 289
(22600, 28000)
315
(3800)
305 (sh)
(5700)
310 (sh)
(9200)
1b
1c
310
(10000)
310 (sh)
(17300)
though the B3LYP/6-31G* calculations do not quite
replicate this, the calculated ∆E for calix[8]arene is not
significantly lower than that for calix[6]arene, despite the
calculated structure showing an increased number of
hydrogen bonds. Clearly, the strength of the hydrogen
bond is important.
The order of acidity 1b pKa(1) < 1c pKa(1) < 1a
pKa(1), etc., can be rationalized by consideration of charge
separation and hydrogen bonding. Insight is provided by
the quantum mechanical calculations whose DFT-calcu-
lated gas-phase acidities (Table 7) show a trend in broad
The experimental ∆pKa is 1.6, corresponding to a
∆(∆E) of <9 kJ mol-1, a value well within the range
attributable to hydrogen bonding. Several factors deter-
mine the strength of a hydrogen bond, but particularly
the O-O distance. Comparison of the individual O-O
distances shows that they are in all cases shorter for
calix[6]arene than for calix[8]arene. Furthermore, for
calix[6]arene, some distances fall into the range in which
the greatest variability in strength is likely. Hibbert and
Emsley have proposed a correlation of hydrogen bond
strength with O-O distance.44 The correlation can be
divided into three categories: weak bonding of the order
of 25 kJ mol-1 that changes little over the O-O distance
of 2.80-2.55 Å, strong bonding of ca. 100 kJ mol-1 that
changes little over the range of 2.40-2.20 Å, and inter-
mediate bonding that changes dramatically from 25 to
100 kJ mol-1 between 2.55 and 2.40 Å. The critical
O-H‚‚‚O-‚‚‚H-O distances for calix[6]arene (2.49, 2.52
Å) are just within the intermediate category, so that the
small but important shortening relative to calix[8]arene
(2.54, 2.55 Å) could easily compensate for the more
numerous (but weaker) hydrogen bonds in the latter.
The variation in pKa(1) is reflected semiquantitatively
in the UV-vis spectrophotometric data (Table 8) with
the trend in ∆λmax: 1b (25, 19 nm) < 1c (27, 21 nm) <
1a (36, 28 nm). This may reflect increased delocalization
of the negative charge into the aromatic ring (to give the
keto resonance contribution) in the less extensively
hydrogen-bonded calix[4]arene compared with the calix-
[8]arene and the calix[6]arene.
As expected all the second pKa values (pKa(2): di-
deprotonation) are higher than all the first, and they span
a much wider range. No pKa(2) is observed for calix[4]-
arene 1a below ca. 33, so clearly the generation of two
negative charges within the restricted confines of the so-
called lower rim (hydroxy rim) is not possible. This is
supported by the B3LYP/6-31G* model, which calculates
the H-calix[4]arene second deprotonation to be much
higher in energy than that in H-calix[6]arene and H-calix-
[8]arene. Experimentally, the pKa(2) for calix[6]arene 1b
is observable in MeCN (at 23.85), but is significantly
higher than the first, by more than 8 units. Computa-
tionally, the winged-winged cone dianion has each -O-
agreement (making due allowance for the complexity of
MeCN
the system) with the experimental pKa
data.
For calix[4]arene, the tetra-t-Bu-depleted H-calix[4]-
arene serves as a model, and the B3LYP/6-31G* structure
shows a cyclic pattern of hydrogen bonding. The B3LYP/
6-31G* structure for the monoanion shows an increase
in the hydrogen bonding to the anionic -O- (ring A) from
the two adjacent -OH groups (rings B and D) as reflected
in the shortening of the O-O distance (see Supporting
Information for structures and tables of distances), but
a decrease in the secondary hydrogen bonding from the
furthest -OH (ring C). This is because as rings B and D
tilt their -OH groups toward the -O-, ring C is forced
to tilt outward, moving the -OH further from the -O-.
This is different than the bifurcated hydrogen bonding
for ring C proposed by Grootenhuis et al., on the basis of
semiempirical MO calculations.38 It shows that full
hydrogen bonding involvement for all -OH groups in
H3A- is not possible, and explains the relatively high pKa
of calix[4]arene.
The B3LYP/6-31G*-optimized pinched cone structure
of H-calix[6]arene also shows all six -OH groups involved
in a cycle of hydrogen bonding. Surprisingly, the changes
on mono-deprotonation are very similar in nature and
magnitude to those calculated for the H-calix[4]arene.
There is a shortening of the distances from the -O- (ring
A) to the adjacent -OH groups (rings B and F), and a
tilting of the -OH of the ring opposite the anion (ring
D) away from the -O-, although it is still just about
within the distance (2.78 Å) for hydrogen bonding to its
neighbor (ring E). So, all five -OH groups are involved
in a “relay” of hydrogen-bonding stabilization of the
anion, and it appears that the increased acidity of calix-
[6]arene relative to calix[4]arene is simply due to the
increased number of hydrogen bonding interactions in
the relayed stabilization of the anion.
However, a simple correlation of acidity with the
number of hydrogen-bonding relays in the anion cannot
be carried too far. Experimentally, the calix[8]arene,
despite the potential for more hydrogen bonds than in
calix[6]arene, shows a significantly higher pKa(1). Al-
(43) Coetzee, J. E.; Padmanabhan, G. R. J. Phys. Chem. 1965, 69,
3193.
(44) Hibbert, F.; Emsley, J. Adv. Phys. Org. Chem. 1990, 26, 267.
(42) Isaacs, N. Physical Organic Chemistry, 2nd ed.; Longman:
Harlow, U.K., 1995; p 239.
9254 J. Org. Chem., Vol. 70, No. 23, 2005

Deprotonation of Calixarenes in Acetonitrile
group (rings A and D) hydrogen-bond-stabilized by only
its two adjacent neighbors, effectively giving two separate
three-oxygen systems, and the -O-‚‚‚-O- (rings A and
D) distance is 8.45 Å.
Commercial MeCN with a supplier-stated water con-
tent of <0.0065% was used. When MeCN was dried by
reflux over CaH2 and distilled, no difference in the result
was apparent.
Trifluoromethanesulfonic acid was used as received.
Typical Procedure for UV-Vis Spectroscopic
Titrations. 4-tert-Butylcalix[4]arene (0.0073 g) was dis-
solved in MeCN (250 cm3) at 20 °C, giving a 45 µM
solution. o-Tolylbiguanide (0.0148 g) was dissolved in
MeCN (50 cm3), giving a 1.55 mM solution. 4-tert-
Butylcalix[4]arene solution (3.0 cm3) was added by pipet
to a cuvette (10 mm), and a UV spectrum was taken at
25 °C on a UV-vis spectrometer. Successive aliquots
(10 mm3) of the o-tolylbiguanide solution were added
to the stoppered cuvette using a microsyringe, and a
UV spectrum was taken after each addition. Additions
were repeated 19 times, until a total of 200 mm3 had
been added (2.3 equiv), and a titration curve was
plotted of Abs/cHA vs cB/cHA. All standard solutions were
prepared at 20 °C, and the titrations were carried out at
25 °C.
Calculation of pKa Values from UV-Vis Spectro-
scopic Measurements. For an acid-base equilibrium
involving HA, A-, B, and BH+, the equilibrium concen-
tration of any species in dilute solution can be calculated
from the observed absorbance (Abs), the various extinc-
tion coefficients (ꢀHA, etc.), and the known stoichiometry
(eqs 2-6).
In contrast to calix[6]arene, the pKa(2) for calix[8]arene
1c at 20.29 is higher than the first by only 3 units.
Consideration of the H-calix[8]arene 1,2,3,4-alternate
conformation (see Supporting Information) shows that
this dianion consists of two separate four-oxygen systems,
although the fourth -OH in each “half-system” (rings C
and G) is quite far from its neighbor (2.82 Å), and may
not contribute much to the hydrogen bonding. So it is
probably the greater separation of the -O- groups (11.01
Å) in the calix[8]arene dianion compared to its calix[6]-
arene analogue that accounts for the much smaller
difference between the pKa(1) and pKa(2) values in the
former.
No higher deprotonations (e.g., pKa(3)) are observed in
MeCN, even with the relatively strong Verkade’s base
2. Since the pKa of 2 is close to the accepted autoprotolysis
constant of MeCN (pKautoprotolysis > 33), it appears that a
di-deprotonated state (mono-deprotonated for calix[4]-
arene) is the maximum possible in this solvent.
In conclusion, the recently established consistent scale
of amine basicity in MeCN has been put to use to
determine all accessible pKa values for calix[4]arene,
calix[6]arene, and calix[8]arene in MeCN. For mono-
deprotonation, the dominant factor in calixarene acidity
is the anion’s ability to form internal hydrogen bonds that
are not only numerous but also strong. For di-deproto-
nation, the ability to spatially separate the anions
becomes important.
Abs ) ꢀHA[HA] + ꢀA-[A-] + ꢀ [B] + ꢀBH+[BH+] (2)
B
[A-] ) [BH+]
[HA] ) cHA - [A-] ) cHA - [BH+]
(3)
(4)
Experimental Section
[B] ) cB - [BH+]
(5)
(6)
Chemicals. The calixarenes were obtained commeri-
cally as puriss grade, and were checked for purity by
Abs - ꢀHAcHA - ꢀBcB
[BH+] )
1
microanalysis, H NMR, and mp before use.
ꢀA- + ꢀBH+ - ꢀHA - ꢀB
The following bases were obtained commercially and
used as received: 2,8,9-triisobutyl-2,5,8,9-tetraaza-1-phos-
phabicyclo[3.3.3]undecane 2, 1,3,4,6,7,8-hexahydro-1-
methyl-2H-pyrimido[1,2-a]pyrimidine 3, 1,8-diazabicy-
clo[5.4.0]undec-7-ene (DBU) 4, pyrrolidine 9, 1-(o-tolyl)-
biguanide 10, triethylamine 12, 4-aminopyridine 14,
benzylamine 15, and 2-amino-1-methylbenzimidazole 16.
The following bases were prepared by literature pro-
cedures using commercially available reagents and start-
ing materials: (2,6-dichlorophenylimino)tripyrrolidinophos-
phorane 13,21 (4-bromophenylimino)tripyrrolidinophos-
phorane 7,21 (2-chlorophenylimino)tris(dimethylamino)-
phosphorane 11,18,45 2-phenyl-1,1,3,3-tetramethylguanide
8,22 (phenylimino)tripyrrolidinophosphorane 6,21 and (4-
methoxyphenylimino)tripyrrolidinophosphorane 5.21 The
laboratory-prepared bases gave satisfactory combustion
analyses and (for solids) sharp melting points close to
the literature values. Salts (either HCl, HPF6, or HBPh4)
of the synthesized phosphorane bases were prepared
to check the stability of the protonated forms, and
gave sharp melting points and satisfactory combustion
analyses. The exceptions were 6, a low melting solid, its
6‚HBPh4 salt, and 8‚HBPh4, for which satisfactory com-
bustion analysis could not be obtained. However, the salts
did give sharp melting points, and base 6 was not used
for any precise pKa calculation.
The extinction coefficients for HA (non-deprotonated
calixarenes), B, and BH+ are known, and the important
value for the anion A- can be determined under condi-
tions of complete deprotonation (of the relevant group
only). In many experiments, the determination of con-
centration, e.g. [BH+]eq, was facilitated by the use of
suitable wavelengths; for example, in some cases, B, BH+,
and HA were not found to absorb above 300 nm, whereas
A- absorbed strongly. In other cases, determination was
at a HA/A- isosbestic point where changes in absorbance
were caused only by B/BH+. For the equilibrium between
acid HA and base B, the equilibrium constant K is given
by eq 7.
[BH+]2
(cHA - [BH+])(cB - [BH+])
K )
(7)
From this, the pKa of HA can be found (eq 8) from the
known value for BH+
pKa(HA) ) pKa(BH+) - log10 K
.
(8)
In some case (see Results), the condition [A-] ) [BH+]
did not hold, and the more general form of the equation
J. Org. Chem, Vol. 70, No. 23, 2005 9255

Cunningham and Woolfall
(eq 9) requiring independent measurement of [A-] and
[BH+] was used.
the target structure was constructed using the build
menu, and minimized by molecular mechanics (MM)
using the MMFF force field. In some cases (e.g., for
the conformationally labile calix[6]arenes), molecular
mechanics (MMFF) was used to obtain the equilibrium
conformer. For DFT calculations, the structure obtained
following MM geometry optimization (including the equi-
librium conformer) was modified to replace the tert-butyl
groups with hydrogen atoms (e.g., H-calix[6]arene), and
then geometry-optimized using the B3LYP/6-31G* model.
The quoted energies are those obtained using the same
model.
[BH+][A-]
(cHA - [A-])(cB - [BH+])
K )
(9)
The equilibrium constant K was calculated at different
levels of cB (typically 10 spaced over 1 equiv), and the
values were averaged. To check that K was being
determined under conditions where reasonable amounts
of all components (HA, A-, B, BH+) were present, the
experimental Abs/cHA vs cB/cHA plot was compared with
one constructed on the basis of “no deprotonation” (i.e.,
[HA] ) cHA, [B] ) cB, [A-] ) 0, [BH+] ) 0), and with
another corresponding to “complete deprotonation” (i.e.,
[HA] ) cA - cB, [A-] ) [BH+] ) cB, [B] ) 0); when feasible,
only points showing significant deviation from these two
extremes were used. Finally, a calculated Abs, derived
from values of the component concentrations calculated
Acknowledgment. This work was supported by
Grant GR/R24784/01 from the EPSRC.
Supporting Information Available: Details of chemicals,
purification and syntheses of bases. Graphical structures and
tables of bond lengths, angles, total energies, and Cartesian
coordinates from computational studies. This material is
available free of charge via the Internet at http://pubs.acs.org.
from the newly determined K, was plotted (as Abscalc/cHA
)
JO051312T
against cB/cHA and visually inspected for closeness of fit
to the experimental plot.
Uncertainties in weight and volume measurements are
negligible compared with the standard deviations in K
and pKa (given in the tables) across the range of cB used.
The final pKa value in each table is the simple average
of those from each base, whereas the quoted uncertainty
in the final pKa value is either the standard deviation of
this average or the simple average of the individual
deviations, whichever is larger.
(45) Tang, J.; Dopke, J.; Verkade, J. G. J. Am. Chem. Soc. 1993,
115, 5015.
(46) Spartan ’04; Wavefunction, Inc.: Irvine, CA. Except for molec-
ular mechanics and semiempirical models, the calculation methods
used in Spartan ’04 have been documented in: Kong, J.; White, C. A.;
Krylov, A. I.; Sherrill, C. D.; Adamson, R. D.; Furlani, T. R.; Lee, M.
S.; Lee, A. M.; Gwaltney, S. R.; Adams, T. R.; Ochsenfeld, C.; Gilbert,
A. T. B.; Kedziora, G. S.; Rassolov, V. A.; Maurice, D. R.; Nair, N.;
Shao, Y.; Besley, N. A.; Maslen, P. E.; Dombroski, J. P.; Daschel, H.;
Zhang, W.; Korambath, P. P.; Baker, J.; Byrd, E. F. C.; Van Voorhis,
T.; Oumi, M.; Hirata, S.; Hsu, C.-P.; Ishikawa, N.; Florian, J.; Warshel,
A.; Johnson, B. G.; Gill, P. M. W.; Head-Gordon, M.; Pople, J. A. J.
Comput. Chem. 2000, 21, 1532.
Computational Calculations. All calculations were
carried out using the Spartan ’04 program.46 Typically,
9256 J. Org. Chem., Vol. 70, No. 23, 2005