INTERPRETATION OF HYDROXYLIC SOLVENT EFFECTS BASED ON CORRELATIONS WITH SOLVENT PARAMETERS. REACTION OF Et3N WITH EtI

Article abstract of DOI:10.1135/cccc19940898

Hydroxylic solvent effects on 20 rate constants, k, of the Menschutkin reaction of Et3N with EtI are unravelled and rationalized by means of multiple linear regression equations.To perform this analysis new k values in 2 mono- and 9 dialcohols are obtained.New values are also presented for the Kamlet and Taft solvatochromic parameters ?*, α and β of 1-hexanol.The results show that the solvent dipolarity, polarizability and cohesive energy density are main properties influencing the reactivity, for the set of studied solvents.

Full text of DOI:10.1135/cccc19940898

898
Calado et al.:
INTERPRETATION OF HYDROXYLIC SOLVENT EFFECTS
BASED ON CORRELATIONS WITH SOLVENT PARAMETERS.
REACTION OF Et3N WITH EtI
a
a
a
Antonio R. T. CALADO , Lidia M. V. PINHEIRO , Lidia M. P. C. ALBUQUERQUE ,
a
b
b
b
Raquel M. C. GONCALVES , Martí ROSES , Clara RAFOLS and Elisabeth BOSCH
a Centro de Electroquimica e Cinetica da Universidade de Lisboa (FFUL, FCUL),
Cd. Bento da Rocha Cabral, 14, 1200 Lisboa, Portugal
b Department de Quimica Analitica,
Universitat de Barcelona, Diagonal 647, 08028, Barcelona, Spain
Received May 24, 1993
Accepted September 20, 1993
Hydroxylic solvent effects on 20 rate constants, k, of the Menschutkin reaction of Et3N with EtI are
unravelled and rationalized by means of multiple linear regression equations. To perform this ana-
lysis new k values in 2 mono- and 9 dialcohols are obtained. New values are also presented for the
Kamlet and Taft solvatochromic parameters π*, α and β of 1-hexanol. The results show that the sol-
vent dipolarity, polarizability and cohesive energy density are the main properties influencing the re-
activity, for the set of studied solvents.
The study of the effect of medium in Menschutkin reactions started many years ago.
However, experimental and theoretical studies of solvent effects on these reactions still
attract a large popularity due to the fact that these reactions exhibits a high sensitivity
to solvent effects and that its quantitative description by empirical parameters was
largely improved during the last decades. Reviews of this matter were recently publish-
ed1,2
.
Accepting the basic principle that solvent effects on chemical reactivity are essen-
tially similar in their nature and that only a few mechanisms of interaction take place
between solvent and solute molecules, a general way of treating data in order to express
these interactions is the method of linear solvation energy relationships (LSER), using
a suitable model to connect the macroscopic solvent parameters and the microscopic
details of the physico-chemical interaction processes. In practice, multiparameter corre-
lations of the logarithm of the rate constants, log k, or the Gibbs energy of activation,
∆G≠, with empirical solvent parameters have been successfully applied to several reac-
tions1 – 11
.
We started our studies of LSER presenting work on the solvolysis of tert-butyl
halides in hydroxylic solvents, a classic example of a unimolecular reaction9 – 11. We
continue our studies of LSER, investigating the Menschutkin reaction of triethylamine
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Hydroxylic Solvent Effects
899
(Et3N) with ethyl iodide (EtI) in water, 10 monoalcohols and 9 dialcohols. From the
kinetic point of view, a conductometric technique was chosen to construct concentra-
tion–time plots, taking account of the actual ionic reaction product, the tetraalkylam-
monium salt (Et4N+I−), as described before12. Rate constants of Et3N/EtI reactions in
mono- and dialcohols were then experimentally determined at 25 °C. Between the most
suitable empirical equations for evaluation of the medium effects, we chose Eqs (1)
(refs9 – 11) and (2) (refs5,6):
log k = a0 + a1 f(ε) + a2 g(η) + a3 ENT + a4 C
(1)
(2)
′
′
′
′
′
log k = a0 + a1 π + a2 α + a3 β + a4 C .
Data for the solvent parameters are from literature2,9,13, except the values for the Kam-
let and Taft solvatochromic parameters π*, α, and β of 1-hexanol. Both multiparameter
approaches take into consideration the three dominant types of interactions: non-spe-
cific and specific solvent–solute interactions and solvent–solvent interactions from the
cavity effect. In Eqs (1) and (2), f(ε) is a measure of the dipolarity effect, g(η) a
measure of the polarizability effect, ETN is the normalized Dimroth and Reichardt par-
ameter which consists of a blend of dipolarity and solvent HBD acidity, C is the solvent
contribution to the cavity term, π* a measure of solvent dipolarity/polarizability, α is a
scale of hydrogen bond donor – HBD acidities and β a scale of hydrogen bond acceptor
– HBA basicities.
The results are discussed in terms of the fundamental solute–solvent–solvent interac-
tions mechanisms and compared with previous results.
EXPERIMENTAL
Apparatus. A Wayne–Kerr B905 conductometric bridge and a modified version of the Shedlovsky
cells were used to obtain the conductometric data in order to calculate the kinetic rate constants. A
Beckman DU-7 spectrophotometer, with a 10 mm cells, was used for acquisition of the absorbance
data of the solvatochromic indicators.
Solvatochromic dyes. We use four solvatochromic indicators sensitive to the solvent polarity π* –
4-nitroanisole (1), 2-nitroanisole (2), 4-ethylnitrobenzene (3) and N,N-diethyl-4-nitroaniline (4), two
sensitive to π* and hydrogen bond acceptor capability β – 4-nitroaniline (5) and 4-nitrophenol (6),
and one particularly sensitive to π* and hydrogen bond donor capability α – 2,6-diphenyl-4-(2,4,6-
triphenylpyridinio)phenolate or Reichardt’s Dye (7) to obtain the π*, α and β values of 1-hexanol*.
Dyes (1) and (5) were from Merck; (2), (3), (4) and (7) were from Aldrich; (7) was from Scharlau.
* π*, α and β values for 1-hexanol have been already published6. However, we repeat the procedure in
order to obtain a coherent list of data.
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

900
Calado et al.:
Substrats and solvents. The triethylamine and ethyl iodide were from Fluka (≥99.5%) and they
were used without further purification. Alcohols were from Merck (≥99%). The water content was
kept lower than 0.02%.
Procedure. The mole fraction–time plots for the ionic reaction product, the tetraalkylammonium
salt, were based on standard curves relating mole fractions to experimental conductances. The cali-
bration method has been exemplified before12. At least five different standard solutions of tetraethyl-
ammonium iodide were used to obtain the relationships. The cells, filled fully with the standard
solutions or the reactant solutions were placed within the thermostat (25.00 ± 0.01 °C). After the
thermal equilibrium had been reached, we started conductance readings for both types of solutions at
regular intervals of time, chosen according to the systems. All the solutions were prepared by weight
after dry oxygen-free nitrogen has been passed through. All the observed kinetics show an ex-
perimental error < 1% in k (mole fraction−1 s−1).
The absorbance data for each indicator were recorded as previously described13. The spectro-
photometer cell was externally thermostatted at 25.0 ± 0.2 °C with a water jacket. The maximum
wavelength of each indicator was obtained after numerical smoothing of the absorbance data. The
parameter π* was obtained from indicators (1) to (4), β from indicators (5) and (6) and the π* pre-
viously computed, and α from indicator (7) and the π* and β values.
RESULTS AND DISCUSSION
In Table I we summarized the available second-order rate constants for the Menschut-
kin reaction Et3N with EtI in water and alcohols. Most of the data in this table are new;
those which are cited from the literature are suitable footnoted. A compilation of the
values of the empirical solvent parameters, f(ε), g(η), ENT , π*, α, β and C for the set of
studied solvents is also shown in Table I.
The non-collinearity assumption between any two of the chosen parameters, with
respect to Eqs (1) and (2), was tested. As before9, a strong collinearity was observed
between f(ε) and ENT . Some influence of solvent dipolarity in ENT is most probably the
reasonable explanation, as stated by several authors using different sets of solvents16 – 18
.
We also observed collinearities between ENT and C, π* and β and α and C. In these
cases, however, we believe, as Glikberg and Marcus pointed out19, that there are no
relevant relationships from a physico-chemical point of view, since much smaller corre-
lation coefficient values can be found for other sets of solvents.
Although both complete versions of Eqs (1) and (2) accommodate successfully the
rate constant values of the reaction under study (r 0.95; s 0.2), the decision about
the best fits was achieved by reference to a suitable criterion. We decided to use the
Ehrenson’s criterion20. The f function was then calculated for the complete and trun-
cated versions of Eqs (1) and (2) showing the lower standard deviations, s, and from the
ratio fj/fj + 1, where j represents the number of estimated coefficients a in the simple and
multiple linear regressions, we were able to conclude about the confidence levels and,
consequently, about the statistically preferred correlations.
With respect to Eq. (1), f2 = 0.003883, f3 = 0.002894 and f4 = 0.002185 (and, conse-
quently, f2/f3 = 1.341 and f3/f4 = 1.325), which means that the hypothesis that a par-
ameter set provides as a good fit as another set obtained by the removal of one
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Hydroxylic Solvent Effects
901
parameter from the previous one must be rejected at a confidence level 99.9%, going
from two to one parameter, and at a confidence level 99.6%, going from three to two
parameters. In conclusion, the statistically preferred correlation is:
log k = −(26.4 ± 2.6) + (41.4 ± 6.2) f(ε) + (10.9 ± 3.1) g(η) +
+ (0.48 ± 0.21) . 10−3
C
(N = 20; r = 0.953; s = 0.22) .
(3)
TABLE I
Rate constants k (mole fraction−1 s−1)a of the Menschutkin reaction of triethylamine with ethyl iodide
in hydroxylic solvents, at 25 °C and selected properties of the solventsb
C/103
MPa
n
Solvent
−log k
f(ε)
g(η)
ETN
π*
α
β
1
Water
2.718c 0.49048 0.20569 1.000
3.999d 0.47738 0.20311 0.762
4.288d 0.47006 0.22147 0.654
4.571d 0.46420 0.23467 0.617
4.389d 0.46327 0.23011 0.546
4.706d 0.45836 0.24210 0.602
4.491d 0.45604 0.24087 0.506
4.897d 0.44792 0.24776 0.568
1.13
0.60
0.55
0.53
0.48
0.54
0.54
0.50
0.51
0.48
1.16
1.09
0.88
0.79
0.68
0.74
0.54
0.73
0.64
0.74
0.50
0.73
0.80
0.85
0.93
0.84
0.91
0.88
0.93
0.91
2.307
0.887
0.703
0.590
0.552
0.485
0.488
0.497
0.482
0.497
2
3
4
5
6
7
8
9
Methanol
Ethanol
1-Propanol
2-Propanol
1-Butanol
2-Butanol
1-Pentanol
2-Methyl-1-butanol 4.980 0.45350 0.24712 0.534
10 3-Methyl-1-butanol 4.845 0.45220 0.24627 0.565
11 1-Hexanol
5.087d 0.44565 0.25190 0.554e 0.52e
0.68e 0.86e 0.471f
12 1,2-Ethanediol
13 1,2-Propanediol
14 1,3-Propanediol
15 1,2-Butanediol
16 1,3-Butanediol
17 2,3-Butanediol
18 1,5-Pentanediol
19 Diethylene glycol
20 Triethylene glycol
3.725 0.48037 0.25927 0.790
3.670 0.47422 0.25958 0.722
3.304 0.47886 0.26345 0.747
3.952 0.46718 0.26240 0.676
3.777 0.47420 0.26407 0.682
3.977 0.46614 0.25885 0.651
3.912 0.47289 0.26843 0.654
3.257 0.47670 0.26745 0.713
3.342 0.46900 0.27173 0.704
0.89
0.76
0.84
0.71
0.75
0.75
0.76
0.92
0.88
0.88
0.83
0.80
0.80
0.76
0.68
0.70
0.72
0.66
0.72
0.78
0.77
0.71
0.74
0.88
0.82
0.67
0.69
0.887
0.881
0.847
0.60
0.562
0.602
0.603
0.615
0.480
a
b
Standard deviation of k < 1%, except for 1-hexanol ( 3%). Values listed in refs2,9,13, except
otherwise indicated. c Value calculated from data in ref.14. d Values from ref.12. e This work. f Value
calculated from data in ref.15.
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

902
Calado et al.:
With respect to Eq. (2), using the same criterion, we consider that the following two
coefficients correlation is the best statistical fit:
log k = −(6.37 ± 0.20) + (3.33 ± 0.29) π
(N = 20; r = 0.939; s = 0.23) .
(4)
′
The statistical quantities a0 and a0 correspond to the value of the solvent dependent
property log k, in the gas phase or in an inert solvent. We may note that hexane shows
the smallest experimentally determined k value, log k = −6.990 (ref.14). In this context,
Eq. (4) seems to present a more direct logical connection between the statistical ap-
proach and the physico-chemical interpretation. Unfortunately, log k value in gas phase is
not available and it is difficult to assign an “inert” solvent for the Menschutkin reactions.
Concentrating our attention in Eq. (3), we may observe, for the set of studied sol-
vents, that: (i) the non-specific solvent–solute interactions, represented by f(ε) and g(η), are
both relevant, i.e., rate constants of Et3N with EtI reaction show a marked dependence
on solvent dipolarity and polarizability; (ii) the specific solvation, represented by ENT , is
unrelevant, i.e., the solvent–solute interaction on account of the solvent HBD acidity does
not influence the Menschutkin reaction to a relative significant extent; (iii) solvent–sol-
vent interactions, measured by the C parameter, are important, i.e. disruption and reor-
ganization of solvent molecules in order to create a suitable cavity to accommodate the
substrate molecule contributes to the overall solvent effect on the reaction under study.
If we extended our analysis to the results obtained by Eq. (4) for the hydroxylic
solvents, we may conclude that: the solvent dipolarity-polarizability (π* term) is the main
factor influencing the rate constants and, conversely, the acidity and basicity of the solvent
(α and β terms) are not significant. These two conclusions are in good agreement with
those taken from Eq. (3). A different interpretation, however, can be drawn from the
relative importance of the cavity term. Nevertheless, it is interesting to note that, ac-
cording to Eq. (1), the best two parameters correlation does not include the C term and that,
according to Eq. (2), the best three parameters correlation contains the C (and the α) term.
The reaction between Et3N and EtI has been examined before by Abraham et al.3,6,
Pytela1, and Bekárek and Nevecna8 who found, for a set of solvents including aprotic
and protic solvents (hydrocarbons, aromatics, alcohols)*, that solvent dipolarity was the
* Although within the hypothesis of LSER nothing is stated about the mechanism of the reactions under
study and, consequently, about the possible configurations of the activated complexes, it seems
desirable to apply LSER to a set of solvents for which we may postulate a certain type of transition
state. This way, solvent effects can be descriminated, compared and interpreted in a more solid basis.
This is why we prefer to consider the influence of hydroxylic solvents (water and alcohols – class 3)21
separated from the influence of other classes of solvents.
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Hydroxylic Solvent Effects
903
over-riding factor influencing the reaction rate, with solvent hydrogen-bond acidity and
basicity being statistically much less significant. Although our set of studied solvents
differs considerably from those used before, our conclusions agree in the fundamental
points. Within the set of hydroxylic solvents used in this work proton-donor ability of
medium did not strongly affect the reaction; however, this was not the case described
by other authors1,8, using several classes of solvents simultaneously.
CONCLUSIONS
In the present work, we focussed our attention on the solvent effects on the bimolecular
reaction of Et3N with EtI in water and alcohols. Dipolarity and polarizability influence
the Menschutkin reaction, most probably stabilising the ion-pair like transition state
and, consequently, increasing the rate constant and, conversely, the effectiveness of the
solvents as hydrogen bond donors seems to be irrelevant in this case. Presumably, as
Abraham et al.6 pointed out, the stabilisation of the leaving iodide ion in the transition
state by electrophilic assistance is counterbalanced by stabilisation of triethylamine in
the initial state. Finally, the nucleophilic solvent assistance is not significant for the
studied Menschutkin reaction in the hydroxylic solvents water, mono- and dialcohols.
We are grateful to the Instituto Nacional de Investigacao Científica of Portugal (project 1C and
3C-CECUL) and to the Direccion General de Investigacion Cientifica y Tecnica of Spain (project PB91-
0262) for financial support.
REFERENCES
1. Pytela O.: Collect. Czech. Chem. Commun. 53, 1333 (1988).
2. Reichardt C.: Solvents and Solvent Effects in Organic Chemistry, 2nd ed..VCH, Weinheim 1988.
3. Abraham M. H.: Pure Appl. Chem. 57, 1055 (1985).
4. Kamlet M. J., Taft R. W.: Acta Chem. Scand., B 39, 611 (1985).
5. Abraham M. H., Doherty R. M., Kamlet M. J., Harris J. M., Taft R. W.: J. Chem. Soc., Perkin
Trans. 2 1987, 913.
6. Abraham M. H., Grellier P. L., Abboud J.-L. M., Doherty R. M., Taft R. W.: Can. J. Chem. 66,
2673 (1988).
7. Nevecna T., Bekarek V.: Collect. Czech. Chem. Commun. 55, 180 (1990).
8. Bekarek V., Nevecna T.: Collect. Czech. Chem. Commun. 56, 874 (1991).
9. Goncalves R. M. C., Simoes A. M. N., Albuquerque L. M. P. C.: J. Chem. Soc., Perkin Trans. 2
1990, 1379.
10. Goncalves R. M. C., Simoes A. M. N., Albuquerque L. M. P. C.: J. Chem. Soc., Perkin Trans. 2
1991, 825.
11. Goncalves R. M. C., Simoes A. M. N., Leitao R. A. S. E., Albuquerque L. M. P. C.: J. Chem.
Res. (S) 1992, 330; (M) 1992, 2701.
12. Viana C. A. N., Calado A. R. T., Pinheiro L. M. V.: J. Chem. Res. (S) 1992, 6; (M) 1992, 173.
13. Goncalves R. M. C., Simoes A. M. N., Albuquerque L. M. P. C., Roses M., Rafols C., Bosch E.:
J. Chem. Res. (S) 1993, 214; (M) 1993, 1380.
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

904
Calado et al.:
14. Abraham M. H., Grellier P. L.: J. Chem. Soc., Perkin Trans. 2 1976, 1735.
15. Wadso I.: Acta Chem. Scand. 20, 544 (1966).
16. Kamlet M. J., Abboud J.-L., Taft R. W.: J. Am. Chem. Soc. 99, 6027, 8325 (1977).
17. Kamlet M. J., Abboud J.-L., Abraham M. H., Taft R. W.: J. Org. Chem. 48, 2877 (1983).
18. Bekarek V., Jurina J.: Collect. Czech. Chem. Commun. 47, 1060 (1982).
19. Glikberg S., Marcus Y.: J. Solution Chem. 12, 255 (1983).
20. Ehrenson E.: J. Org. Chem. 44, 1793 (1979).
21. Dutkiewicz M.: J. Chem. Soc., Faraday Trans. 86, 2237 (1990).
Collect. Czech. Chem. Commun. (Vol. 59) (1994)