Hydrolysis of esters of oxy acids: pKa values for strong acids; Brflnsted relationship for

attack of water at methyl; free energies of hydrolysis of esters of oxy acids; and a linear
relationship between free energy of hydrolysis and pKaholding over a range of 20 pK units
J . PETERGUTHRIE'

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

Depnrt~nentof Chemistry, Ur~ir'ersityof Western Ontario, London, Ont., Canada N6A 5B7
Received October 26, 19772

J. PETERGUTHRIE.
Can. J. Chem. 56.2342 (1978).
By combining various kinds of evidence from the literature it is possible to derive a n internally consistent set of pKa values for the strong mineral acids and the arenesulfonic acids; the
values are referred to dilute aqueous solution a s standard state and a r e expected to be correct
within 0.5 log unit. Using these pKa values and literature data for hydrolysis of methyl esters of
acids of the type Y-X03Me,,, where Y is 0 , O H , OMe, alkyl, or aryl and X is C1, S, o r P, a
Brwnsted plot can be constructed with slope equal to 1.02 0.04. From the free energies of
hydrolysis for dimethyl sulfate and the methyl phosphates it is possible t o calculate rate constants for the microscopic reverse reaction. These define a Brwnsted line of slope 0.27 ? 0.3,
from which rate constants for the formation of the esters of perchloric and various sulfonic
acids may be estimated. This permits calculation of free energies of hydrolysis for these esters.
Thermochemical data in the literature permit calculation of the free energies of hydrolysis of
dimethyl sulfate, trimethyl arsenite, and tetraethyl orthosilicate. In the case of dimethyl
sulfate the calculation (using the previously reported eq. [I]) leads to a pK, value in close
agreement with theoretical expectation, confirming that eq. [I] is valid for acids of pK, 2 -3.
For tetraethyl orthosilicate the thermochemical data are less precise but are in satisfactory
agreement with the predictions of eq. [I]. The free energies of hydrolysis derived f r o m the
Bransted correlations are also in good agreement with expectation based on eq. [I].
For acids where resonance phenomena are important either in the acid itself (boric acid) or
in the anion (nitric acid, nitrous acid, carboxylic acids) the experimental free energies of
formation fall far from the line defined by eq. [I]. It is concluded that eq. [I] is limited to
species where there is no n delocalization involving the reacting oxygen, but where this condition is satisfied, the equation holds over the entire accessible range of oxy acid pK, values, i.e.,
from -6.4 to 16.

J. PETERGUTHRIE.
Can. J. Chem. 56,2342 (1978).
En combinant divers types d e donnees provenant de la littbature, il est possible d'obtenir
une sCrie autocoherente de valeurs de pK, pour les acides inorganiques forts et les acides arenesulfoniques; les valeurs se referent aux solutions aqueuses dilutes comme Ctat de reference et
on les croit correctes a k0.5 unites log. Faisant appel a ces valeurs d e pK, et aux donnCes de
la litttrature pour l'hydrolyse des esters methyliques des acides d u type Y-X03Me,,
oh
Y = 0 , O H , Me, alkyle ou aryle, et X = CI, S ou P, on peut Ctablir une droite de Bronsted
avec une pente tgale a 1.02 0.04. Utilisant les energies libres d'hydrolyse du sulfate du
mtthyle et des phosphates de methyle, il est possible de calculer les constantes de vitesses pour
la reaction microscopique inverse. Ces valeurs definissent une droite de Bransted d e pente
0.27 k 0.3, a partir de laquelle on peut evaluer des constantes de vitesse pour la formation des
esters des acides perchloriques et de divers acides sulfoniques. Ces valeurs permettent de
calculer les energies libres d'hydrolyse d e ces esters.
Les donnees thermochimiques retrouvees dans la litttrature permettent de calculer les
energies libres d'hydrolyse du sulfate de dimethyle, de l'arsenite d e trimtthyle et de I'orthosilicate de tetraethyle. Dans le cas du sulfate d e dimethyle, les calculs (utilisant I'Cq. [ I ] rapportte anterieurement) conduisent a une valeur de pK, qui est raisonnablement pres de la
valeur attendue d'une facon thkorique; ceci confirme que I'eq. [ l ] est valide p o u r les
acides de pK, 2 - 3. Dans le cas de I'orthosilicate de tetraethyle, les donntes thermochimiques
sont moins precises mais elles presentent une concordance satisfaisante avec les predictions de
1'Cq. [I]. Les Cnergies libres d'hydrolyse obtenues a partir des correlations de Bransted sont
aussi en bon accord avec les valeurs attendues en se basant sur 1'Cq. [I].
Dans le cas des acides oh le phenomene d e la resonance est important, soit dans I'acide
lui-mCme (acide borique) ou dans I'anion (acide nitrique, acide nitreux, acides carboxyliques),
les energies libres obtenues exptrimentalement pour la formation sont tres loin de la ligne
definie par I'eq. [I]. On en conclut que I'eq. [I] est limitte a des especes oh il n'y a pas d e
delocalisation pi impliquant I'oxygene qui reagit; toutefois dans le cas oh cette condition est
satisfaite, I'Cquation est valable pour la gamme complete des valeurs d e pK, des acides oxygtnes
soit de - 6.4 a 16.
[Traduit par le journal]
'Alfred P. Sloan Fellow, 1975-1979.
ZRevision received May 17, 1978.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

Introduction
Recently (1) we reported a linear correlation between the free energy change for replacing a hydroxyl
group in a molecule by an alkoxyl group and the
electronic properties of the rest of the molecules.
Initially (1) this was done for ethers using the cr*
values for the substituents on the carbon atom
bearing OH or OR, but this has since been extended
to other classes of compound by using the pK, of the
OH compound as a measure of the electron withdrawing properties of the rest of the molecule (2).
Equation [ l ] gives the relationship between the free

Ij

Ii
I'
1
1

the entire accessible range of acids which can exist in

aqueous solution and within its range of applicability
can be used with confidence to estimate unknown
equilibrium constants.

Results and Discussion

Thermodynamic properties for all compounds discussed in this paper (including quantities calculated
in this work) are found in Table 1. Literature data
were supplemented where necessary by estimated
quantities calculated using accepted procedures.
Details are found in the Appendix.

pK, Values for Strong Acids

AG" = -4.78(f 0.28) + 0.336(&0.024)pKa
[l]
For acids with pK, values lower than -2 there is
energy change, corrected for any steric or symmetry no generally satisfactory way to measure the pK,,
effects and the pK, of the hydroxyl compound for referred to aqueous solution. Table 2 summarizes
process [2]. Since this correlation worked well for various values which have been suggested. Attempts
have been made using nmr data (3-5) but these
X-OR + HZO = XOH + HOR
[2]
methods involve severe assumptions and long extraphosphate esters it seemed worthwhile to see how far polations, and generally lead to values for the pK,
it could be extended and what limitations exist for it. which seem unduly high. Although Raman (6, 7) and
Accordingly we wish to report the results of an infrared (8, 9) measurements can lead to concentraexamination of the hydrolysis equilibria for several tions of undissociated acid even for perchloric (6) and
esters of inorganic acids. The correlation appears to sulfuric acids (7-9), these measurements are only
be useful for any acids where resonance is unim- possible in very concentrated acid solutions, so that a
portant in stabilizing the anion but does not work at long and uncertain extrapolation would be needed to
all for acids for which resonance plays a role.
refer them to dilute aqueous solution. Direct meaTo extend the method to acids stronger than surements on dilute aqueous solutions of very strong
phosphoric, it was necessary to have pK, values for acids seem unlikely to lead to pK, values unless some
strong mineral acids such as sulfuric and perchloric new technical breakthrough occurs.
Attempts have been made to calculate pK, values
acid, since from these pK, values and previously
reported kinetic studies of the hydrolysis of the for the strong mineral acids using a number of
methyl esters of these acids it would be possible to approaches. Pauling's rules (ref. 10, cited in ref. 25)
make indirect estimates of the free energies of and some elaborations of them (11) are empirical
hydrolysis of the esters which could be compared correlations which work well for acids with pK, values
with the predictions of eq. [I]. A review of the data in the measurable range. Pauling's rules are: (a)
in the literature permits the calculation of a mutually pK, = 7 - 5n, where rz is taken from the structural
consistent set of pK, values for acids as strong as formula XO,,(OH),; (b) K,, K,, K,, etc., are in the
fluorosulfonic acid, referred to dilute aqueous solu- ratio 1:
10-lo. Use of the first rule suggests
tion as standard state. These pK, values permit that for H,SO,, pKl = - 3, and for HCIO,, pK, =
construction of a Brernsted plot for rates of hydrolysis - 8.
Slightly more elaborate was the attempt to calcuof methyl esters of tetracoordinate oxy acids, which
serves as an additional test of the correctness of the late pK, values for inorganic acids reported by
pK, values deduced for the strong acids. A much less Kossiakoff and Harker (12), who employed a simple
satisfactory Brernsted plot can be constructed for the electrostatic model based on the distribution of
(unobservable) reverse reaction, but the number of formal charges deduced from the octet rule. Their
points is inadequate to specify the line very well. calculations were remarkably successful and led to
Nevertheless, this line permits the calculation of the predicted pK, values of -3.0 for H 2 S 0 4 and -7.3
rate constants for the reverse reactions, and so the for HC10,. Schwartzenbach (13) employed a similar
calculation of the equilibrium constants for hy- model to calculate pK, values for these acids and
drolysis of the esters. These equilibrium constants supported his results by consideration of the series
are in agreement with the values anticipated from the HP0,'-,
pK = 12.3; HS04-, pK = 1.9; HCIO,,
linear free energy relationship between the free pK = -8.6; similarly, H2P04-, pK = 7.2; H2S04,
energy of hydrolysis and the pK, of the acid. Thus pK = - 3.1. Thus for each increase of the formal
the results in this paper demonstrate that the free charge on the central atom by 1 unit, the pK deenergy relationship which we have reported covers creases by about 10 units.

CAN. J . CHEM. VOL. 56, 1978

TABLE
1. T h e r l n o d y n a r n i c

d a t a for c o m p o u n d s discussed i n this paper"

( u ) Compounds for which free energies o f formation have not been reported

AH,O(g)

Compound

S0(g)'

AGr"(g)

AH, b .

AHr"(I)"

AGln(aq)b . '

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

( 6 ) Compounds for which free energies o f formation in aqileous solution have been reported
AGr"(aq)

Compound

Compound

AGlo(aq)

Compound

AGf"(aq)

( c ) Compounds for which the free energies of formation in aqueous solution have been calculated in this work
Conlpound

AGlo(aq)

Compound

AGro(aq)

Compound

AGr"(aq)

aAt 25'C, standard states are ideal gas at I otm, pure liquid, and 1 M aqueous solution with a n infinitely dilute reference state, unless otherwise n o t e d .
keal mol-1.
=In cal deg-' mol-I.
W e a t o f vaporization.
'Unless otherwise noted, caleulated from AGrn(g) a n d the AG, value in Table 6.

JRefcrencc
.
........- - 55
...

9Reference 56.
"Calculated from otlier values in this table.
'Calculated usins a value for AG, estimated as described in the Auuendix.
..
jReference 57.
kCalculated using the entropy o f liquid methyl nitrate (58) and the solubility o f the liquid in water (59)
'Reference
60.
~
-.~
~
mCalculaled using the atomic eontribulions method o f Benson and Buss (51).
"Calculated using the bond contributions given in ref. 52.
'Calculated from the vapour pressure data given in ref. 61, using data for pressures between 1 and 4 0 T o r r ; Bradley el a/. (62) report AH, = 11.25 k c a l
mol-' but did not specify the pressure range employed.
nRecalculated from the data o f Flitcroft and Skinner (42) using the modern value for AHro of amorphous SiOl (63).
qFree energy of formation o f H + and the monoanion.
'Reference 53.
~

wRcference 37.
xCaleulated from the free energy o f formation of the anion and the ph', (Table 2).
calculated a s descritied in the text.
'Standard state is the pure liquid.

All of these calculations are based upon considerations of formal charge; although this approach
has been very successful at least for acids of measurable pK,, including bisulfate, extension to perchlorate involves a particular kind of extrapolation
unique to this system and therefore not generally
recognized, namely the assumption that a formal
charge of plus three can be interpreted in the same
sense as one of plus two, i.e., that the electronic
effects of the central atom continue to be linear in the
formal charge even when this is three. The possibility
that nonlinearity might set in seems not to have been
considered. The electroneeativitv of the central
chlorine would be so drastically increased by the
formal charge of plus three (14) that charge redistribution by inductive withdrawal from the
neighboring oxygens could well lead to a smaller than

anticipated increment in effective charge on the

central atom upon changing from SV1to CIV". This
effect need not be dependent upon n bonding involving d orbitals, although the orbital contracting
effect of a large formal charge could well make such
TG bonding more important in C10,- than in other
oxy acids of the nonmetals.
Of these various predictions, that which seems t o
be on the firmest ground is the prediction of the pK,
of H2S04by Kossiakoff and Harker (12), although in
fact the other methods lead to very similar values.
The reason for preferring Kossiakoff and Harker's
value is that from their table of ~redictedDK values
one can see that they can predict the difference
between successive pK, values within one log unit for
those cases which are well behaved acids (without
resonance or tautomerization effects to confuse the

TABLE2. pK, values which have been reported for strong acids
Acid

PK,

Reference

Method

HC104

-9 . 9
-8.6
-7.3
-7.7
-2.7
-1.6
-5.4
-4.1
-6.2
-1.3
-4.0
-3.1
-3.0
-6.0
-9.4
-8.3

34
13
12
32
4
3
32
66
67
68
34
13
12
32
69
70

Based on pK in sulfi~ricacid
Theoretical
Theoretical
Based on pK in organic solvents
Nuclear magnetic resonance
Nuclear magnetic resonance
Based on pK in organic solvents
Solubility
Spectrophotometric; based on Hot'
Nuclear magnetic resonance
Based on pK in formic acid
Theoretical
Theoretical
Based on pK in organic solvents
Based on acidity function
Curve fitting to nmr data

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

p-TsOH

HzSO,

issue), and that they successfully predicted the

second pK, of sulfuric acid. Thus it seems reasonable
to expect that the first pK, of sulfuric acid will be
-3 f 1.
It was also desirable to have pK, values for a
number of sulfonic acids, referred to dilute aqueous
solution. Organic folklore suggests that FSO,H,
CF3S03H, and the various arenesulfonic acids have
very negative pK, values. However, consideration of
both theoretical expectations and experimental results shows that even FS03H is unlikely to have a
pK, less than -7.5.
The starting point for the theoretical argument is
the experimental value for the pK, of methanesulfonic acid in water, which has been measured as
- 1.92 f 0.01 (15). In the same investigation the pK,
value for ethanesulfonic acid was found to be
- 1.68 f 0.02.
As a model for the effect of substituents upon the
pK,, we consider the behavior of phosphonic acids;
Kresge and Tang (16a) have reported that for the
thermodynamic second pK, values of alkyl phosphonates, there is a good correlation with o*:
pK,

(8.10 f 0.10) - (1.26 f 0.07)0*

Martin and Griffin (16b) have reported a similar

correlation with o* = 1.12 for the first ionization but
they used appa;ent pK, values measured at unspecified ionic strength, and it is not clear whether the
techniques used to determine these rather low pK,
values were adequate. Although it is quite possible to
measure pK, values less than 2.5 in water by titration,
it is imperative to correct for titration of H + either
by calculation (17) or by differential titration (18).
pK2 values for some arenephosphonic acids are
available but only for benzene phosphonic acid is
there a thermodynamic value. Makitio and Konttinen (19) report that pK2 is 6.99 at an ionic strength

of 0.1 M and 7.43 at zero ionic strength. Jaffk et al.

(20) have reported apparent pK, values for a number
of arenephosphonic acids, including benzenephosphonic acid for which they give pK2 = 7.07. Their
values were corrected making the assumption that
pK, - pK,' will be the same for all arenephosphonates. o* values for substituted aryl groups can
be calculated from the pK, values of the appropriate
arylacetic acids (21) using the equation reported by
Charton (22); (o, was converted to o* using the
standard factor of 6.23 (22)). The pK, values of Jaff6
et al. (20) fall very close to the line defined by
Kresge's data when these o* values are used (see
Fig. 1). We conclude that the second pK, values of
both alkyl and arylphosphonic acids can be correlated and presumably predicted by the same equation
and that it is very probable that the thermodynamic
first pK, will show a very similar p*. Thus the preferred method for predicting the pK, values of
sulfonic acids should be to use a p* value of 1.26, and
extrapolate from the known value for methanesulfonic acid. The value of p* for sulfonic acids
should not be expected to be exactly equal to that for
phosphonic acids but consideration of the range of
p* values encountered for various classes of acids
Y-X-OH
suggests that it is extremely unlikely to
vary by more than f0.5. Although this uncertainty
could give rise to a range of f 1.5 in pK, for o* = 3,
it still places stringent limits on the reasonable values
for the pK, of, for example, FSO,H, for which
o,* = 3.24 (23) and excludes any value for pK, more
negative than -7.5.
If the apparent pK2 values for phosphate monoesters (2, 19) (including phosphoric acid with a
suitable symmetry correction) are similarly plotted
against o* values for the alkoxy or aryloxy (calculated from pK, values for the aryloxyacetic acids (24))
substituents, a good linear correlation is obtained,

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

2346

CAN. J . CHEM. VOL. 56, 1978

FIG.1. Dependence of pK, upon o* for phosphonic and sulfonic acids: (-)
least-squares line for alkyl phosphonic
acid pK2 (16a), ( 0 ) alkyl phosphonic acids, (W) aryl phosphonic acids (ref. 20; see text); (---) least-squares line for
phosphate monoesters pK2, (A) monoesters (ref. 2; see text),
(r)monofluorophosphoric acid (ref. 82); (----) predicted line
for sulfonic acids; (0)
alkylsulfonic acids, (0)
arylsulfonic
acids, ( 0 )halosulfonic acids, ( A ) sulfuric acid and monomethyl sulfate; see text for sources of pK, values.

with pK, = 9.27 - 1.360*. It should be noted that

carbon and oxygen substituents give different though
rather similar lines; thus we must not expect to find
sulfate esters or sulfuric acid itself falling on the line
for sulfonic acids. The pK, for fluorophosphate falls
between the lines for phosphate esters and phosphonic acids.
The considerations discussed above lead to predicted values for the pK, values of sulfonic acids as
follows (error limits are imposed by the assumed uncertainty of 0.5 in p* or are set at 0.5 log unit, whichever is larger): .Ph-S03H,
-2.7 f 0.5; pBrC6H4-S03H, - 3.0 f 0.5; pN02-C6H4-S03H,
-3.6 f 0.7; C1-S03H, -5.2
1.3; CF3S03H,
- 5.2 f 1.3; FS03H, -6.0 f 1.6.
Measurements in 100% sulfuric acid as solvent
have led to dissociation constants of 1.0 x
m
for HClO, (25), 8.0 x
m for CF3S03H (26),
m for
2.3 x
m for FS03H (27), and 9.0 x
ClS03H (27). These results lead to pK, values relative
to HClO, (the weakest acid in the series) of - 1.4 for
FS03H, - 1.0 for C1S03H, and -0.9 for CF3S0,H.
For a series of structurally similar acids it seems
reasonable to expect that relative ionization constants in highly ionizing solvents such as water or
sulfuric acid will be closely similar; this is equivalent
to assuming that the only significant difference in the
two solvents is the dramatic difference in solvent
basicity.
Bessikre (28,29) has determined the ionization constants for sulfuric and perchloric acids in trifluoro-

acetic acid as solvent; his values for pKi are 2.4 and
0.8, respectively. It must be borne in mind that trifluoroacetic acid is a solvent of low dielectricconstant
( E = 8.4) (28), and that the processes of ionization,
i.e., heterolytic cleavage of the H-X bond must be
considered separately from dissociation to give free
ions, the latter being quite unfavorable. Bessikre has
also reported 'global dissociation constants' for
formation of free ions from ion pairs plus unionized
acid; the corresponding pK values are 4.5 for
perchloric acid, 6.3 for sulfuric acid, and 7.1 for
methanesulfonic acid (28). Kolthoff and Bruckenstein (30) have reported the ionization constant for
perchloric acid in acetic acid as p K i = 1.0 and have
reported pK values for global ionization of 4.87 for
perchloric acid, 7.24 for sulfuric acid, and 8.46 for
toluenesulfonic acid (3 1). The global ionization constants seem to be more widely dispersed in acetic
acid; unfortunately, there seems to have been n o
determination of a value of pKi for sulfuric acid in
acetic acid. The values of pKi, since they refer to a
single process, are to be preferred as a measure of
acidity; these values lead to a pK, difference of 1.6
between perchloric and sulfuric acids.
From a study of the reaction I + HA = IHA,
where I is an indicator, Bessikre (20) has constructed
a scale of relative acidities in trifluoroacetic acid; this
scale places sulfuric acid 1.7 log units less acidic than
perchloric acid and both methanesulfonic acid and
p-toluenesulfonic acid 2.3 log units less acidic than
perchloric acid. (Chlorosulfonic acid is placed
slightly less acidic than sulfuric acid but in the
absence of control experiments t o demonstrate that
chlorosulfonic acid is the actual acid species in
mixtures of chlorosulfonic and trifluoroacetic acids,
this can not be considered definitely established. I n
sulfuric acid as solvent, chlorosulfonicacid is stronger
than sulfuric acid (27).) Since these acidities are
based on experiments which d o not distinguish
between unionized acid, HA, and the ion pair,
H f A-, the interpretation is less clear than would be
desirable. Furthermore, it is not entirely clear how
the relative acidities reported were actually calculated; if the data for indicators only are employed
(Tables I11 and IV of ref. 29) somewhat different
values are obtained, namely pK, differences relative
to perchloric acid of 1.9 for sulfuric acid and 3.1 for
methanesulfonic acid and p-toluenesulfonic acid.
The agreement between different approaches t o
these pK, differences is considerably less perfect than
might be desired but all available data can be summarized by stating that sulfuric acid is 2.0 f 0.4 p K
units less acidic than perchloric acid and that
methanesulfonic acid and toluenesulfonic acids
(which appear indistinguishable in these experiments)
are 3.1 f 0.5 pK units less acidic than perchloric acid.

GUTHRIE

TABLE3. pK, values for acids discussed in this paper"

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

Acid

P Kc,

Acid

PK~

HNO3
HNOz
H2CO3
CH3OCOzH
HCOOH
CH,COOH

~ ~ e f e r e n li ie:
*Estimated as described in the text.
<Reference 72.
'Reference 73.
gReference 74.
*Reference 75.
'Reference 33.

It is somewhat surprising that toluenesulfonic acid

and methanesulfonic acid appear so similar, since
both on theoretical (o* values) and experimental
(leaving group behavior) grounds one expects
toluenesulfonic acid to be the stronger. However,
measurements of ion pair formation could well be
perturbed by noncovalent interactions involving the
benzene rings of the indicator bases and the toluenesulfonic acid. Thus the relative acidity of toluenesulfonic acid must be considered less reliable.
We now have enough information to construct a
scale of acidity values. Starting with methanesulfonic acid, which has a well-established pK, of
- 1.92, we can calculate that sulfuric acid has a pK,
of -2.8 rt_ 0.5, and that perchloric acid has a pK, of
-5.0
0.5 from the data in acetic and trifluoroacetic acids. Then from the data in sulfuric acid we
can calculate that the pK, of chlorosulfonic acid is
- 6.0 f 0.5, that of trifluoromethanesulfonic acid is
- 5.9 0.5, and that of fluorosulfonic acid - 6.4
0.5. These pK, values, and all others used in this
paper, are found in Table 3.
Figure 1 shows that these values are in good agreement with the predicted line, which increases our
confidence in the assignment. For the arenesulfonic
acids, there is actually increased uncertainty, because
of the apparent near identity of the pK, values of
methane- and toluenesulfonic acids; for the moment
the best estimates for the pK, of the arenesulfonic

acids seem to be those based on the predicted line; to

allow for the inconsistent result in acetic and trifluoroacetic acids the error limits are increased to
f 1.0 pK, unit.
Bessiire (32) has proposed a rather different set of
pK, values based upon his data; this scale is determined by the choice of HCI as reference acid, and by
the choice of the pK;, for aqueous HCI. There appear
to be three fundamental objections to these choices :
(a) the pK, of HCl in aqueous solution is imperfectly defined, and values very different from that
chosen by Bessiere (33) have been proposed;3 (6) C1is quite different in structure and presumably in
ions used in this work
solvation from the Y-X0,to construct a scale, so that the danger of serious
3Bessiere chose a value o f - 3.7 for the pK, o f HCI in water;
this value is derived from a scale of acidity in absolute ethanol
(34). T h e scale for absolute ethanol has been based on 'global
ionization constants,' which in the case of weak acids such a s
acetic acid are probably very similar t o ionization constants
but which may be very different for strong acids which will
ionize t o give ion pairs in a solvent of low polarity such a s
ethanol; since the apparent ionization constants for HCI a n d
for CsCl are very similar in ethanol (35), it seems probable that
the global ionization constant for HC1 is really measuring
dissociation of an ion pair. This argument would lead to a
much more negative scale based on Bessiere's data if HCI is
used a s reference acid; using the more generally accepted value
of - 7 for the pK, in water of HCI (33), the pK, in water for
HC104 becomes -11. This value is rejected for reasons
discussed in the text.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

2348

C A N . J . CHEM. VOL. 56. 1978

variations in solvation effects is considerable; in

particular, the balance between ionization and
dissociation constants is likely to be quite different
and thus the meaning of the global ionization constant will not be the same for different structural
types; (c) it has not been demonstrated as far as one
can tell that HCI in trifluoroacetic acid is actually
HCI and has not to some considerable extent undergone covalent addition to the acid. HCI has a
significant tendency to add to carbonyl groups as
shown by chloroalkyl ether formation (36), when the
equilibrium is not shifted by dissociation into Hf and
Cl-, and trifluoroacetic acid is far more prone to
undergo nucleophilic addition than are less electron
deficient carboxylic acids (37).

tively, with a total of - 12.46 kcal rnol-', in excellent

agreement with the thermochemical value.

Br~nstedCorrelations for Methyl Ester. For/~~atior~

and
Soluolysis
It is possible to test the entire set of pK, values for
strong acids derived above, by taking advantage of
the large amount of work which has been done
studying the kinetics of hydrolysis of the methyl
esters of these acids. A plot of log k vs. pK, (a
Brsnsted plot) is expected to show a simple relationship provided that-the pK, values are correct but
should show no pattern or a physically unreasonable
one if they are seriously in error.
It is quite generally found that rates of nucleophilic substitutions are related t o the pK, of the
Sulfuric Acid and its Methyl Esters
nucleophile and leaving group and give linear
It is possible to test the pK, value derived for Brsnsted plots, provided that the reactions correlated
sulfuric acid in aqueous solution by comparing the all have the same rate determining step and provided
value of the free energy of hydrolysis of dimethyl that the variable reagents are of similar structural
sulfate derived from this pK,, using eq. [ l ] with the type (44). This approach has been more frequently
value which can be deduced from the heat of forma- employed in substitution at carbonyl (49, phostion of the ester. (Whether this is best regarded as a phoryl (46), or sulfonyl (47) centres than at aliphatic
test of the pK, value or of eq. [I] depends upon which carbon simply because of the scarcity of data satisis considered the better established but in any case fying the second criterion. A set of data now exists,
serious disagreement would cast doubt on one if not as a result of the present work and recent work from
both.)
this laboratory, and Fig. 2a shows that a good
Equation [ I ] predicts the free energy of hydrolysis, Brsnsted correlation exists for the rate constants for
for the process converting an ester to the unionized uncatalyzed hydrolysis of methyl esters of acids of the
acid in aqueous solution. Since sulfuric acid and its type Y-XO,H,,
but that methyl nitrate deviates
monoesters are strong acids the available thermo- badly. Unlike the other acids nitric acid derives some
dynamic data refer to the ions. A pK, value for considerable part of its acidity from resonance
sulfuric acid in aqueous solution has been derived; stabilization of the anion.
it now remains to correct for the effect of replacing
The observation that the points for the esters of
OH by OR, in order to estimate a value for the pK,, the strong acids fall within their uncertainty limits of
of monomethyl sulfate. This was done by a modifica- the line determined by methyl mesylate and trition of a method which we have previously e~nployed methyl phosphate constitutes a strong confirmation
(2). The effect was assumed to be the same as the of the assignment of pK,, values t o these acids. H a d
average (symmetry corrected) change in pK, observed these acids been as strong as has sometimes been sugfor phosphoric acid and its mono and dimethyl gested (see Table 2) then a straight line Bronsted plot
esters, using data collected in ref. 2. This effect is 0.89 could not have been obtained. For example, had
in pK,; this value was used to estimate the other pK, perchloric acid had a pK, of ca. - 10 as has been
values for inorganic esters reported in Table 3.
suggested, then the average slope of the Brsnsted
For the overall hydrolysis, the thermodynamic plot from methanesulfonate to perchlorate as leaving
data in Table 1 lead to a free energy change of group would have been -0.37 corresponding to a
- 12.82 kcal mol-I.
marked curvature. Although curved Bransted plots
are by no means unknown (44), it is without precedent to have curvature so extreme at such low values
When free energies of hydrolysis for the two stages of of the observed rate constant, when there is no prehydrolysis are calculated using eq. [I], the values equilibrium process preceding a fast rate deterobtained are -6.74 and -5.72 kcal mol-l, respec- mining step. The reactions being considered here are
believed to be simple S,2 solvolyses (79), so no such
[4] MeO-SO,-OMe
+ H 2 0=$ MeOH
explanation is possible. Thus a linear Bransted plot is
what would be expected.
MeO-SO2-OH
+ H,O =$ MeOH + HzSO,
Although it is possible to construct a Bronsted plot
[5]

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

GUTHRIE

FIG.2. (0)B r ~ n s t e dplot for hydrolysis of methyl esters of oxy acids, (@) acids without resonance effects, (0)methyl
nitrate. The line was determined by weighted least squares: log kll,o = -7.39(*0.09) - 1.02(+0.04)pKa. (b) Br~nsted
plot for the calculated rate constants for formation of methyl esters by attack of oxy anions upon protonated methyl.
The best line as well as the extreme lines allowed by the data are shown.

for the reverse reaction this is less satisfactory

because the available points do not cover a wide
enough range. Since the rate determining step for the
hydrolysis is attack of water on the methyl carbon
leading to protonated methanol as the initial product
along with the leaving group as anion, the rate constants for ester formation should be calculated for the
microscopic reverse of this reaction, i.e., attack of the
anion on protonated methanol. For all of the cases
where the rate of hydrolysis is measureable, the
equilibrium lies far on the side of hydrolysis; however, the rate constant for ester formation can be
calculated if the rate constant and the equilibrium
constant for hydrolysis are known. This condition is
satisfied for dimethyl sulfate and the methyl phosphates. Since the equilibrium constants are calculated
from thermochemical data and so are imprecise, the
rate constants for ester formation are subject to uncertainties of at least an order of magnitude. Figure26
shows this Brsnsted plot and also the extreme lines
which are allowed by the error limits on the data
points. The extreme lines are drawn so that they will
just pass within the error bars for the data; these lines
are used to assess the error bounds for extrapolations
of the line beyond the range specified by the data.
Within this range, it was assumed that the bounds on
the line were the same as the bounds on the data, and
lines parallel to the best line were drawn. Rate
constants for the formation of the esters for which
hydrolysis rate constants are available were taken
from this correlation line, witherror limits taken from

the extreme lines; these rate constants and free

energies of hydrolysis derived from them are found in
Table 4. Although these values are subject to much
greater uncertainty than is desirable, there has been
to our knowledge no other method reported to give
any information concerning these free energies of
hydrolysis.
The Brsnsted line for the attack of oxy anions on
CH,OH,+ has a rather small slope, subject to large
uncertainties. Unfortunately there do not seem to be
any other studies of the kinetics of nucleophilic
attack of a series of oxygen monoanions on a cationic
substrate, so that it is difficult to assess the reasonableness of the result. One related study is by Kevill
and Wang (480) of the nucleophilic attack of a series
of arenesulfonate anions upon methyl triflate in
acetonitrile. Using the pK,, values for arenesulfonates
derived in this work, Kevill's p value can be converted to a Bransted slope of +0.69; this is similar to
the upper limit for the slope of Fig. 26. Since
Kevill's data refer to a reaction with a neutral substrate in a dipolar aprotic solvent, rather than a
cationic substrate in protic solvent, it is not clear
how similar the slopes should be expected to be. A
somewhat more analogous system is the reaction of
anions with S-(2-chloroethyl)thiironium ion (486);
from the data in ref. 48b it may be calculated that for
sulfate and phosphate dianions, the Brsnsted slope
for nucleophilic attack is f0.23. This is also an
imperfect model since the reaction is one of dianions
with a highly strained substrate but it is encouraging

2350

CAN. J. CHEM. VOL. 56, 1978

TABLE
4. Data for Bransted plots

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

Ester

pK, of acida

log klr,ob

log krcvc

AGhydrold

AGOc

nFrom Table 3.
bPseudo-first-order rate constant for hydrolysis in water at 25'C (s-I) divided by the number of hydrolyzable methyl groups.
CSecond-order rate constant for attack o f the anion upon protonated methanol.
'Free energy change for the process: XOCH, + H 2 0 = XO- -+ H20CH3+;the standard state for water is the pure liquid, and for solutes is 1 M
aqueous solution.
=Free energy change for the process: XOCH,
H1O = XOH
HOCH,.
,Reference 76.
9Estimated from rate constants for the ethyl ester at 25C in 100 and 80% aqueous ethanol, (77) assuming that log k was linear in Y and that the
methyl to ethyl ratio was the same as for ROSOICl (76).
"eference 78.
'Reference 79.
)Reference 2.
'Values taken from the best line in Fig. 2b; error limits taken from the extreme lines.
'Calculated from k H l oand equilibrium constant, as described in the text.

that a reaction with similar charge types and the

same solvent shows a similar Brsnsted slope.
Free Energies of Hydrolysis of Some Other Esters
To supplement the Set of compounds X-0-R
which can be used to test the range of applicability of
eq. [I] some further free energies of hydrolysis for
esters of acids of known PK, can be derived from
thermochemical data in the literature.
Arsenious Acid and its Methyl Esters
Arsenious oxide (As203) is sparingly soluble in
water and it is considered that in aqueous solution
". . . the predominant species is probably As(OH),"
(38). The acid species does not have an As-H bond
by contrast to H3P03 (38). The heat of hydrolysis of
trimethyl arsenite has been measured and it is reported that the ester hydrolyzed extremely rapidly in
water (39). Thus in calculating the free energy of
formation of the ester in aqueous solution it was
necessary to estimate the free energy of transfer as
described in the Appendix. In the absence of experimental knowledge concerning the thermodynamics of
solution of organic arsenic compounds there is unavoidably some additional uncertainty in the numerical value of the free energy of formation of the ester
in solution. The thermodynamic data in Table 1 lead

to an overall free energy of hydrolysis of - 11.03

1.69 kcal mol-'. Application of eq. [l] leads t o
(MeO),As

+ 3 H z 0 e 3MeOH + H3As03

the following values for the first, second, and third

stages of hydrolysis: -2.46, -2.46, and -2.42 kcal
rnol-' or -7.34 kcal rnol-' overall. The discrepancy
is disturbingly large, amounting to 3.7 kcal overall,
or somewhat over 1 kcal per step. However, it is
probably unrealistic to expect predictions to be much
better than to within a kcal for the extent of structural variation involved in extending a relationship
originally derived for carbon compounds to a n
arsenic derivative. Furthermore the thermodynamic
data upon which the 'experimental' free energy of
hydrolysis is based are subject to more than the usual
uncertainty because of our imperfect knowledge of
the solution chemistry of arsenic. It seems not unreasonable to conclude that eq. [l] gives a useful,
albeit imprecise, estimate of the free energy of
hydrolysis even for an arsenic derivative.
Silicic Acid and its Ethyl Esters
It is now accepted that the species in solutions of
S i 0 2 or its hydrates at low pH is Si(OH), (40), with
free energy of formation equal to - 312.57 kcal
mol-' (41). The heat of hydrolysis of tetraethyl

GUTHRIE

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

orthosilicate in acidic solution has been measured

(42) but is subject to an uncertainty of 2 kcal mol-'
because of uncertainty concerning the nature (colloidal solution or amorphous precipitate) of the silica
produced (42). The first pK, for silicic acid is reported
to be 9.93 (43); this is in fair agreement with expectations from Pauling's rule and the discrepancy can
be rationalized in terms of the low electronegativity
of the central atom. If pK, values for the mono-, di-,
and triester's are estimated as already described, the
free energies of hydrolysis can be calculated for each
step. Steric effects are expected to be small, and for
lack of any better approach were estimated on the
assumption that they would be the same as for tetraethyl orthocarbonate. The predicted value for the
overall free energy of hydrolysis for process [6] is
-8.75 $_ 0.71 kcal mol-', which is to be compared
2.3 kcal
with the experimental value of -6.31
mol-'. The agreement is within the uncertainties of
the quantities concerned. We conclude that eq. [ l ] is
valid for orthosilicic acid and its esters. Free energies
of formation of the mono-; di-, and triesters were
calculated using eq. [l] and are included in Table 1.

Methyl Borate
~ r o mthe free energies of formation for trimethyl
borate and boric acid in Table 1 and the usual
assumption concerning the acid strengthening effect
of replacing hydroxyl by methoxyl, use of eq. [I]
leads to a predicted pK, for boric acid of 9.6. It
should be noted that this refers to the hypothetical
process shown in eq. [7] and not to the process
shown in eq. [8] which is believed to represent what

I 7~
181

H3B03 e H +
H3B03

+ HZB03-

+ H,O e H' + B(OH),-

actually happens in aqueous solution (33). The

calculated pK, is close to the observed value but this
is probably accidental. Just as eq. [ l ] leads to anomalously high pK, values for acids for which the anion
shows increased resonance, so it should lead to
anomalously low pK, values for acids, such as boric,
which should show a decrease in resonance upon
ionization. Using one third of the observed free
energy change for complete hydrolysis as the free
energy change per step, free energies of formation of
mono and dimethyl borate in aqueous solution were
calculated. These species are obligatory intermediates in the hydrolysis of trimethyl borate, but
nothing seems to be known about their stability.
Methyl Nitrite
Although the free energy of formation of the
gaseous ester is well defined, it is necessary to esti-

2351

mate the free energy of transfer because of the rapid

hydrolysis of alkyl nitrites. (Transesterification in
neutral alcohol is reported to be very fast (49).) From
the free energies of formation of the gaseous acid (50)
and the hypothetical aqueous acid (calculated from
the free energy of formation of nitrite ion and the
pK, reported for nitrous acid) a value of the free
energy of transfer of the acid could be obtained; it
was then assumed that the increment in AG, on going
from acid to ester would be the same as for formic
acid. This procedure, though plausible, must lead to
considerable uncertaintv in the final value of the free
energy of formation of the aqueous ester. From. the
values in Table 1, a free energy of hydrolysis of
$0.78 kcal mol-' can be obtained, which leads to a
pK, of 15, which is markedly different from the
observed value of 3.3. The difference of 12 log units
may be attributed to the effect of resonance delocalization in the anion upon the acidity of nitrous acid.
Methyl Nitrate
The data in Table 1 lead to free energy of hydrolysis
of - 0.78 kcal mol- ' and a calculated pK, of 11.9 for
nitric acid; this differs from the observed pK, value
of - 1.4 by 13.3 log units, which is a measure of the
importance of resonance delocalization upon the
acidity of nitric acid.

Methyl Esters of Carboxylic Acids

Equilibrium constants for hydrolysis of methyl
formate and acetate have been measured. These lead
to pK, values of 11.3 and 11.6, respectively. Data
from the literature permit the calculation of the
overall free energy of hydrolysis of dimethyl carbonate as -2.64 kcal mol-'. This leads to a pK, of 11.8.
Cqrrelation of Free Energy of Ester Hydrolysis with
Acid pK,
The free energies of hydrolysis derived in this work
(Table 5), as well as those reported earlier, are
plotted as a function of the pK, of the acid produced
on hydrolysis in Fig. 3. Clearly, for those acids where
resonance effects are not important, the fit to the
previously reported line is very good. Equally
clearly, the deviations for those acids which do have
important resonance effects are very substantial. The
correlation holds for the entire range of pK, values
for Brmsted oxygen acids which can exist in water
and therefore it seems appropriate to conclude that
we have a general method for calculating the equilibrium constant for a defined, and large, class of
hydrolysis reactions.
Free Energies of Transfer for Inorganic Acids
From thermodynamic data for the gaseous acids
(50) and free energies of formation o f the aqueous
acids from the present work, the following values for

2352

C A N . J . CHEM. VOL. 56. 1978

TABLE5. Data for Fig. 3

PK~

AG""

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

X-OH

FIG.3. Test of eq. [I]; the line is calculated using the slope
points for compounds
and intercept reported in ref. 2 : (0)
previously reported in ref. 2, (0)
points for inorganic acids
without resonance, ( A ) points for acids in which resonance is
important.

negative than the values for the corresponding methyl

esters and must reflect the very powerful hydrogen
bonding of strongly acidic protons to water.
Experimental
The solubility of dimethyl carbonate in water at 25C was
determined by nmr analysis of a saturated solution, integrating
the methyl signal relative to the solvent peak. The solvent w a s
0.001 N aqueous HCI.
Equilibrium constants for ester formation in water at 25C
were determined by nmr analysis as previously described (83);
values of the equilibrium constant of 0.097 & 0.015 M-' for
formic acid plus methanol (lit. (84) v a l ~ l e0.122 0.003 M-')
a n d 0.1 1 1 k 0.006 M-' for acetic acid plus methanol were
obtained.

Acknowledgements
Financial assistance from the National Research
Council of Canada and the Alfred P. Sloan Foundation is gratefully acknowledged. I thank Dana
Zendrowski for technical assistance.

"Free energy o f hydrolysis, with neutral reactants and products, in

aqueous solution, corrected for steric and symmetry effects.
*Reference 2.
=Thiswork.
dAverage AG" per step based on AGrOvalues.

the free energies of transfer from the gas phase to

aqueous solution can be calculated: HCIO,, - 21 4
kcal mol-'; H,SO,, - 19 2 kcal mol-'; HNO,,
-7.0 kcal mol-'. (The uncertainties reflect the uncertainty in AH:(g).)
These values are much more

1 . J . P. GUTHRLE.
Can. J. Chern. 53, 898 (1975).
2. J . P. GUTHRIE.
J . Am. Chern. Soc. 99,3991 (1977).
3. G. C. HOODand C. A. REILLY.J . Chern. Phys. 32, 127
(1960).
4. R. W. DUERST.J . Chem. Phys. 48, 2275 (1968).
5. G. C. HOODand C. A. REILLY.J. Chern. Phys. 27, 1126
(1957).
J . G. FREEMAN,
and T . M.
6. J. W. AKITT,A. K . COVINGTON,
LILLEY.
Trans. Faraday Soc. 65,2701 (1969).
7. ( ( I ) T . F. YOUNGand J. E. WALRAFEN.Trans. Faraday
Soc. 57,34 (1961);(b) N. G. ZARAKHANI
and M. I. VINNIK.
Russ. J. Phys. Chem. 37,260(1963).
8. V. D. MAIOROVand N. B. LIBROVICH.RUSS.J . Phys.
Chern. 47,989 (1973).
RUSS.J . Phys.
9. V. D. MAIOROVand N. B. LIBROVICH.
Chern. 47. 1295 (1973).
10. L . PAULING.
~ e n e r achemistry.
l
Freeman, San Francisco,
CA. 1947. p. 394.
J. E. RLCC1. J. A m Chern. Sot. 70, 19(1948).
12. A. KOSSIAKOFF
and D. HARKER.J . Am. Chern. Soc. 60,
2047 (1938).
13. G. SCHWARTZENBACH.
Z. Phys. Chem. 176,133 (1936).
14. (a) R. T. SANDERSON.
J. Chern. Educ. 31,2 (1945); (b) J.

15.
16.
17

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

18.
19.
20.
21.
22.
23.
24.
25.
26.
27.

28.
29.
30.

31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.

47.
48.

HINZE,M. A. WHITEHEAD,
and H. H. JAFFE.J . Am.
Chern. Soc. 85, 148 (1963).
A. K. COVINGTON
and R. THOMPSON.
J . Solution Chern. 3,
603 (1974).
( a ) A . J. KRESGEand Y. C. TANG.J . Org. Chern. 42,757
and C. E. GRIFFIN.
J . Organornet.
(1976); (b) D. J. MARTIN
Chern. 1,292 (1964).
A. ALBERTand E. P. SERJEANT.
Ionization constants of
acids and bases. Methuen, London. 1962.
T . V. PARKE
and W. W. DAVIS.Anal. Chern. 26,642 (1954).
0 MXKITIOand V. KONTTINEN.
Acta Chern. Scand. 23,
1459 (1969).
H. H. JAFFE,L. D. FREEDMAN,
and G. 0 . DOAK.J . Am.
Chern. Soc. 75,2210 (1953).
J. F. J. DIPPYand F. R. WILLIAMS.
J. Chern. Soc. 161
(1934).
M. CHARTON.
J. Org. Chern. 29, 1222 (1964).
R. W. TAFT, E. PRICE,I. R. FOX, I . C. LEWIS,K. K.
ANDERSON,
and G. T. DAVIS.J . Am. Chern. Soc. 85, 709
(1903).
N. V. HAYESand G. E. K. BRANCH.
J. Am. Chern. Soc. 65,
1555 (1943).
R. J. GILLESPIE.
J. Chern. Soc. 2537(1950).
D. G. RUSSELLand J . B. S ~ N I O R
Can.
. J. Chern. 52, 2975
(1974).
R. J. GILLESPIE,
T. E. PEEL,and E. H. ROBINSON.
J . Am.
Chern. Soc. 93,5083 (1971).
J. B E S S I ~ RBull.
E . Soc. Chirn. Fr. 3353 (1969).
J. B E S S I ~ RBull.
E . Soc. Chirn. Fr. 3356(1969).
1. M. KOLTHOFFand S. BRUCKENSTEIN.
J. Am. Chern.
SOC.78, l(1956).
S. BRUCKENS~EIN
and I . M. KOLTHOFF.
J. Am. Chern.
SOC.78, 2974 (1956).
J. BESSIERE.
Anal. Chern. Acta, 52,55 (1970).
R. P. BELL.The proton in c h e r n ~ s t ~2nd
y . ed. Cornell Unlversity Press, Ithaca, NY. 1973.
G. CHARLOT
and B. TREMILLON.
Tr~it~slated
by P. J. J.
Harvey. Chemical reactions in solvents and melts. Pergarnon Press, Oxford. 1969.
A. M. ELAGGAN,
D. C. BRADLEY,
and W. WARDLAW.
J.
Chern. Soc. 2092 (1958).
R. LEIMER
and P. SALOMAA.
Acta Chern. Scand. 1, 353
(1947).
J. P. GUTHRIE.
Can. J. Chern. 54,202(1976).
J. D. SMITH.In Comprehensive inorganic chemistry. Vol.
2. Edited by A. F. Trotrnan-Dickenson. Pergarnon Press,
Oxford. 1973. p. 608.
T . CHARNLEY,
C. T . MORTIMER,
and H. A. S K I N N E RJ..
Chern. Soc. 1181 (1953).
V. A. POPEIKIN
and A. S. MIKHAILOV.
Dokl. Akad. Nauk
SSSR, 191,917(1970);Chern. Abstr. 73,39220d (1970).
H. C. HELGESON.
Am. J. Sci. 267,729 (1969).
T . FLITCROFT
and H. A. S K I N N E RJ.. Chern. Soc. 3355
(1956).
B. N. RYZHENKO.
Geochern. Int. 4,99(1967).
W. P. JENCKS.Catalysis in chemistry and enzymology.
McGraw-Hill, New York, NY. 1969. p. 81.
W. P. JENCKS.Catalysis In chemistry and enzymology.
McGraw-Hill, New York, NY. 1969. pp. 51 1 and 533.
(a) W. P. JENCKS.
Catalysis in chemistry and enzymology.
McGraw-Hill, New York, NY. 1969. p. 83; ( b ) A .J. KIRBY
and M. YOUNAS.
J . Chern. Soc. B, 1165 (1970); (c) S. A.
K H A Nand A. J. KIRBY.
J. Chern. Soc. B, 1 I72 (1970).
0. ROGNE.J. Chern. Soc. B, 1855 (1971); 0. ROGNE.J.
Chern. Soc. B, 727 (1970).
(a) D. N. KEVILLand A. WANG.J. Chern. Soc. Chern.
Cornrnun. 618(1976); (b) C. G. S W A I Nand C. B. SCOTT.J.
Am. Chern. Soc. 75, 141 (1953).

49. S. COFFEY(Editor.). Rodd's chemistry of carbon cornpounds. 2nd ed. Vol. 1 B. Elsevier, Amsterdam. 1965. p. 57.
50. S. W. BENSON.Therrnochernical kinetics. 2nd ed. Wiley,
New York, NY. 1976.
51. S . W. BENSONand J. H. Buss. J. Chern. Phys. 29, 546
(1958).
52. H. E . O'NEALand M. A. RING.I n o g . Chern. 5,435(1966).
53. F. D. ROSSINI,D. D. NAGMAN,
W. E. EVANS,S. LEVINE,
and I . JAFFE.Natl. Bur. Stand. (U.S.)Circ. 500(1952).
54. N. C. DENOand H. E. BERKHEIMER.
J. Chern. Eng. Data,
5 , l(1960).
55. J. D. Cox and G. PILCHER.Therrnochernistry of organic
and organornetallic compounds. Academic Press, New
York, NY. 1970.
56. J. D. RAY and A. A. GERSHON.
J . Phys. Chern. 66, 1750
(1962).
57. J. D. RAY and R. A. OGG, JR. J. Phys. Chern. 63, 1522
(1959).
58. P. GRAYand P. L. SMITH.J . Chem. SOC.2380(1953).
59. G. DESSEIGNE.
Mern. poudres. 30,59(1948); Chern. Abstr.
4 4 , 7 2 2 1 (1950).
~
J. Org. Chern. 96, 2924
60. J. H I N Eand A. W. KLUEPPEL.
(1974).
61. D. R. STULL.Ind. Eng. Chern. 39,517(1947).
62. D. C. BRADLEY,
R. C. MEHROTRA,
J . D. SWANICH,
and W.
WARDLAW.
J. Chem. Soc. 2025 (1953).
and B. S. ISAND.CATCH tables for silicon
63. J. B. PEDLEY
compounds. University of Sussex, Brighton. 1972.
64. W. M. LATIMER.
The oxidation states of the elements and
their potentials in aqueous solution. 2nd ed. Prentice Hall,
Engelwood Cliffs, NJ. 1952.
J. Am. Chern. Soc. 95,6999 (1973).
65. J. P. GUTHRIE.
66. J. HORYNA.
Coll. Czech. Chern. Cornrnun. 24, 1596(1959).
67. H . CERFONTAIN
and B. W. SCHNITGER.
Recl. Trav. Chirn.
91, 199 (1972).
68. R. H. DINIUSand G. R. CHOPPIN.J. Phys. Chern. 66,268
(1962).
Can. J . Chern. 42, 1 1 13
69. S . K. HALLand E. A. ROBINSON.
(1964).
70: E . HOGFELDT.
Acta Chern. Scand. 17,785 (1963).
71. L. S. LEVITTand B. W. LEVITT.Tetrahedron, 27, 3777
(1971).
In Handbook of
72. W. P. JENCKSand J . REGENSTEIN.
biochemistry. 1st ed. H . A. Sober (Editor). Chemical Rubber Publishing Co., Cleveland, OH. 1968.
D. R. LLEWELLYN,
K. G. OLDHAM,
and C .
73. C. A. BUNTON,
A. VERNON.
J. Chern. Soc. 3574(1958).
74. C . A. BUNTON,
M. M. MHALA,K. G. OLDHAM,
and C. A.
VERNON.
J. Chern. Soc. 3293 (1960).
Critical stability con75. R. M. SMITHand A. E. MARTELL.
stants. Vol. 4. Plenum Press, New York, NY. 1976.
76. E . C. F. KO and R. E . ROBERTSON.
Can. J . Chern. 50,434
(1972).
JR., C. L. WILKINS, and E.
77. A. STREITWIESER,
KIEHLMANN.
J. Am. Chern. Soc. 90, 1598 (1968).
78. R. E. ROBERTSON,
A. ANNESSA,
and J. M. W. SCOTT.Can.
J. Chern. 5 3 , 3 106(1975).
79. R. E. ROBERTSON.
Prog. Phys. Org. Chem. 4,213(1967).
80. BEILSTEIN.
Handbuch der organischen Chernie. (a) Ser. 11,
Vol. 1 , p. 272 (1941); (b) Ser. 11, Vol. 1, p. 271 (1941); (c)
S h . 1 1 1 , Vol. 1 , p. 1198 (1958); (d) Ser. I V , Vol. 1 , p. 1259
(1973).
and L . M. DENNIS.
J. Am. Chern. Soc. 55,
81. S. H. WEBSTER
3233 (1933).
82. I. G. Ryss and V. B. TULCHENSKII.
RUSS.J. Inorg. Chern.
6 , 9 4 7 (1961).
83. J. P. GUTHRIE.
Can. J. Chern. 55,3562(1977).
84. A. R. FERSHT.J. Am. Chern. Soc. 93,3504 (1971).

CAN. J. CHEM. VOL. 56, 1978

TABLE
6. Free energies of transfer from the gas phase to aqueous solution

Compound

(MI

Vapor pressure
(Torr)

AGt
(kcal mol-I)

(CH3O)zSOz
(CH30)3As
(CH,O),B
(CH3CH20)4Si
(CH3O),CO

0.22"
1.35'
3.5'
0.0096*
1 .5gh

0.36*
9. 454
132'
1 .92"
61.5'

-3.20
-2.78
- 1.78
-0.79
-1.77

Can. J. Chem. Downloaded from www.nrcresearchpress.com by UNIV GUELPH on 10/16/14

For personal use only.

Solubility

EReference 80.
bCalculated by extrapolation of a line through vapor pressure data from Beilstein
(806 and c).
CCalculatedas described in the text, using C H ( O C H 3 ) (65)
~ as the model for evaluating
the effect of hydrogen bonding.
dCalculated from the equation relating log p and T given in ref. 39.
cCalculated from the equation relating log p to T given in ref. 81.
'Calculated as described in the text, using C(OCH2CH3)4(2) as the model for evaluating
the effect of hydrogen bonding.
RInterpolated from a plot of In p vs. 1/T based on data from ref. 61.
hThis work, see Experimental.
'Estimated as described in the Appendix.

Appendix
Estimation of Entropies
Standard entropies for the gaseous esters had to be
estimated in most cases; this was done using the
additivity schemes proposed by Benson and Buss
(51); wherever possible this was done using bond
contributions (51, 52) but in some cases no parameters were available and it was necessary to employ
the atomic contributions scheme (51). For this purpose terms for tetracoordinate chlorine, tricoordinate
boron, and tricoordinate arsenic were needed; these
were calculated from the available entropy data
(51, 53) (for perchloric acid, BCl, and BF,, and
AsC1, and AsF,) and are - 35.3, - 13.7, and - 5.2
cal deg- ' mol- ', respectively.
Estitnatio~of Solubilities
Since many of these inorganic esters hydrolyze
rapidly, it is not possible to determine their solubilities experimentally. Solubilities were estimated employing a modification of the method of Deno and
Berkheimer (54). This method is based upon a relationship between solubility and the parachor expressed as
log (cI/c2) = 0.01 3P - EA

where c, is the concentration of the liquid solute in

itself, c2 is the molar solubility in water, P is the
parachor of the solute, and EA is a correction for
hydrogen bond formation between solute and water.
For each inorganic ester of the M(OR),, type, EAwas
assumed to be equal to the value calculated from the
solubility and parachor for the analogous orthoester, C(OR), or HC(OR),,, with the same R and n.
Parachor values were calculated from density a n d
surface tension data found in Beilstein. The solubility
values so estimated are found in Table 6.
Estimation of Vapor Pressures
The vapor pressures of the esters were taken
wherever possible from published equations relating
log p to T. For the cases of dimethyl sulfate a n d
tetraethyl orthosilicate it was necessary to draw the
best lines through plots of In p vs. 1/T (in the case of
tetraethyl orthosilicate using only data at low pressure) and so estimate the vapor pressure at 25'C. I n
the case of dimethyl carbonate, only the boiling point
at atmospheric pressure was available so that it was
necessary to estimate the vapor pressure using the
heat of vaporization (60) and the heat capacity of
vaporization taken as - 12 cal deg-I mol-I (546).
The vapor pressures are found in Table 6.