The Hydrogen-Bond Basicity pKHB Scale of Peroxides and Ethers

Article abstract of DOI:10.1002/(SICI)1099-0690(199805)1998:5<925::AID-EJOC925>3.0.CO;2-F

Using 4-fluorophenol as a reference hydrogen-bond donor, equilibrium constants, K_f, for the formation of 1:1 hydrogen-bonded complexes have been obtained by FTIR spectrometry for 39 ethers of widely different structure (cyclic and acyclic ethers, crown ethers, glymes, acetals, orthoesters, and disiloxane) and 3 peroxides, in CCl₄ at 298 K. The pk_HB scale of monoethers extends from 1.44 for 2,3-diadamant-2-yloxirane to -0.53 for hexamethyldisiloxane. The main effects explaining the variation of the hydrogen-bond basicity of sp³ oxygen atoms are (i) the electron-withdrawing field-inductive effect [e.g. in (CF₃)₂CHOMe], (ii) the electron-withdrawing resonance effect (e.g. in EtOCH=CH₂) (iii) the steric effect (e.g. in tBu₂O), (iv) the lone-pair-lone-pair repulsion (e.g. in cyclic peroxides), and (v) the cyclization giving the basicity order: oxetane > tetrahydrofuran > tetrahydropyran > oxirane. A spectroscopic scale of hydrogen-bond basicity is constructed from, the infrared frequency shift Δv(OH) of methanol hydrogen-bonded to peroxides and ethers. The thermodynamic pK_HB scale does not correlate with the Δv(OH) scale because of (i) statistical effects in polyethers and peroxides (ii) secondary hydrogen-bond acceptor sites (e.g. in benzyl ether), (iii) variations of the s character of oxygen lone pairs either by conjugation or cyclization, (iv) steric effects, (v) lone-pair-lone-pair repulsions, and (vi) anomeric effects. The v(OH...O) band shape reveals two stereoisomeric complexes, the most stable being tetrahedral at the ether oxygen atom.

Full text of DOI:10.1002/(SICI)1099-0690(199805)1998:5<925::AID-EJOC925>3.0.CO;2-F

FULL PAPER
The Hydrogen-Bond Basicity pKHB Scale of Peroxides and Ethers
Michel Berthelot*, Francois Besseau, and Christian Laurence
¸
Laboratoire de Spectrochimie, Faculte´ des Sciences et des Techniques, Universite´ de Nantes,
BP 92208, F-44322 Nantes Cedex 3, France
berthelo@chimie.univ-nantes.fr
Received November 21, 1997
Keywords: Basicity scale / Hydrogen bonding / Ethers / Peroxides / FTIR
Using 4-fluorophenol as a reference hydrogen-bond donor, peroxides), and (v) the cyclization giving the basicity order:
equilibrium constants, Kf, for the formation of 1:1 hydrogen-
bonded complexes have been obtained by FTIR spectrometry
for 39 ethers of widely different structure (cyclic and acyclic
ethers, crown ethers, glymes, acetals, orthoesters, and disil-
oxetane Ͼ tetrahydrofuran Ͼ tetrahydropyran Ͼ oxirane. A
spectroscopic scale of hydrogen-bond basicity is constructed
from the infrared frequency shift ∆ν(OH) of methanol hydro-
gen-bonded to peroxides and ethers. The thermodynamic
oxane) and 3 peroxides, in CCl4 at 298 K. The pKHB scale of pKHB scale does not correlate with the ∆ν(OH) scale because
monoethers extends from 1.44 for 2,3-diadamant-2-yloxirane of (i) statistical effects in polyethers and peroxides (ii) secon-
to –0.53 for hexamethyldisiloxane. The main effects explain- dary hydrogen-bond acceptor sites (e.g. in benzyl ether), (iii)
ing the variation of the hydrogen-bond basicity of sp3 oxygen
atoms are (i) the electron-withdrawing field-inductive effect
[e.g. in (CF3)2CHOMe], (ii) the electron-withdrawing res-
variations of the s character of oxygen lone pairs either by
conjugation or cyclization, (iv) steric effects, (v) lone-pair–
lone-pair repulsions, and (vi) anomeric effects. The ν(OH···O)
onance effect (e.g. in EtOCH=CH2) (iii) the steric effect (e.g. band shape reveals two stereoisomeric complexes, the most
in tBu2O), (iv) the lone-pair–lone-pair repulsion (e.g. in cyclic stable being tetrahedral at the ether oxygen atom.
Since the work of Taft et al.[1] and Arnett et al.[2], 4- been studied. We present here the pKHB scale for peroxides
fluorophenol has proved to be an excellent reference hydro- and (mainly) ethers.
gen-bond donor for the establishment of a thermodynamic
To our knowledge, no formation constant for the hydro-
hydrogen-bond basicity scale for organic bases B. This
gen-bonded complex between any hydrogen-bond donor
scale, denoted by pKHB, is defined as the logarithm of the
and any peroxide seems available. In the case of ethers, nu-
merous measurements of hydrogen-bond formation con-
formation constant Kf of the 1:1 hydrogen-bonded complex
4-FC6H4OH···B in CCl4 at 298 K [eqs. 1Ϫ3].
stants have been reported[17][18][19][20][21][22][23][24]. However
they generally refer to a too limited number of compounds
for giving a wide view on the influence of structure on ether
hydrogen-bond basicity. Moreover, they were carried out
with different hydrogen-bond donors (methanol[19], phen-
ols[17][18][20][22][23][24], isothiocyanic acid[21]), different sol-
vents (cyclohexane[24] and carbon tetrachloride) and differ-
ent temperatures. Statistical procedures should be used to
set up an homogeneous basicity scale from these data and
this would inevitably result in a loss of fine structural infor-
mations. We prefer building a scale defined from a reference
process (eq. 1) rather than a statistical scale.
4-FC6H4OH ϩ B v 4-FC6H4OH···B
Kf / l molϪ1 ϭ [complex] / [4-FC6H4OH]·[Base]
pKHB ϭ log10Kf
(1)
(2)
(3)
The choice of these standard conditions allows the accu-
rate determination of Kf in a wide basicity range by various
techniques (19F NMR[1], UV[2], calorimetry[3]) giving al-
most identical results[4]. Linear free-energy relationships
indicate[5] that the lowest limit of the pKHB scale measur-
able in CCl4 at 298 K is ca. Ϫ1.1. For neutral bases we have
indications[6] that the pKHB scale can be extended to ca. 6.
Thus, this basicity scale extends, at least, on a 7 pK units
range and a 40 kJ molϪ1 Gibbs energy range.
We have used FTIR spectrometry in this work since this
technique brings many additional informations on hydro-
gen-bonded complexes.
The building of the pKHB scale contributes not only to
the increasing efforts towards a quantitative description of
The equilibrium concentrations of the different species in
the hydrogen bond, but also to the difficult and unachieved eq. (1) can be obtained from the absorbances of the OϪH
task of measuring quantitatively the strength of organic stretching of 4-fluorophenol at 3614 cmϪ1 for different in-
Lewis bases. We have already measured pKHB for nitro- itial base concentrations. Therefore, we have first deter-
gen[7], oxygen, sulfur[8], and π[9] bases. In the case of oxygen mined the pKHB values of 3 peroxides and 39 ethers includ-
bases, alcohols[10], esters[11], amides[12], ketones[13], nitro ing crown ethers, glymes, acetals, orthoesters, and one dis-
bases[14], sulfonyl bases[15], and amidates[16] have already iloxane. The case of aromatic ethers, which are not only
Eur. J. Org. Chem. 1998, 925Ϫ931
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0505Ϫ0925 $ 17.50ϩ.50/0
925

M. Berthelot, F. Besseau, C. Laurence
FULL PAPER
oxygen bases but also π bases, will be studied in a future they have almost the same basicity, it is easy to refer to
paper. We have found that the ether pKHB scale extends on the basicity of one oxygen atom by applying the statistical
ca. 2.5 pK units and the main effects governing these varia- correction Ϫlog n to pKHB. The results are shown in pa-
tions is discussed.
rentheses in Table 1.
Secondly, we have measured the methanol and 4-fluoro-
According to eq. (4), we have to substract Kf(X) in order
phenol OϪH stretching wavenumber shifts between the free to obtain the true ether basicity Kf(O). We have evaluated
and the hydrogen-bonded OH group. These shifts are often Kf(X) by using relationships between pKHB(X) and
considered as spectroscopic scales of hydrogen-bond ba- ∆ν(OH···X) established in the families of π bases[9] and of
sicity and their comparison with the thermodynamic pKHB chloroalkanes[31]. The results are presented in Table 2, and
scale reveals steric and anomeric effects and variations in pKHB(O) are reported in parentheses in Table 1 for dibenzyl
the hybridization state of the oxygen lone pairs.
ether (10), 2-chloroethyl ethyl ether (13) and bis(2-chloro-
At last, IR spectrometry has revealed, from the ethyl) ether (16). The pKHB values remain uncorrected for
ν(OH···O) band shape and through deconvolution pro- the compounds 9, 15, 17, 20, 32, 35, 37, and 38 because it
cedures, the existence of two stereoisomers for many hydro- does not appear any detectable ν(OH···π), ν(OH···Cl), or
gen-bonded ethers. This stereoisomerism of the hydrogen- ν(OH···F) bands in the IR spectra of their complexes with
bond formation, already known for ketones[25][26][27], does 4-fluorophenol.
not seem to have been previously described for ethers.
Literature measurements[17][18][19][20][21][22][23][24] of the
formation constants K for 1:1 hydrogen-bonded complexes
between ethers and various hydrogen-bond donors have
been compared to our pKHB scale via the linear free-energy
relationship (6)
Results
The pKHB values are compiled in Table 1. For a few X-
substituted ethers bearing a second hydrogen-bond ac-
ceptor site X, the presence in the IR spectra of a second
ν(OH···X) band, in addition to the ν(OH···O) band (Figure
1), shows that two 1:1 hydrogen-bonded complexes are
formed in solution, the OH···O and the OH···X complexes.
pKHB ϭ α logK ϩ β
(6)
and the results are presented in Table 3. Our pKHB scale
agrees (mean correlation coefficient rm ϭ 0.983) with the
literature, and the agreement is excellent (rm ϭ 0.995) when
very few important outliers are excluded. The fact that
these excluded data disagree only in one data set (i.e. agree
with pKHB in all other sets) indicates a greater reliability of
our results.
Figure 1. IR spectrum of the association of 4-fluorophenol (0.004
) with bis(2-chloroethyl)ether (0.6 ) in CCl4; two 1:1 complexes
are observed and attributed to adducts to the chlorine atom (Cl)
and to the oxygen atom (O)
The wavenumber shifts ∆ν1(OH) of methanol and
∆ν2(OH) of 4-fluorophenol are also reported in Table 1.
For methanol the ν1(OH···O) band is roughly symmetrical
if we exclude the peroxides and epoxides. For 4-fluoro-
phenol the ν2(OH···O) band appears in many cases as flat
and abnormally broad or presents a shoulder. These shapes
signify the presence of two overlapping bands. The existence
of two bands is particulary evident in epoxides and per-
oxides, and they are even resolved in diadamantyloxirane
21. We have mathematically resolved these broad bands into
two Gauss-Lorentz-type components, as illustrated in Fig-
ure 2, and the wavenumber shifts corresponding to the com-
ponent maxima are reported in Table 1.
Figure 2. IR spectrum of the association of 4-fluorophenol (0.004
) with 2,3-diadamant-2-yloxirane (0.02 ) in CCl4; the two 1:1
complexes are attributed to structures A and B shown in Figure 3
It is easy to demonstrate that the measured formation
constant Kf (with ether in excess in order to avoid the for-
mation of 2:1 complexes) is a global constant correspond-
ing to the sum of the formation constants of two 1:1 com-
plexes:
Kf ϭ Kf(X) ϩ Kf(O)
(4)
In the same way, eq. (5) applies to polyethers and per-
oxides. In this case,
n
Kf ϭ Σ Kf(Oi)
(5)
i ϭ 1
if the oxygen atoms are equivalent or if we assume that
926
Eur. J. Org. Chem. 1998, 925Ϫ931

The Hydrogen-Bond Basicity pKHB Scale of Peroxides and Ethers
FULL PAPER
Table 1. Hydrogen-bond basicity of ethers and peroxydes: IR shifts of methanol [∆ν1(OH)/cmϪ1] and 4-fluorophenol [∆ν2(OH)/cmϪ1
and pKHB values
]
[a]
[b]
[c]
Statistically corrected. Ϫ Value corrected for the presence of a second accepting group. Ϫ See formulae.
Discussion
stretching bands of the 4-fluorophenol[D] complexes keep
the same shape that of the OH ones: this makes improbable
the attribution of the doublet to a Fermi resonance. Sec-
ondly, temperature effects (from Ϫ6 to ϩ55°C in CCl4)
show that the two band intensities decrease reversibly at the
benefit of the free band at 3614 cmϪ1 as the temperature of
The Two 1:1 Complexes of 4-Fluorophenol with Ethers
We have attributed the two OH stretching bands of the
4-FC6H4OHϪether complexes to complexes of different
geometries on the following grounds. Firstly the OD
Eur. J. Org. Chem. 1998, 925Ϫ931
927

M. Berthelot, F. Besseau, C. Laurence
FULL PAPER
Table 2. Evaluation of the oxygen hydrogen-bond basicity Kf(O) lowing sequences in the percentage of B from infrared
when a second basic center X is present in the molecule
intensities:
No. Compound
Kf[a][b] ∆ν(OH···X)[c] Kf(X)[a] Kf(O)[a][d]
2,2,5,5-tetramethyltetrahydrofuran > tetrahydrofuran
tBu2O > iPr2O > Et2O
tBu2O > tBuOEt > tBuOMe
10
13
16
dibenzyl ether
2-chloroethyl ethyl ether 3.55
bis(2-chloroethyl) ether 1.85
5.25
55.6
87
0.39[e]
0.78[f]
0.46[f]
4.48
2.78
0.94
67
Moreover, the crystallographic scatterplots illustrating
[a]
[c]
[d]
l molϪ1. Ϫ [b] Kf ϭ 10 pKHB. Ϫ cmϪ1. Ϫ Kf(O) ϭ Kf Ϫ n the position of hydrogen-donor atoms involved in hydrogen
Kf(X). Ϫ
Calculated[9] from pKHB(π) ϭ 0.0109 ∆ν(OH···π) Ϫ
[e]
bonding to sp3 oxygen atoms show[38] a cluster correspond-
ing to geometry B more marked for epoxides than for other
1.02. Ϫ Calculated[31] from pKHB(Cl) ϭ 0.0115 ∆ν(OH···Cl) Ϫ
[f]
1.11.
Table 3. Results of the correlations pKHB ϭ α log K ϩ β between the equilibrium constants for the association of ethers with 4-
fluorophenol (pKHB) and other hydrogen-bond donors (log K)
Set
no.
Hydrogen-bond donor,
temperature, solvent
α
β
N[a]
r[b]
s[c]
Ref.
[18]
1
2
3
4
5
Phenol, 293 K, CCl4
0.967
0.452
1.307
0.981
1.225
Ϫ0.062
0.184
11
5
13
11
10
0.994[d]
0.999[e]
0.990
0.055
0.006
0.087
0.056
0.024
[19]
Methanol, 298 K, CCl4
HNCS, 298 K, CCl4
[21]
0.183
[20][22]
[24]
Phenol, 298 K, CCl4
4-Chlorophenol, 293 K, C6H12
Ϫ0.087
0.160
0.996[f]
0.996[g]
[a]
[b]
[c]
Number of common ethers. Ϫ
Correlation coefficient. Ϫ Standard deviation. Ϫ [d] After excluding furan, Ph2O, and nBu2O. Ϫ
[e]
[g]
After excluding dioxane. Ϫ [f] After excluding tetrahydropyran. Ϫ After excluding tBuOMe.
the medium is increased. It is important to note that the cyclic ethers. We also observe, from the relative intensities
most shifted band is more sensitive to temperature varia- of OH bands attributed to complexes A and B, that the
tions, showing that the complex with the greatest ∆ν(OH) percentage of B increases from cyclic ethers to epoxides.
has also the greatest formation enthalpy, in agreement with
the Badger-Bauer rule[32][33]
.
Structure؊Hydrogen-Bond Basicity Relationships
We assume geometries A and B shown in Figure 3 for
the most stable and the least stable complexes, respectively.
The geometry A, tetrahedral at the ether oxygen atom, i.e.
with an OH group in the direction of a conventional sp3
oxygen lone pair, has been found in the solid state for the
pentafluorophenolϪdioxane complex[34], and in the gas
phase for the hydrogen-bonded dimers of oxetane, oxirane,
The pKHB scale of ethers extends from 1.44 for 2,3-diada-
mant-2-yloxirane, the most basic monoether, to Ϫ0.53 for
hexamethyldisiloxane, the least basic one. Other weakly
basic ethers are (CF3)2CHOMe (Ϫ0.41) and furan[*]
(Ϫ0.40). These examples illustrate the importance of steric
effects (Me3SiOSiMe3), resonance effects (furan) and field-
inductive effects [(CF3)2CHOMe] on the hydrogen-bond
basicity of sp3 oxygen atoms.
or 2,5-dihydrofuran with various molecules HX[35][36][37]
.
However, a crystallographic analysis[38] of the directionality
of hydrogen bonding to sp3 oxygen atoms show that the
preference for a hydrogen-bond donor group to align with
the axes conventionally associated with the oxygen lone
pairs is quite weak because it is readily masked by lattice
forces and steric interferences between the R or RЈ substitu-
ent and the OH group. Since most of the hydrogen bonds
crystallographically surveyed do tend to lie within the plane
of the two oxygen lone pairs, a trigonal geometry B is sug-
gested for the second isomer.
The field-inductive effect can be rationalized by means of
the field-inductive substituent constant σF[41]. For the series
of acyclic non-conjugated ethers RORЈ (1Ϫ8, 10, 11, 13, 16,
17, and 20), the pKHB values, corrected when necessary for
the association on a second site, are correlated to the sum
of the σF constants of substituents R and RЈ (eq. 7).
pKHB ϭ 1.033 Ϫ 3.367 ΣσF
N ϭ 14; r ϭ 0.942; s ϭ 0.18
(7)
This correlation is rather satisfactory if we consider that
steric, polarizability, and conformational effects are not
taken into account by the σF parameter. The reaction con-
stant ρF is significantly larger for ethers (3.37) than for es-
ters RCOOEt (1.79)[11] or amides RCONMe2 (2.74)[12], as
expected for a direct substitution on the hydrogen-bond ac-
ceptor atom itself.
Figure 3. Two possible structures for the complex etherϪhydrogen-
bond donor
[*] Rotational spectroscopy shows[39] that HCl is hydrogen-bonded
to the furan oxygen atom along the C2 symmetry axis of the
molecule. We expect that furan is also an oxygen base towards
4-fluorophenol (and not a π base as assumed in ref.[40]).
We expect that complex A is more destabilized than B by
bulky R and/or RЈ groups and we indeed observe the fol-
928
Eur. J. Org. Chem. 1998, 925Ϫ931

The Hydrogen-Bond Basicity pKHB Scale of Peroxides and Ethers
If we assume that the lower basicity of peroxides ROORЈ
FULL PAPER
This non monotonic variation of pKHB with the ring size
compared to ethers is mainly the consequence of the elec- might be the result of two opposite effects. When the
tron-withdrawing field-inductive effect of each oxygen atom CϪOϪC angle diminishes with the ring size, the steric hin-
on the other, and since the sum of the σF substituent con- drance of the oxygen lone pairs by the α-methylene groups
stants of the alkyl R and of the alkoxy ORЈ is equal to must decrease[24], and consequently pKHB increases. But
0.25[41], we calculate from eq. (7): pKHB ϭ 0.16 for one oppositely the closure of the CϪOϪC angle must increase
oxygen atom and (adding log 2 ϭ 0.30) 0.46 for the whole the s character of the oxygen lone pairs[17][20] reducing
peroxide. This predicted value agrees very well with the ex- pKHB. The rehybridization of the oxygen lone pairs might
perimental value (0.43) of tBuOOtBu, but this agreement be significant mainly in epoxides[20] and this would explain
may be fortuitous since the prediction does not take into that the three-membered ring structure causes a definite de-
account the steric effects of the tBu groups. On the contrary crease in basicity.
the predicted pKHB of cyclic peroxides is by far too low
when compared to the experimental values. We suspect that
this difference originates in lone-pairϪlone-pair repulsions
The pKHB-∆ν(OH) Relationship
between the two adjacent oxygen atoms. These repulsions
are not taken into account by the σF constant and are more
effective in cyclic than in acyclic peroxides where they can
be released by rotation around the OϪO bond. Lone-pair
repulsions increase the basicity as described by Taft et al.
The correlation between the Gibbs energies of hydrogen-
bond formation (i.e. pKHB) and infrared shifts has often
been studied. It is well established[18] that a general corre-
lation does not exist for all kinds of bases, but that family-
dependent relationships do hold[7][8][9][10][11][12][14][15] if a
number of conditions are obeyed. In this context, a family
is defined as a group of bases that possess the same hydro-
gen-bond acceptor atomic center in roughly the same hy-
bridization state. It has been observed that the various fam-
ilies of bases trace roughly parallel lines in the
pKHB,∆ν(OH) plane, disposed from left to right with the
increasing p character of the hydrogen-bond accepting lone
pair(s)[7b]. In other words, for the same pKHB value ∆ν(OH)
increases from nitriles (sp N) to pyridines (sp2 N) and to
amines (sp3 N).
for adjacent nitrogen atoms[42]
.
The importance of steric effects on the hydrogen-bond
basicity of ethers is well established[24] and explains that the
most crowded ether tBu2O is the least basic of alkyl ethers.
However, West et al.[20] have shown that chain branching α
to the oxygen atom has two opposite effects. It increases the
hydrogen-bond formation enthalpy because of the elec-
tronic effects of alkyl groups and this should increase ba-
sicity in the order: Me < Et < iPr < tBu. It also increases
the hydrogen-bond formation entropy because of the steric
effect and this should reduce basicity in the order: Me > Et
> iPr > tBu. The pKHB sequence: tBuOtBu < tBuOEt <
tBuOMe corresponds to the predominance of the steric ef-
fect, but in the following order: tBu2O < Et2O < iPr2O the
second inequality corresponds to the predominance of elec-
tronic effects over the steric effect. The severe steric require-
ment of two SiMe3 groups also explains, but not totally
(vide infra), the very low basicity of (Me3Si)2O.
The electron-withdrawing resonance effect of the vinyl
and ethynyl groups adds to their electron-withdrawing field-
inductive effect for explaining the low basicity of EtOCHϭ
CH2 and EtOCϵCH. A transfer of pπ electrons from the
oxygen atom to the dπ orbitals of the silicon atoms[43] also
reduces the oxygen basicity of (Me3Si)2O.
As seen in Figure 4, the hydrogen-bond basicity of cyclic
ethers varies with the ring size in the following order: 4- >
5- > 6- > 3-membered rings.
The conditions for a good family-dependent ∆G-∆ν re-
lationship to be observed are the following. Firstly, the
steric requirements of the hydrogen-bond acceptor site must
remain almost constant. Secondly, the hydrogen-bond ac-
ceptor site must remain unique; if not, the pKHB must be
corrected (by the statistical factor Ϫlog n if there are n
equivalent sites). In any case, and at last, these sites must
not be adjacent, because lone-pair(s)Ϫlone-pair(s) repul-
sions increase more importantly pKHB than ∆ν(OH)[44]
Ethers and peroxides constitute interesting groups of bases
for checking these rules and for showing how the pKHB
.
-
∆ν(OH) plot reveals or confirms structural features govern-
ing the hydrogen-bond basicity. Since methanol gives more
symmetrical absorptions (vide supra) leading to more accu-
rate ∆ν values, ∆ν1(OH) is preferred to ∆ν2(OH) for this
plot.
In the pKHB vs. ∆ν1(OH) plot (Figure 5), we observe a
scattering of points, and calculate a very low correlation
coefficient (r ϭ 0.692, n ϭ 40). In applying to pKHB the
polysites corrections (eqs. 4 et 5), we improve significantly
the correlation coefficient (r ϭ 0.80). The remaining scatter
is partly caused by steric effects to which the ∆G-∆ν corre-
lation is very sensitive[*]. In order that they do not obscure
other structural effects, we have focused on cyclic ethers and
peroxides, which are less dependent on steric effects than
Figure 4. Variation of pKHB with the ring size of cyclic ethers
[*] For the reason that ∆G is more sensitive to steric effects,
through its entropic term, than ∆ν, roughly related to ∆H, ac-
cording to the Badger-Bauer relationship[32][33]
.
Eur. J. Org. Chem. 1998, 925Ϫ931
929

M. Berthelot, F. Besseau, C. Laurence
FULL PAPER
open-chain ethers since cyclization pulls back the alkyl
chains.
Finally, we note in Table 4 that tBuOϪOtBu behaves
similarly as hindered ethers. This confirms that the confor-
mational flexibility of acyclic peroxides allows a significant
release of the lone-pairs interactions between the adjacent
oxygen atoms, unlike the cyclic peroxides.
Figure 5. Family-dependent relationship between pKHB and
∆ν1(OH) for cyclic ethers; numbers refer to Table 1
Experimental Section
Materials: Most compounds were purchased from Aldrich. Their
purity was checked by gas chromatography and, when necessary,
they were distilled, and dried over molecular sieves. 4-fluorophenol
was sublimed over P2O5 at 60°C and 13 Pa. Spectroscopic-grade
˚
CCl4 was passed through a column of freshly activated 4-A molecu-
lar sieves before use. Spectroscopic-grade methanol was kept over
˚
3-A molecular sieves. Ϫ Ascaridole (38) and MeOCH(CF3)2 (20)
have been generously given to us by D. G. Morris (Glasgow) and
M. H. Abraham (London), respectively. Di-tert-butyl ether (8) was
synthesized in the laboratory of Prof. C. Reichardt (Marburg) ac-
cording to the method of Erickson and Ashton[28]. Diada-
mantyloxirane 21 and diadamantyldioxetane 39 have been synthe-
tized by J.-Y. Le Questel in the department of chemistry of the
University of Glasgow under the supervision of D. G. Morris ac-
cording to the procedure shown in Figure 6. Details of the prep-
arations and physical properties of 21 and 39 have been reported
in the Le Questel thesis[29]
.
Figure 6. Synthetic pathway for compounds 21 and 39
Figure 5 clearly shows:
1) Separate lines AϪD for six-, five-, four-, and three-
membered rings ranging from right to left when the oxygen
lone pairs gain s character with the closure of the
CϪOϪC angle.
2) A line E for 1,3,5-trioxane, 1,3-dioxane, and 1,3-diox-
olane, cyclic ethers where the 1,3 positions of the oxygen
atoms allow the anomeric effect[45] to operate.
3) A line F for conjugated cyclic ethers. Such conjugation
of the oxygen lone pairs with a π system represents a partial
rehybridization giving larger s character to the lone pairs.
Consequently, the line F is situated on the left side of lines
A and B.
Infrared Spectra: Infrared measurements were carried out with
Bruker IFS 48 or Nicolet 510 M FTIR instruments, selecting a 1-
cmϪ1 resolution and an accumulation of 256 scans. The 1-cm (for
the absorbance measurements) and 4-cm (for the shifts measure-
ments) cells were thermostatted at 25 ± 0.2°C. Variable-tempera-
ture experiments were performed from Ϫ6 to ϩ55°C with a Peltier
system. All operations, including the filling of the cells, were con-
ducted in the dry atmosphere of a glove box. The spectra were
treated with softwares including deconvolution and mathematical
decomposition of overlapping or dissymetrical bands into Gauss-
Lorentz-type components.
4) The position of cyclic peroxides at the upper limit of
the plot (line G), confirming that lone-pairsϪlone-pairs re-
pulsion of the adjacent oxygen atoms increases pKHB
.
5) A rough parallelism of lines AϪF.
Now steric effects appear clearly as deviations under the
appropriate lines. If we refer open-chain ethers to the line
A of unstrained rings, the deviations collected in Table 4
represent a rough estimation of the steric effect in some
crowded ethers.
Equilibrium Constants: The formation constant of 1:1 complexes
(c) of 4-fluorophenol (a) with bases (b) is defined as Kf ϭ Cc/Ca·Cb
where the equilibrium concentrations are on a molar scale. The
initial molar concentration of 4-fluorophenol, Ca0, was kept under
the limit of 3·10Ϫ3 mol lϪ1 in order to prevent self-association,
while the initial molar concentration of the base, Cb0, was adjusted
so that the complexation of 4-fluorophenol was not less than 20%
and not more than 80%. For example, for 2-methyltetrahydrofuran
(Kf ϭ 21.8 l molϪ1), Cb0 was varied from 1.4·10Ϫ2 mol lϪ1 (23% of
the phenol is complexed) to 5.6·10Ϫ2 mol lϪ1 (53% of the phenol
is complexed). The equilibrium concentration Ca was obtained
from the absorbance of the 3614 cmϪ1 band of 4-fluorophenol (ε ϭ
Table 4. Vertical deviations (∆) of some acyclic ethers and a per-
oxide from the line A of Figure 5 corresponding to the unstrained
rings
Compound
∆
Compound
∆
EtϪOϪEt
Ϫ0.11
Ϫ0.13
Ϫ0.15
Ϫ0.19
Ϫ0.23
tBuϪOϪEt
Ϫ0.28
Ϫ0.30
Ϫ0.50
Ϫ0.66
Ϫ0.20
PhCH2ϪOϪCH2Ph
ClCH2CH2ϪOϪEt
tBuϪOϪMe
nBuϪOϪnBu
Me3SiϪOϪSiMe3
tBuϪOϪtBu
iPrϪOϪiPr
tBuϪOϪOϪtBu
930
Eur. J. Org. Chem. 1998, 925Ϫ931

The Hydrogen-Bond Basicity pKHB Scale of Peroxides and Ethers
FULL PAPER
[11]
[12]
[13]
237 l molϪ1 cmϪ1) and the other equilibrium concentrations were
deduced from Ca. The maximum error in pKHB is estimated to be
± 0.04.
F. Besseau, C. Laurence, M. Berthelot, J. Chem. Soc., Perkin
Trans. 2 1994, 485Ϫ489.
J.-Y. Le Questel, C. Laurence, A. Lachkar , M. Helbert, M.
Berthelot, J. Chem. Soc., Perkin Trans. 2 1992, 2091Ϫ2094.
F. Besseau, M. Lucon, C. Laurence, M. Berthelot, J. Chem.
¸
In the case of polyethers, we must maintain the ether concen-
tration at least in tenfold excess, in order to favor the formation of
1:1 complexes over n:1 complexes. The constancy of the equilib-
rium constants determined at different ether concentrations on the
basis of 1:1 complex formation, indicates that the 1:1 complex is
the primary species formed (Table 5).
Soc., Perkin Trans. 2 1998, 101Ϫ107.
[14] [14a]
C. Laurence, M. Berthelot, M. Lucon, D. G. Morris, J.
¸
Chem. Soc., Perkin Trans. 2 1994, 491Ϫ 493. Ϫ [14b] A. Chardin,
C. Laurence, M. Berthelot, D. G. Morris, Bull. Soc. Chim. Fr.
1996, 133, 389Ϫ393.
[15]
[16]
[17]
A. Chardin, C. Laurence, M. Berthelot, D. G. Morris, J. Chem.
Soc., Perkin Trans. 2 1996, 1047Ϫ1051.
A. Chardin, M. Berthelot, C. Laurence, D. G. Morris, J. Phys.
Org. Chem. 1994, 7, 705Ϫ711.
0
Table 5. Variation of Kf (l molϪ1) as a function of Cb0/Ca for the
E. Lippert, H. Prigge, Justus Liebigs Ann. Chem. 1962, 659,
81Ϫ89.
4-fluorophenol/18-crown-6 complex; Kf is constant when the ratio
0
Cb0/Ca is over 10
[18]
[19]
T. Gramstad, Spectrochim. Acta 1963, 19, 497Ϫ508.
E. S. Hanrahan, F. Deskins, Proc. West Va. Acad. Sci. 1967,
39, 371Ϫ374.
R. West, D. L. Powell, M. K. T. Lee, L. S. Whatley, J. Am.
Chem. Soc. 1964, 86, 3227Ϫ3229.
T. M. Barakat, M. J. Nelson , S. M. Nelson, A. D. E. Pullin,
Trans. Faraday Soc. 1966, 62, 2674Ϫ2684.
R. West, L. S. Wilson, D. L. Powell, J. Organomet. Chem. 1979,
178, 5Ϫ9.
T. Iijima, H. Kakiuchi, Tetrahedron 1979, 35, 299Ϫ302.
L. Bellon, R. W. Taft, J.-L. M. Abboud, J. Org. Chem. 1980,
45, 1166Ϫ1168.
H. Fritzsche, Spectrochim. Acta 1965, 21, 799Ϫ813.
C. Laurence, M. Berthelot, M. Helbert, Spectrochim. Acta 1985,
41A, 883Ϫ892.
A. Massat, P. Guillaume, J. P. Doucet , J. E. Dubois, J. Mol.
Struct. 1991, 244, 69Ϫ86
J. L. E. Erickson, W. J. Ashton, J. Am. Chem. Soc. 1941, 63,
1767Ϫ1771.
J.-Y. Le Questel, Ph. D. Thesis, University of Nantes, 1991.
D. Clotman, D. Van Lerberghe, T. Zeegers-Huyskens, Spectro-
chim. Acta 1970, 26A, 1621Ϫ1631.
M. Berthelot, C. Laurence, unpublished work.
R. M. Badger, S. H. Bauer, J. Chem. Phys. 1937, 5, 839Ϫ851.
R. M. Badger , J. Chem. Phys. 1940, 8, 288Ϫ289.
T. Gramstad, S. Husebye, K. Maartmann-Moe, Acta Chem.
Scand. 1985, B39, 767Ϫ771.
A.C. Legon, D. J. Millen , Chem. Soc. Rev. 1987, 16, 467Ϫ498.
A.C. Legon, J. C. Thor, Chem. Phys. Lett. 1994, 227, 472Ϫ479.
A.C. Legon, Chem. Phys. Lett. 1995, 247, 24Ϫ31.
P. Murray-Rust, J.P. Glusker, J. Am. Chem. Soc. 1984, 106,
1018Ϫ1025.
0
Cb0/Ca
Kf
0.42 0.85 1.73 3.36 7.15 10.7 13.7 15.8
142
131
122
114
97.1 95.9 95.7 95.8
[20]
[21]
[22]
1:1, 2:1, ... complexes[a]
1:1 complexes: Kf ϭ 95.8
[a] By assuming the presence of only a 1:1 complex and a 2:1 com-
plex (2 molecules of 4-fluorophenol for 1 crown ether), we obtain,
by the Huyskens method[30], Kf(1:1) ϭ 93.7 l molϪ1 in good
agreement with the value obtained when the crown ether is in ten-
to sixteenfold excess.
[23]
[24]
[25]
[26]
Infrared Shifts: The wavenumber shifts of the OH bands of meth-
[27]
[28]
anol at 3644 cmϪ1 and 4-fluorophenol at 3614 cmϪ1 are defined as
∆ν1(OH)
ϭ 3644 Ϫ ν1(OH···O) and ∆ν2(OH) ϭ 3614 Ϫ
ν2(OH···O), respectively. They are known to ± 2 cmϪ1 if the
ν(OH···O) band is almost symmetrical, but asymmetry and math-
ematical decomposition cause larger errors.
[29]
[30]
[31]
[32]
[33]
[34]
[1]
D. Gurka , R. W. Taft, J. Am. Chem. Soc. 1969, 91, 4794Ϫ4801.
[2]
E. M. Arnett, L. Joris , E. Mitchell , T. S. S. R. Murty, T. M.
Gorrie, P. v. R. Schleyer, J. Am. Chem. Soc. 1970, 92,
2365Ϫ2377.
[35]
[36]
[37]
[38]
[3]
M. D. Joesten, R. S. Drago, J. Am. Chem. Soc. 1962, 84,
2037Ϫ2039.
[4]
D. Gurka, R. W.Taft, L. Joris, P. v. R. Schleyer, J. Am. Chem.
Soc. 1967, 89, 5957Ϫ5958.
[5]
M. H. Abraham, P. L. Grellier, D. V. Prior, J. J. Morris, P. J.
[39]
[40]
[41]
J. A. Shea, S. G. Kukolich, J. Chem. Phys. 1983, 78, 3545Ϫ3551.
Z. Yoshida, E. Osawa, J. Am. Chem. Soc. 1966, 88, 4019Ϫ4026.
R.W. Taft, R. D. Topsom, Progr. Phys. Org. Chem. 1987, 16,
1Ϫ83.
Taylor, J. Chem. Soc. Perkin Trans. 2 1990, 521Ϫ529.
[6]
A. Chardin, Ph. D. Thesis, University of Nantes, 1997.
[7] [7a]
M. Berthelot, M. Helbert, C. Laurence, J.-Y. Le Questel, J.
[7b]
Phys. Org. Chem. 1993, 6, 302Ϫ 306. Ϫ
E. D. Raczynska,
[42]
[43]
[44]
[45]
R. W. Taft, F. Anvia, M. Taagepera, J. Catalan, J. Elguero, J.
Am. Chem. Soc. 1986, 108, 3237Ϫ3239.
C. Laurence, P. Nicolet, J. Chem. Soc., Perkin Trans. 2 1988,
1491Ϫ1494.
R. West, L. S. Whatley, K. J. Lake, J. Am. Chem. Soc. 1961,
83, 761Ϫ764.
[8]
[9]
C. Laurence, M. Berthelot, J.-Y. Le Questel, M. J. El Ghomari,
J. Chem. Soc., Perkin Trans. 2 1995, 2075Ϫ2079.
F. Besseau , C. Laurence, M. Berthelot, Bull. Soc. Chim. Fr.
1996, 133, 381Ϫ387.
M. Berthelot, C. Laurence, M. Safar, F. Besseau, J. Chem. Soc.,
Perkin Trans. 2, in press.
P. P. Graczyk, M. Mikoajczyk, Top. Stereochem. 1994, 31,
159Ϫ349.
[10]
C. Laurence, M. Berthelot, M. Helbert, K. Sraıdi, J. Phys.
¨
Chem. 1989, 93, 3799Ϫ3802.
[97356]
Eur. J. Org. Chem. 1998, 925Ϫ931
931