Nucleophilic catalysis of nitrosation: relationship between nitrosating agent equilibrium constant...
Transcript of Nucleophilic catalysis of nitrosation: relationship between nitrosating agent equilibrium constant...
Nucleophilic catalysis of nitrosation:
relationship between nitrosating agent
equilibrium constant and catalyst nucleophilicity
Gabriel da Silva, Eric M. Kennedy* and Bogdan Z. Dlugogorski
Process Safety and Environment Protection Research Group, School of Engineering,
The University of Newcastle, Callaghan, NSW 2308, Australia; Fax: (+61 2) 4921
6920, E-mail: [email protected]
This paper develops a relationship between the equilibrium constant for the formation
of a nitrosating agent and nucleophilicity of the catalyst, to predict equilibrium
constants which until present have proven unobtainable by experimental means. This
relationship is deduced from a fundamental consideration of the Gibbs free energy of
formation of relevant species and includes Edwards’ nucleophilic parameter. A large
set of experimental measurements collected from the literature is applied to obtain the
parameters of the model. Subsequently, the relationship is employed to estimate the
equilibrium constants for the formation of nitrosyl iodide and nitrosyl acetate,
allowing the examination of their reactivities with various amine substrates.
Introduction
It is well known that, nitrosation reactions are catalysed by a range of nucleophilic
species (designated X-), thought largely to follow the mechanistic interpretation of
Ridd1. The mechanism involves the nucleophile dehydrating nitrous acid to form a
nitrosating agent of the form ONX, according to equation (1). The equilibrium
constant for this reaction, KONX, is given by equation (2).
HNO2 + X- + H+ ↔ ONX + H2O (1)
]][H][XHNO[]ONX[
2ONX +−=K (2)
Catalysts that have previously been utilised for nitrosation reactions include certain
halides (bromide, chloride and iodide) and some sulphur containing compounds
(thiocyanate, thiourea and thiosulphate). The nitrosating agents formed from these
catalysts subsequently affect the nitrosation of a substrate (designated S) according to
equation (3), where kN is the rate constant for the second-order nitrosation reaction.
ONX + S → ONS+ + X- (3)
During nitrosation under mildly acidic conditions, formation of the nitrosating species
is generally found to be at equilibrium, while the nitrosation reaction itself is rate
determining. This has led to the establishment of the following rate equation to
describe the overall rate of nitrosation.
KONX
kN
]][H][X[S][HNO2NONX+−= kKr (4)
Equation (4) shows that the efficiency of a particular catalyst is partitioned between
its equilibrium and rate constants. Of these two constants, KONX shows the greatest
variation between catalysts, thus making it the critical rate-determining factor. As
such, knowledge of the equilibrium constant is crucial in predicting the effect of a
particular catalyst. Also, to determine the overall rate of nitrosation, one must first
establish the equilibrium constant of the particular catalyst used. In spite of this
requirement, equilibrium constants for a number of common nitrosating agents have
not been well established in the literature. Other equilibrium constants, such as those
for nitrosyl iodide and nitrosyl acetate, are minuscule and cannot be obtained using
conventional experimental techniques, and only their rough estimates are available.
The next section of this paper presents a derivation of a thermodynamic model to
predict values of KONX based on Edwards’ nucleophilic En parameter. This is
followed by a review of literature on the available experimental measurements of
KONX to justify a linear relationship between ln.KONX and En. Finally, the model is
applied so as to provide previously unobtainable equilibrium constants, which are
then used to facilitate a kinetic re-analysis of some existing rate data.
Theoretical
It is well appreciated that KONX increases with increasing nucleophilic strength, yet
there have been no prior attempts to develop a quantitative relationship. Many scales
have been proposed to quantify nucleophilic strength, with perhaps the more
commonly used being those of Pearson2 and Swain and Scott3. These scales are
kinetic in nature and are not expected to apply to the current system. A
thermodynamic measure of nucleophilicity is, however, provided by Edwards’
nucleophilic parameter4, En. While this parameter has been applied previously with
considerable success, it has been assigned little fundamental significance.
We propose that the thermodynamic potential of a nucleophile can be obtained from
the free energies of formation of both X- and X2. This idea is represented by equation
(5) with the corresponding free energy change of the reaction expressed by equation
(6).
ON + αnX- +
−
21 nα X2 ↔ ONX + αne- (5)
oooooON,X,nX,
nONX,5 22
1ffff GGGGG ∆−∆−∆
−
−∆=∆ −αα (6)
Here, αn is a parameter that measures the ON-X bond order, and lies somewhere
between zero (X-) and unity (X-X). We propose that for a wide range of catalysts αn
is approximately constant, and will argue this point further in discussion section. It
will be shown later in the current chapter that, this implies a linear relationship
between KONX and En.
We now introduce equation (7), which leads to the formulation of Edwards’ En
parameter. The standard free energy change for this reaction is expressed in equation
(8)
2X- ↔ X2 + 2e- (7)
ooo−∆−∆=∆ X,X,7 2
2 ff GGG (8)
Upon substitution of equation (8) into equation (6) and some rearrangement, we
obtain equation (9).
oooooON,57
nX,ONX, 2
1fff GGGGG ∆+∆+∆
−
+∆=∆ −
α (9)
Edwards noticed that nucleophilic strength was a function of the standard electrode
potential Eo, which is related to the free energy change of reaction (7) by equation
(10), where v is the number of electrons transferred in the reaction (in this case 2) and
F is the Faraday constant (96.485 C mol-1). Equation (10) may now be substituted
into equation (9), to produce equation (11).
ovFEGo −=∆ 7 (10)
( ) oooooON,5X,ONX, 1 fnff GGEFGG ∆+∆+−−∆=∆ − α (11)
The Edwards nucleophilic constant is related to the standard electrode potential by
equation (12), where the numerical constant of 2.6 corresponds to the electrode
potential of water. Nucleophilic En constants have been tabulated for most common
nucleophilic species, including those used to catalyse nitrosation4.
60.2n += oEE (12)
We may replace the standard electrode potential in equation (11) with Edwards’
nucleophilic constant, as illustrated by equation (13).
( ) ooooON,5nnnX,ONX, )1(60.21 fff GGFEFGG ∆+∆+−+−−∆=∆ − αα (13)
We now turn our attention to the formation reaction for ONX, as given by equation
(1). The standard free energy change for this reaction is specified by equation (14).
oooooo−+ ∆−∆−∆−∆+∆=∆ X,H,HONO,OH,ONX,1 2 fffff GGGGGG (14)
Replacing the standard free energies of formation of H2O(l) (-237.129 J mol-1),
HNO2(aq) (-55.649 J mol-1) and H+(aq) (0 J mol-1), we obtain equation (15); all
thermodynamic values invoked in this paper have been taken from reference 5 and
apply to aqueous solution.
480181X,ONX,1 −∆−∆=∆ −ooo
ff GGG (15)
Substituting (13) into (15) leads to:
( ) 480181)1(60.21 ON,5nnn1 −∆+∆+−+−−=∆ ooofGGFEFG αα (16)
Equation (16) may be further simplified to give equation (17), by substituting the free
energy of formation of ON(aq). This has been determined as 111.300 J mol-1 from the
gas phase value of 86.600 J/mol, along with the known Henry’s constant of 1.9×10-3
M atm-1 6.
( ) 18070)1(60.21 5nnn1 −∆+−+−−=∆ oo GFEFG αα (17)
Using the relationship between o1G∆ and KONX, equation (17) is transformed into
equation (18).
( ) 18070)1(60.21ln 5nnONX +∆−−−−= oGFEFKRT n αα (18)
With a minor rearrangement, we obtain equation (19) that takes the form of a linear
relationship between ln.KONX and En. As such, a plot of ln.KONX vs En should give a
straight line, where the slope may be used to calculate αn, and the intercept to obtain
oG5∆ .
( )RT
FGE
RTFK
)1(60.2180701ln n5n
nONX
αα −−∆−+
−
=o
(19)
Review of Experimental Measurements
A comprehensive review of literature values of KONX has been undertaken to identify
the experimental measurements required to calculate the parameters of the model.
Note that all equilibrium constants quoted in the current manuscript invoke the
standard state of a hypothetical ideal solution of unit molality, and, unless otherwise
stated, are at 25oC.
Nitrosyl Chloride
The equilibrium constant for formation of nitrosyl chloride has been experimentally
studied by Schmid and Hallaba7, who determined it to be 1.1×10-3 M-2. Turney and
Wright8 used thermodynamic cycles in determining the equilibrium constant to be
6×10-4 M-2, somewhat smaller than the experimentally determined value. A
thermodynamic estimate may also be made from the free energy change of the
reaction (16.9 J/mol) as calculated from standard free energies of formation. This
gives a thermodynamic equilibrium constant of 1.1×10-3 M-2, agreeing with the
experimentally derived value.
Nitrosyl Bromide
Schmid and Fouad9 experimentally determined the equilibrium constant for nitrosyl
bromide as 5.1×10-2 M-2. Turney and Wright8 have again assigned the equilibrium
constant a thermodynamic value of 9×10-2 M-2. Both the experimental and calculated
values are seen to exhibit reasonable agreement.
Nitrosyl Thiocyanate
The earliest, and the most often cited, determination of the equilibrium constant for
nitrosyl thiocyanate formation is that of Stedman and Whincup10 who conducted tests
in perchloric-acid media over a range of ionic strengths. Using an estimate of the
activity coefficient, the equilibrium constant was evaluated as 32 M-2 at 20oC. The
enthalpy change for the reaction was also determined to be -12 kJ/mol, allowing us to
adjust the equilibrium constant to 29 M-2 at 25oC. Further experiments were
conducted in sulphuric-acid media, where the equilibrium constant was determined to
be 18.5 M-2 at 20oC (i. e. 17.0 M-2 at 25oC), though it was reported that this value was
unduly influenced by solution activity. Jones et al.11 determined an average value of
9.9 M-2 for nitrosyl thiocyanate over a range of perchloric acid concentrations,
contrasting somewhat with the values of Stedman and Whincup10. We used the free
energies of aqueous thiocyanate and nitrosyl thiocyanate formation to determine the
equilibrium constant as 30.8 M-2, agreeing well with the widely accepted value of
Stedman and Whincup10.
S-Nitrosothiourea
While thiourea yields one of the more efficient nitrosating agents, the equilibrium
constant for its formation remains poorly resolved. Al-Mallah et al.12 determined it as
5.000 M-2, at an ionic strength of 0.3. Adjusting for activity coefficients using the
Debye-Hückel approximation, we obtain a corrected value of 9.500 M-2.
Nitrosyl Thiosulphate Ion
Garley and Stedman13 conducted experiments into the formation of the nitrosyl
thiosulphate ion in weakly acidic solutions, and obtained an average equilibrium
constant of 1.66×107 M-2. Again, we have adjusted this result to account for activity
coefficients using the Debye-Hückel approximation, yielding a thermodynamic
equilibrium constant of 1.82×107 M-2. Abia et al.14 also determined the equilibrium
constant for the formation of nitrosyl thiosulphate, obtaining an average value of
2.5×106 M-2. Once adjusted for solution activity, this gives a value of 4.3×106 M-2,
being somewhat lower than that of Garley and Stedman13.
Dinitrogen Trioxide
During nitrosation in the absence of a nucleophilic catalyst, reaction proceeds through
the nitrosating agent dinitrogen trioxide. The formation reaction for dinitrogen
trioxide is generally written as in equation (20).
2HNO2 ↔ N2O3 + H2O (20)
The equilibrium constant for equation (20) has been well studied both from
thermodynamic15,16 and experimental17,18 standpoints, returning respective KONX
values of 2.3×10-3, 9.0×10-3, 3.0×10-3 and 7.5×10-3 M-2. However, these studies have
expressed the equilibrium constant in the form of equation (20), rather then in the
form of equation (21), with nitrous acid reacting with a nitrite ion.
HNO2 + NO2- + H+ ↔ N2O3 + H2O (21)
We now see that equation (21) resembles equation (1), where the nitrite ion is acting
as the nucleophile. To adjust the free energy change of the reaction (and subsequently
the equilibrium constant) from the form of equation (20) to that of equation (21), we
simply replace the free energy of formation of nitrous acid (-55.600 J/mol) with that
of a nitrite ion (-37.200 J/mol). This yields thermodynamic equilibrium constants of
3.8, 15.1, 5.0 and 12.6 M-2 for the four studies mentioned above. We may also
determine the equilibrium constant to be 4.9 M-2 from the free energies of formation
of the reacting species.
Discussion
The equilibrium constants reviewed in the previous chapter are plotted in figure 1 as a
function of En, together with the equilibrium constant for OH- acting as the
nucleophile, which is, by definition, unity. The linear relationship of equation (19) is
also drawn in figure 1, in the least-square sense, to produce the linear relationship of
equation (22). The comparison of equations (19) and (22) yields αn = 0.46 and oG5∆
= 8.906 J/mol, with αn indicating that the ON-X bond order is intermediate between
that of X2 and X-.
Figure 1. Comparison of the model with the experimental measurements.
35.2988.17ln nONX −= EK (22)
As both of the parameters plotted in figure 1 are related to thermodynamic properties,
our relationship may be considered a linear free energy relationship (LFER). A LFER
of this form generally indicates that a single mechanism dominates interaction within
Nitrosyl Thiocyanate
Nitrous Acid
En
1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
lnK O
NX
-10
-5
0
5
10
15
20
Cl-
Br-
OH-
NO2-
SCN-
SC(NH2)2
S2O32-
the reaction series. Interpreting figure 1 in this context reveals that bond order is the
dominant factor in determining the magnitude of KONX, and that other interaction
mechanisms do not play a significant role. This is particularly interesting when one
considers the range of species to which the relationship has been applied, including
both the nitrosyl halides and the sulphur-centred nitrosating agents. This range also
contains a number of larger molecules(e. g. thiourea and thiosulphate), species
possessing two possible nucleophilic sites (such as the S and N atoms in thiocyanate
and thiourea) and those with negative, neutral and positive charges. In spite of this
wide range of species, no secondary effects such as those attributable to steric or
resonance factors seem to significantly influence nucleophilicity.
Table 1 shows the equilibrium constants calculated from the model, along with
literature values for each catalyst. As expected, all of the equilibrium constants used
to calculate the parameters of the model are well predicted, usually within the ten-fold
variation witnessed in the experimental measurements. Included in table 1 are
estimates for the equilibrium constants of nitrosyl fluoride, nitrosyl nitrate and
nitrosyl sulphate formation; three compounds which have previously been found to
have little or no catalytic effect on nitrosation22,23. The relatively low equilibrium
constants predicted for nitrosyl fluoride and nitrosyl sulphate disclose the reason why
these agents do not catalyse nitrosation. The literature value for the formation of
nitrosyl nitrate, which is six orders of magnitude greater than that predicted here, is
clearly in error, as the nucleophilic catalysis of nitrosation by the nitrate ion cannot be
observed in experiments as a consequence of the low equilibrium constant.
Table 1. Comparison of literature and predicted equilibrium constants for common
nitrosating agents. Calculations made using En values of reference 4, except that of
ONF, which is taken from reference 19.
ONX KONX (literature), M-2 KONX (model), M-2 Reference(s)ONF - 1.5×10-15 -ONNO3 4×10-5 3.3×10-11 5ONSO4
- - 7.0×10-11 -CH3COONO ≈1.4×10-8 4.4×10-6 20ONCl 1.1×10-3; 6×10-4; 1.1×10-3 7.8×10-4 4; 5; This StudyONBr 5.1×10-2; 9×10-2 9.7×10-2 7; 5ONNO2 3.8; 15; 5.0; 13; 4.9 5.0 13; 14; 15; 16; This StudyONSCN 29; 17; 9.9; 31 30 8; 8; 9; This StudyONI >100 1.800 21(NH2)2CSNO+ 9.500 1.5×104 10S2O3NO- 1.8×107; 4.3×106 6.7×106 11; 12
Dix and Williams24 studied the iodide catalysed nitrosation of various aniline
derivatives, but were unable to present rate constants due to the absence of an
equilibrium constant for nitrosyl iodide. These rate constants have now been
estimated (using the equilibrium constant from table 1), and are presented in table 2.
Figure 2 shows how the rate constants for nitrosation via nitrosyl iodide vary as a
function of pKa. For comparison, we also included the published results obtained for
nitrosation via nitrosyl chloride and nitrosyl thiocyanate. It can readily be seen that,
nitrosyl iodide is the least reactive nitrosating agent, and that the reaction rates do not
approach the encounter-controlled limit (as they do for nitrosyl chloride). In other
words, nitrosation via nitrosyl iodide is controlled by chemical kinetics, rather than by
diffusion.
Table 2. Rate constants for the nitrosation of various aniline derivatives by nitrosyl
iodide.
kN, M-1 s-1
p-Methoxyaniline 2.6×107
p-Chloroaniline 2.9×106
Aniline 2.6×106
m-Methoxyaniline 1.9×106
p-Carboxyaniline 8.8×104
Figure 2. ln.kN versus pKa for the nitrosation of various aniline derivatives by ONI
(•), ONCl 24 (◊) and ONSCN 22 (□). Dashed line indicates the encounter-controlled
limit.
pKa
2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0
lnk n
4
5
6
7
8
9
10
Nitrosyl acetate is another compound for which a KONX value has not been
experimentally obtainable due to its low value. The only available value is an
estimate that has been inferred from the nitrosation of several secondary amines20.
Nitrosyl acetate has long been implicated in nitrosation reactions, and there exist
several studies on its role, many with considerable disagreement. A complex
mechanism has been proposed and validated20,27,28 in which nitrosyl acetate not only
acts as a nitrosating agent, but also catalyses the formation of dinitrogen trioxide.
Table 3 shows calculated rate constants for nitrosation via nitrosyl acetate, extracted
from the data of Casado et al.20 Included in the table, for comparative purposes, are
published data26 for nitrosation via dinitrogen trioxide. While the rate constants for
dinitrogen trioxide are all relatively constant at ca. 10-8 M-1 s-1, those for nitrosyl
acetate are considerably lower, and vary by approximately three orders of magnitude.
It appears as though nitrosation via dinitrogen trioxide proceeds at the encounter-
controlled limit, thus indicating that a similar limitation should operate for nitrosyl
acetate, due to its greater reactivity. The fact that the experimental results do not
support this observation seems to indicate either erroneous experimental data
(possibly due to the intricate kinetic analysis brought about by the complex reaction
mechanism) or incorrect mechanistic assumptions.
Table 3. Rate constants for the nitrosation of three secondary amines by nitrosyl
acetate, as compared to nitrosation via dinitrogen trioxide.
kN (CH3COONO), M-1 s-1 kN (N2O3), M-1 s-1
Morpholine 2.8×106 2.2×108
N-Methylaniline 5.6×105 4.0×108
Piperazine 3.5×104 1.3×108
Acknowledgements
We would like to thank Orica Pty. Ltd. for project funding, and the Australian
Research Council for the provision of a student scholarship to G. D.
References
1 J. H. Ridd, Quart. Revs., 1961, 15, 418.
2 R. G. Pearson, H. Sobel and J. Songstad, J. Am. Chem. Soc., 1968, 90, 319.
3 C. G. Swain and C. B. Scott, J. Am. Chem. Soc., 1953, 75, 141.
4 J. O. Edwards, J. Am. Chem. Soc., 1954, 76, 1540.
5 D. D. Wagman, W. H. Evans, V. B. Parker, I. Halow, S. M. Bailey and R. H.
Schumm, NBS Technical Note 270-3, National Bureau of Standards,
Washington, D. C., 1968.
6 J. L. Durham, J. H. Overton and V. P. Aneja, Atmos. Environ., 1981, 15(6),
1059.
7 H. Schmid and E. Hallaba, Monatsh. Chem., 1956, 87, 560.
8 T. A. Turney and G. A. Wright, Chem. Rev., 1959, 59, 497.
9 H. Schmid and M. G. Fouad, Monatsh. Chem., 1957, 88, 631.
10 G. Stedman and P. A. E. Whincup, J. Chem. Soc., 1963, 5796.
11 E. Jones, C. G. Munkley, E. D. Phillips and G. Stedman, J. Chem Soc., Dalton
Trans., 1996, 1915.
12 K. Al-Mallah, P. Collings and G. Stedman, J. Chem. Soc., Dalton Trans.,
1974, 2469.
13 M. S. Garley and G. Stedman, J. Inorg. Nucl. Chem., 1981, 43, 2863.
14 L. Abia, A. Castro, E. Iglesias, J. Leis and M. Pena, J. Chem. Res. (S), 1989,
106.
15 V. M. Gratzel, S. Taniguchi and A. Henglein, Ber. Bunse-Ges. Phys. Chem.,
1970, 74(5), 488.
16 T. A. Turney and G. A. Wright, J. Chem. Soc., 1958, 2415.
17 G. Y. Markovits, S. E. Schwartz and L. Newman, Inorg. Chem., 1981, 20,
445.
18 G. Stedman, Adv. Inorg. Chem. Rad., 1979, 22, 113.
19 W. M. Latimer, The Oxidation States of the Elements and Their Potentials in
Aqueous Solutions, 2 ed., Prentice-Hall: New York, 1952.
20 J. Casado, A. Castro, M. Mosquera, M. F. R. Prieto and J. V. Tato, Monatsh.
Chem., 1984, 115, 669.
21 L. Dozsa, I. Szilassy and M. T. Beck, Inorg. Chim. Acta, 1976, 17, 147.
22 E. D. Hughes and J. H. Ridd, J. Chem. Soc., 1958, 88, 82.
23 B. L. Wedzicha and W. Tian, Food Chem., 1989, 31, 189.
24 L. R. Dix and D. L. H. Williams, J. Chem. Res. (S), 1984, 96.
25 J. H. Ridd, Adv. Phys. Org. Chem., 1978, 16, 1.
26 M. R. Crampton, J. T. Thompson and D. L. H. Williams, J. Chem. Soc. Perkin
Trans. 2, 1979, 18.
27 G. Stedman, J. Chem. Soc., 1960, 1702.
28 J. Casado, A. Castro, J. R. Leis, M. A. Lopez Quintela and M. Mosquera,
Monatsh. Chem., 1983, 114, 639.