Concentration effects in the nucleophilic reactions of tertiary amines in aqueous solutions. Alkylation of amines with ethylene chlorohydrin

Article abstract of DOI:10.1134/S0036024410120113

During the alkylation of tertiary amines with ethylene chlorohydrin in aqueous solutions, the initial rates and degrees of conversion attained within control periods of time showed nonstandard dependences on the initial concentrations of reagents. The character of the concentration dependences depended on the structure of the hydrocarbon fragments and the available sets of functional groups in amines. On the basis of our results and viscosimetric data for the model systems, we assumed that these effects were due to pre-reaction association involving the reagents and products.

Full text of DOI:10.1134/S0036024410120113

ISSN 0036ꢀ0244, Russian Journal of Physical Chemistry A, 2010, Vol. 84, No. 12, pp. 2071–2076. © Pleiades Publishing, Ltd., 2010.
Original Russian Text © O.A. Kazantsev, D.S. Baruta, K.V. Shirshin, A.P. Sivokhin, D.M. Kamorin, 2010, published in Zhurnal Fizicheskoi Khimii, 2010, Vol. 84, No. 12,
pp. 2265–2270.
PHYSICAL CHEMISTRY
OF SOLUTIONS
Concentration Effects in the Nucleophilic Reactions
of Tertiary Amines in Aqueous Solutions. Alkylation of Amines
with Ethylene Chlorohydrin
O. A. Kazantsev, D. S. Baruta, K. V. Shirshin, A. P. Sivokhin, and D. M. Kamorin
Alekseev State Technical University, Nizhni Novgorod, Russia
Dzerzhinsk Polytechnical Institute, Nizhni Novgorod oblast, Russia
eꢀmail: dariabaryta@mail.ru
Received February 12, 2010
Abstract—During the alkylation of tertiary amines with ethylene chlorohydrin in aqueous solutions, the iniꢀ
tial rates and degrees of conversion attained within control periods of time showed nonstandard dependences
on the initial concentrations of reagents. The character of the concentration dependences depended on the
structure of the hydrocarbon fragments and the available sets of functional groups in amines. On the basis of
our results and viscosimetric data for the model systems, we assumed that these effects were due to preꢀreacꢀ
tion association involving the reagents and products.
DOI: 10.1134/S0036024410120113
INTRODUCTION
nomethacrylates [9] and amino(meth)acrylamides
[10]; initiated copolymerization of these salts with
acrylamide reactions [11]. These association effects
showed themselves as distinct effects of the initial conꢀ
centrations of reagents inconsistent with the classical
concepts. These effects were therefore expected to be
also significant for nonradical nucleophilic substituꢀ
tion reactions of different tertiary amines in water.
Importantly, studies of the concentration effects not
only lead to the optimization of the reaction condiꢀ
tions, but also provide a method of investigating assoꢀ
ciation in complex systems with multifunctional comꢀ
ponents.
Quaternary ammonium salts (QASs) are widely
used in industry and house keeping. They are generally
obtained by the Menschutkin reaction, i.e., Nꢀalkylaꢀ
tion of tertiary amines with haloalkyls [1]. For the
reaction carried out in homogeneous aqueous soluꢀ
tions, the reagents should contain hydrophilic funcꢀ
tional groups in addition to the reaction centers to
ensure their high solubility. One of accessible waterꢀ
soluble halogen derivatives is ethylene chlorohydrin
(ECH). The products of its interactions with tertiary
amines are the derivatives of choline chloride [2, 3]:
+
–
R3N + ClCH2CH2OH
[R3NCH2CH2OH]Cl. (1)
The aim of this work was to study the effects of
reagent concentrations on the pathway of the Mensꢀ
chutkin reaction of ECH and amines with different
structures in aqueous solutions at low and high conꢀ
versions. As amines, we mainly used reagents with
(CH3)2NCH2CH2X reaction centers of the same type.
They are presented together with notation in the table.
As is well known [12], the activity of amines in quaterꢀ
nization with alkyl halides primarily depends on the
spatial structure and induction effects of substituents
at the nitrogen atom. By varying the X substituent
lying far from the amino group and containing the
alcohol, nitrile, and unsubstituted or substituted
amide groups, we could change the type of the interꢀ
molecular association interaction. The induction and
steric factors of the reaction center did not change;
this was confirmed by insignificant fluctuations in the
The preꢀreaction association of reagents is one of
the insufficiently studied problems concerned with
reactions (1) in homogeneous media. It has recently
attracted more interest and was considered for differꢀ
ent systems. The series of papers [4, 5] described a
model of liquidꢀstate reactions in associated media
using the similarity in the behavior of alcohol associꢀ
ates with flexibleꢀchain polymers in solutions. The
model was constructed on the basis of the reactions of
alcohols with isocyanates. Aminolysis of esters was the
object of studies described in [6, 7]. It was shown that
these processes could not be described adequately
without taking into account associative interactions.
For tertiary (meth)acryl amines, it was recently
shown that association of reagents in aqueous soluꢀ
tions produced a strong effect on both nonradical
(hydrolysis of aminomethacrylates [8]) and radical
total steric RS and induction Σσ* constants of subꢀ
Σ
(spontaneous homopolymerization of the salts of amiꢀ stituents at the amine center, determined by the
2071

2072
KAZANTSEV et al.
Tertiary amines used and the calculated total inductions (Σσ*) and steric (
R3N Formula
DE HOCH2CH2N(CH3)2
ΣRS) constants of substituents at the amine nitrogen
Σσ
*
–
Σ
RS
0.1023
0.3607
0.1655
0.1724
0.1725
0.1726
–
4.29
4.53
4.61
4.72
5.31
4.95
4.43
DN
NCCH2CH2N(CH3)2
DA
H2NCOCH2CH2N(CH3)2
DA
CH2=CHCONH(CH2)3N(CH3)2
CH2=CHCONHC(CH3)2CH2CH2N(CH3)2
CH2=C(CH3)CONH(CH2)3N(CH3)2
N(CH2CH2)3N
DDAA
DMA
DABCO
Galkin–Cherkasov method [13, 14]. For substituents
with any structure,
RESULTS AND DISCUSSION
In the first series of experiments, we studied the
effects of the structure of amines on the concentration
effects in the quaternizations of tertiary amines with
ECH in aqueous solutions. For all amines except
DABCO (whose solubility in water restricted the posꢀ
sibility of obtaining high initial concentrations, over
1–2 mmol/g), we observed abrupt deceleration of the
reaction. The dependences of the initial rates on the
initial concentrations of reagents were determined
from the kinetic data. For convenient analysis, the
amines were divided into two groups: the derivatives
with a functional group as an X substituent and Nꢀ
n
σ
(
)
A
i
σ* =
,
(2)
(3)
∑
r2
i =1
i
Ri2
n
⎡
⎤
R = 30log 1 −
,
∑
s
⎢
⎣
⎥
i ⎦
i =14r2
where
(
σА i
)
is the ability of the ith atom of the substitꢀ
uent to show an induction effect, ri is the distance
between the th atom and the reaction center (Å), and
Ri is the covalent radius of the th atom (Å).
i
i
(3ꢀdimethylaminoalkyl)(meth)acrylamide
mers with a methyl substituent at the third or sixth
position relative to the reaction center or without it.
monoꢀ
The above procedure for calculating the steric and
induction constants of substituents was developed for
the tetracoordinated carbon atom as a reaction cenꢀ
ter. In earlier studies of the Nꢀalkylations of tertiary
amines, the use of the procedure was shown to be
correct for the nitrogen atom as a reaction center as
well [15].
The results of this study clearly showed that the
concentration effects differed considerably between
amines with related structures. Figure 1 shows the
effect of the initial reagent concentration on the initial
reaction rate for DABCO and three amines with difꢀ
ferent X acceptor substituents in the third position relꢀ
ative to the amine reaction center (an alcohol group in
DE, nitrile in DN, and unsubstituted amide group in
DA). DN was the least active of all amines used in this
study over the whole range of concentrations. Its quatꢀ
ernization at 303 K was so slow that the temperature
had to be raised to 343 K. In other nucleophilic reacꢀ
tions in aqueous solutions (with acrylic acid [16], or
chloroacetic acid (see the next paper of this series), its
activity was comparable to that of DE or DMA.
EXPERIMENTAL
For experiments we used N,Nꢀdimethylaminoethꢀ
anol (DE), N,Nꢀdimethylaminopropionitrile (DN),
N,Nꢀdimethylaminopropionamide (DA), Nꢀ(3ꢀdiꢀ
methylaminopropyl)methacrylamide (DMA), Nꢀ(3ꢀ
dimethylaminopropyl)acrylamide (DAA), Nꢀ(1,1ꢀdiꢀ
methylꢀ3ꢀdimethylaminopropyl)acrylamide (DDAA),
1,4ꢀdiazabicycloꢀ[2.2.2]octane (DABCO), and ECH
with at least 98% main substance. The reactions of
ECH with tertiary amines were carried out at 303 K
The other three amines from the first group reacted
(the reaction with DN was performed at 343 K); the at comparable rates in dilute solutions. For DE, the
reagents were taken in equimolar ratios. To preclude initial rate increased almost linearly at higher concenꢀ
radical polymerization in the reactions of monomer trations of the starting solutions, but gradually
amines, we introduced hydroquinone in the reaction decreased at concentrations higher than 2 mmol/g.
mixtures. The Menschutkin reaction was monitored For other amines, the reaction abruptly accelerated
chromatographically and according to the consumpꢀ when the concentration increased to 0.7–1.0 mmol/g.
tion of free amino groups, determined by acid–base For DABCO, the acceleration continued until the
titration. The isolated products were identified by concentration reached its maximum possible level.
chemical analysis and NMR and IR spectroscopy. The For DA, the initial rate at first abruptly increased and
specific viscosity of solutions was measured with an then quickly decreased at concentrations of over
Ubbelode viscosimeter (the capillary diameter was 1.1 mmol/g. For DN, the rate also increased to a maxꢀ
chosen to be 0.34–0.99 mm) at 298 K.
imum (corresponding to 1.2 mmol/g) and then relaꢀ
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 12
2010

CONCENTRATION EFFECTS IN THE NUCLEOPHILIC REACTIONS
2073
ν0 × 104, mmol/(g min)
200
tively smoothly decreased. The maximum rates varied
significantly (from 29.5
10–4 mmol/(g min) for DE to
116.7
10–4 mmol/(g min) for DA).
×
2
×
The amines under study thus behaved differently
when the initial concentrations increased in the quatꢀ
ernization of ECH. It is not surprising, therefore, that
the differences showed themselves at high conversions
as well (i.e., when the products accumulated in the
systems). It was unexpected, however, that at a control
synthesis time of 1.5 h, the activity of amines was quite
different from the activity observed at the start of the
reaction (Fig. 2). For example, at an initial concenꢀ
tration of 2 mmol/g the activity series was DA >
DABCO > DE according to the initial rates, but
DABCO > DE > DA according to the attained converꢀ
sions. The dependences of conversion on the initial
concentration had maxima for all amines, but the
positions of the maxima were different for different
reagents and strongly different from those at the initial
concentrations corresponding to the highest initial
rates. In the most concentrated solutions, conversions
were very low, less than 2% for DA and DE and less
than 4% for DN.
7
160
5
120
80
4
3
40
1
6
0
2
4
6
c, mmol/g
Fig. 1. Effects of the initial concentrations of reagents
on the initial rates of the reactions of ECH with (
DN, ( ) DA, ( ) DABCO, and ( ) DE in aqueous
solutions at (
calculated curves.
1, 2)
3
,
4
5
7
6, 7
The differences were still more dramatic for three
Nꢀ(3ꢀdimethylaminoalkyl)(meth)acrylamides (Figs. 3
and 4). The small differences in their structure could
not significantly affect the steric accessibility and elecꢀ
tron density of the amine nitrogen (the latter was conꢀ
3
⎯
) 303 and (
1
,
2
) 343 K. Dashed lines:
X
, %
firmed by the close рКа values (9.23, 9.25, and 9.39 for
20
16
12
8
DAA, DMA, and DDAA, respectively [17]). The
dependences of the initial quaternization rates of these
amines, however, were very different, although all of
them passed through a maximum. For DAA, which
has the least branched framework, the rate increased
monotonously when the concentration increased to
2 mmol/g and then quickly decreased. The introducꢀ
tion of the methyl substituent in the vinyl group
(DMA) led to a shift of the maximum toward low conꢀ
centrations. In the narrow range of the initial concenꢀ
trations 0.5–1.0 mol/g, the rate at first abruptly
increased and then abruptly decreased. The maximum
initial rate was similar to one attained for DAA. In
highly concentrated solutions, the reaction rate was
identically low for both amines.
3
4
1
2
4
0
2
4
6
When two methyl substituents appeared in the
α
c, mmol/g
position of the amide group (DDAA), the maximum
rate increased nearly threefold and the corresponding
concentration of reagents shifted toward large values
(3 mmol/g). When the initial concentration increased
to 3.15 mmol/g (i.e., only by 5%), the rate dropped by
40% and continued to do so as the concentration
increased further, but still remained one order of magꢀ
nitude higher than for the other two aminoamides
under the same conditions.
Fig. 2. Effects of the initial concentrations of reagents on
the conversions in the reactions of ECH with ( ) DN,
) DA, ( ) DABCO, and ( ) DE in aqueous solutions at
) 303 and ( ) 343 K for 90 min.
1
(
(
2
3
4
2–
4
1
DMA, but in moderately concentrated ones for DAA
and DDAA. The activity series according to the maxiꢀ
mum conversion after 1.5 h was DDAA (48%) > DAA
(34%) > DMA (16%). The conversion was minimum
(less than 2.5%) in the most concentrated solutions
The conversion dependences (after the control
period of 1.5 h) differed substantially from the aboveꢀ
described dependences. The maximum conversion
was attained in the least concentrated solutions for
(с0 ~ 4 mmol/g).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 12
2010

2074
KAZANTSEV et al.
ν0 × 104, mmol/(g min)
rates were 25 to 250 times lower than those determined
by the kinetic equation. For moderate initial concenꢀ
trations, however, there were segments (for DA, DH,
and DMA) on which the experimental initial rates
markedly exceeded the calculated ones.
2
5
300
In constructing the calculated curves, we neglected
the change in the character of the medium in different
experiments. As is known, more polar media generally
produce a positive effect on the Menschutkin reaction,
though the character of this effect was not always
unambiguous [18, 19]. The systems under study can
therefore be substantially changed by reducing the
mass fraction of water or by increasing the contents of
less polar organic compounds (amines and ECH). We
performed a series of experiments using mixtures of
water and alcohol in different ratios. After the introꢀ
duction of up to 40% isopropyl alcohol or ethanol into
the solvent, the initial alkylation rates of different
amines were reduced sixꢀ to eightfold, but did not
markedly change when the alcohol fraction was
increased further (Fig. 5). It follows from these data
that in low and moderately concentrated solutions,
the kinetically justified growth of the initial rate due to
higher initial concentrations should be largely comꢀ
pensated by a decrease in the rate due to lower
medium polarity; in more concentrated solutions, the
change in polarity should not drastically lower the
reaction rate.
200
100
4
c
3
1
0
2
4
,
mmol/g
Fig. 3. Effects of the initial concentrations of reagents on
the initial rates of the reactions of ECH with ( ) DMA,
) DAA, and ( ) DDAA in aqueous solutions (303 K).
Dashed lines: calculated curves.
1, 2
(3
4, 5
X
, %
50
30
10
3
Consequently, the inclusion of medium polarity
does not decrease the discrepancy between the forꢀ
mally expected and real data on the concentration segꢀ
ments with the experimental rates higher or lower than
the calculated ones. The same refers to the attained
conversions and the large differences in the concentraꢀ
tion dependences recorded in the quaternizations of
tertiary amines with different structures. One logical
explanation of these effects is the influence of noncoꢀ
valent associations with reagents on the systems under
study.
2
1
0
2
4
c, mmol/g
Fig. 4. Effects of the initial concentrations of reagents on
the conversions in the reactions of ECH with ( ) DMA,
) DAA, and ( ) DDAA in aqueous solutions (303 K,
90 min).
These interactions can be strongly affected even by
small changes in the hydrocarbon fragments of moleꢀ
cules, giving rise to local regions whose composition
differs substantially from the volume average. Associaꢀ
tion of organic compounds in liquid media has
recently stirred the growing interest. The motive force
of these interactions are noncovalent intermolecular
interactions involving polar groups and nonpolar
hydrocarbon fragments [20, 21]. For aqueous soluꢀ
tions of substances, these effects are generally considꢀ
ered less significant. For simple polar organic comꢀ
pounds (acetone, methanol, formamide), however,
the possibility of forming multimolecular associates in
water has been shown using novel experimental techꢀ
niques and computer simulation [22]. For example,
for mixtures of water with formamides in the medium
range of compositions, associates can even give rise to
1
(2
3
To evaluate the concentration effects more adeꢀ
quately, we constructed the calculated dependences of
the initial rates on the concentrations of the starting
reagents. Additional experiments were carried out in
dilute aqueous solutions under the pseudoꢀfirstꢀorder
conditions (with a large excess of one reagent). As
would be expected, the Menschutkin reaction was
firstꢀorder for both reagents under these conditions.
The initial rates were then determined by the second
order kinetic equation for the whole range of concenꢀ
trations using the rate constants found for dilute soluꢀ
tions. A comparison of the experimental and calcuꢀ
lated data (Figs. 1 and 3) showed that the estimated regions with a microheterogeneous structure.
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 12
2010

CONCENTRATION EFFECTS IN THE NUCLEOPHILIC REACTIONS
2075
ν0 × 104, mmol/(g min)
ηsp
ηsp
12
60
1
3
1
20
8
15
10
5
40
20
0
2
5
4
4
2
0
2
4
6
8
0
40
80
c
c, mmol/g
, wt %
Fig. 5. Dependences of the reaction rate constants on the
water content in the solvent for reactions of ECH with
Fig. 6. Dependences of the specific viscosity on the conꢀ
centrations of organic compounds for aqueous solutions of
(
1
) DMA and (
2
) DE in (
1
) water–isopropyl alcohol and
(
1
) DMA, (
and DN, and (
amine ratio was equimolar).
2
3) ethanol and DMA, (4) ethanol
(2
) water–ethanol mixtures at reagent concentrations of
0.5 mmol/g at 303 K.
Association is often confirmed by the nontrivial follows from the analysis of the experimental and calꢀ
character of the experimental concentration depenꢀ culated kinetic data, there are concentration regions
dences of the physical properties of solutions, among of “favorable” (accelerating) and “unfavorable”
which viscosity is the most sensitive property [23]. We (decelerating) effects of association on the Menschutꢀ
measured the specific viscosities of aqueous tertiary kin reaction in aqueous solutions. Associations can
amines and alcohols (to avoid chemical reactions in well be recorded experimentally, but the particular
the course of analyses, we used model systems, in structures of associates in aqueous solutions of binary
which ethylene chlorohydrin was replaced with ethaꢀ organic systems is a complex problem that will probaꢀ
nol) (Fig. 6). The obtained viscosity isotherms had bly remain open to discussion for a long time, espeꢀ
maxima for all compounds, as typically observed for cially because they can differ substantially in solutions
systems with active association [23]; the concentraꢀ with different ratios of water and organic compounds.
tions of the components and viscosities corresponding This will be treated in our next papers on the basis of
to the maxima differed strongly. This can be explained other compounds.
by the fact that noncovalent interactions are individual
for amines with different sets of functional groups or
REFERENCES
different structures of the carbon skeleton. The viscosꢀ
ity isotherm (Fig. 6, curve ) showed a smoothened
4
1. Chemical Encyclopedy, in 5 vols., Ed. by I. L. Knuꢀ
nyants (Sov. entsiklopediya, Moscow, 1998), vol. 1
[in Russian].
maximum for a mixture of ethanol with DN (its nitrile
group can form one hydrogen bond). A more distinct
maximum was observed for water–ethanol solutions
of DE (the hydroxyl group of the amino alcohol can
form two hydrogen bonds). The most distinct maxiꢀ
2. O. S. Sverdlikovskaya, M. V. Burmistr, and V. Kh. Shapka,
Vopr. Khim. Khim. Tekhnol., No. 2, 60 (2009).
mum was found for DMA solutions (curve
amide group contains three potential hydrogen bondꢀ
ing sites. For comparison, data for aqueous ethanol
3
), whose
3. K. V. Shirshin, S. A. Kazakov, O. A. Kazantsev, and
S. M. Danov, Zh. Obshch. Khim. 68, 1038 (1998)
[Russ. J. Gen. Chem. 68, 989 (1998)].
(curve 2) and DMA (curve 1) are presented. Note that
the association of aminoamide in water is much stronꢀ
ger than in aqueous alcohol solutions.
4. D. N. Tarasov, R. P. Tiger, and S. G. Entelis, in Polyꢀ
mers, Collected vol. (IKhF RAN, Moscow, 2000),
Vol. 2, p. 59 [in Russian].
To summarize, viscosimetric data showed that
associations took place in aqueous solutions of various
tertiary amines and hydroxylꢀcontaining organic comꢀ
pounds (not only ECH, but also QASs). The specific
character of the associations probably led to differing
concentration effects in the chemical reactions. As
5. R. P. Tiger, M. A. Levina, S. G. Entelis, and
M. A. Andreev, Kinet. Katal. 43, 703 (2002) [Kinet.
Catal. 43, 662 (2002)].
6. N. V. Ptitsyna, R. P. Tiger, and S. G. Entelis, Vysokoꢀ
mol. Soedin., Ser. B 40, 1552 (1998) [Polymer Sci., Ser.
B
40, 315 (1998)].
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 12
2010

2076
KAZANTSEV et al.
7. A. N. Zelenyuk, Physical Interactions in Chemically 16. A. P. Malyshev, K. V. Shirshin, O. A. Kazantsev, et al.,
Controlling Systems (MFTI, Moscow, 1991) [in Rusꢀ
sian].
Zh. Prikl. Khim. 79, 1650 (2006) [Russ. J. Appl. Chem.
79, 1629 (2006)].
8. N. A. Kuznetsova, K. V. Shirshin, O. A. Kazantsev,
et al., Zh. Prikl. Khim. 76, 1150 (2003) [Russ. J. Appl.
Chem. 76, 1117 (2003)].
17. O. A. Kazantsev, K. V. Shirshin, S. A. Kazakov, and
S. M. Danov, Zh. Obshch. Khim. 66, 2018 (1996)
[Russ. J. Gen. Chem. 66, 1958 (1996)].
9. O. A. Kazantsev, N. A. Kuznetsova, K. V. Shirshin,
et al., Vysokomol. Soedin., Ser. A 45, 572 (2003) [Polyꢀ
mer Sci., Ser. A 45, 338 (2003)].
18. C. Reichardt, Solvents and Solvents Effects in Organic
Chemistry (Wiley VCH, Weinheim, 2003; Mir, Moscow,
1991).
10. O. A. Kazantsev and K. V. Shirshin, Polymer 45, 5021
19. S. G. Entelis and R. P. Tiger, Reaction Kinetics in the
Liquid Phase (Khimiya, Moscow, 1973; Wiley, New
York, 1976).
(2004).
11. A. P. Sivokhin, Extended Abstract of Candidate’s Disꢀ
sertation in Chemistry (Nizhegor. Gos. Univ., Nizh.
Novgorod, 2008) [in Russian].
12. A. A. Matveev, Zh. P. Piskunova, V. A. Pal’m, and
A. F. Popov, Reakts. Sposobn. Org. Soedin. 22, 110
(1985).
20. G. V. Korolev, A. A. Il’in, E. A. Sizov, and M. M. Moꢀ
gilevich, Zh. Obshch. Khim. 69, 1636 (1999) [Russ. J.
Gen. Chem. 69, 1563 (1999)].
21. G. V. Korolev, A. A. Il’in, E. A. Sizov, and M. M. Moꢀ
gilev, Zh. Obshch. Khim. 69, 1630 (1999) [Russ. J.
Gen. Chem. 69, 1568 (1999)].
13. V. I. Galkin, R. A. Cherkasov, R. D. Sayakhov, et al.,
Zh. Obshch. Khim. 65, 458 (1995).
22. Yu. G. Bushuev, Doctoral Dissertation in Chemistry
(Ivanov. Gos. Khim.ꢀTekhnol. Univ., Ivanovo, 2001).
14. V. I. Galkin, R. D. Sayakhov, and R. A. Cherkasov, Usp.
Khim. 60, 1617 (1991).
15. S. A. Kazakov, O. A. Kazantsev, K. V. Shirshin, et al., 23. Yu. Ya. Fialkov, Solvent as an Agent of Chemical Process
Zh. Org. Khim. 35, 363 (1999). Control (Khimiya, Leningrad, 1990) [in Russian].
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A
Vol. 84
No. 12
2010