KINETICS AND MECHANISM OF ACID CATALYZED DECOMPOSITION OF 1-PHENYL-3,3-DIALKYLTRIAZENES

Article abstract of DOI:10.1135/cccc19940401

Five model 1-phenyl-3,3-dialkyltriazenes (methyl, ethyl, 2-propyl, butyl, cyclohexyl) have been synthesized and their acid-catalyzed decomposition kinetics have been investigated spetrophotometrically in aqueous ethanol (40 vol.percent) with pivalic acid as the catalyst.The results show that the rate-determining step is catalyzed by the proton.The decrease in the observed rate constant at higher concentrations of pivalic acid is explained by the formation of an unreactive complex of the nondissociated acid and respective triazene.The steric effect of alkyl groups on the catalytic rat constants is discussed.

Full text of DOI:10.1135/cccc19940401

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
401
KINETICS AND MECHANISM OF ACID CATALYZED DECOMPOSITION
OF 1-PHENYL-3,3-DIALKYLTRIAZENES
Miroslav LUDWIG, Pavla VALASKOVA and Oldrich PYTELA
Department of Organic Chemistry,
University of Chemical Technology, 532 10 Pardubice, The Czech Republic
Received March 29, 1993
Accepted April 29, 1993
Five model 1-phenyl-3,3-dialkyltriazenes (methyl, ethyl, 2-propyl, butyl, cyclohexyl) have been syn-
thesized and their acid-catalyzed decomposition kinetics have been investigated spectrophotometri-
cally in aqueous ethanol (40 vol.%) with pivalic acid as the catalyst. The results show that the
rate-determining step is catalyzed by the proton. The decrease in the observed rate constant at higher
concentrations of pivalic acid is explained by the formation of an unreactive complex of the nondis-
sociated acid and respective triazene. The steric effect of alkyl groups on the catalytic rate constants
is discussed.
1
2
3
Triazenes of general formula I where R is most frequently an aryl, R is alkyl, and R
is alkyl, aryl, acyl, halogen, hydroxyl, or hydrogen, are decomposed at the conditions
1
– 21
of acid catalysis in various solvents
. The decomposition products involve the dia-
3
zonium salts and corresponding complementary compounds having a bond N −H in-
3
2
stead of N −N . From the point of view of mechanism of acid-catalyzed decomposition,
the triazene molecule is interesting by the fact that, in principle, each of the three
nitrogen atoms can be protonated. The quantum-chemical calculations indicate either
1
3
22
1
the same probability of protonation at N and N (ref. ) or preferable protonation at N
2
3
3
(
ref. ). The triazene chain is split only in the case of the protonation at the N atom,
whereas the protonations of the other nitrogen atoms mostly represent side equilibria .
1
9
1
1
2
3
2
R −N =N −N −R
R3
I
As for the catalysis type, this can be either the specific acid catalysis involving a pre-
equilibrium – the protonation of substrate – and subsequent decomposition to products
3
or the general acid catalysis involving simultaneous formation of N −H bond and split-
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

4
02
Ludwig, Valaskova, Pytela:
3
2
ting of N −N bond. In literature there is no unequivocal conformity in the view on the
mechanism of acid-catalyzed decomposition of triazenes, the type of mechanism ob-
1
2
3
viously also depends on the character of substituents R , R , R (formula I). Some
papers, especially the older ones, prefer the specific catalysis going by the monomole-
1
– 5,12,14
cular mechanism A1 (refs
). On the other hand, a number of results observed
1
1
indicate the bimolecular mechanism of A-S 2 type with general acid catalysis : beside
E
the criterion of isokinetic temperature also the salt effects in the decomposition of
3
-acetyl-1,3-diphenyltriazenes and 3-(N-methylcarbamoyl)-1,3-diphenyltriazenes
1
0
2
3
(
ref. ) indicating the existence of ion pairs, hence an easy splitting of the N −N bond.
The mathematical modelling of decomposition mechanism of 3-methyl-1,3-diphenyl-
triazenes in buffers formed by general acids indicates the general catalysis by A-S 2
E
1
7
mechanism involving the proton-catalyzed splitting . The investigation of substituent
effects in the latter substrate types did not bring unequivocal results with respect to the
reaction course . Investigations of the effect of medium on decomposition of 1,3-
bis(4-methylphenyl)triazene catalyzed with trichloroacetic acid in amphiprotic and
1
6
1
9 – 21
aprotic media
revealed, in various types of solvents, three types of the dependence
between the observed rate constant and concentration of the catalyst trichloroacetic
acid. The fastest reactions were those with the proton, slower ones with undissociated
acid, and the slowest with the acid dimer. The reaction rate in the individual media
depended on the predominant form of the catalyst acid in the given medium. The reac-
tions with the undissociated acid, the proton, as well as that with the acid dimer pro-
ceed in this case by the mechanism of general acid catalysis.
The aim of the present work is to study the effect of magnitude of alkyl groups
3
bound to N atom upon the kinetics of acid catalyzed decomposition of 1-phenyl-3,3-
1
2
3
dialkyltriazenes I (R = phenyl, R = R = methyl, ethyl, isopropyl, butyl, cyclohexyl) in
the medium of 40 vol.% aqueous ethanol and to verify the mechanism of general acid
catalysis at these conditions. For the catalyst acid we chose the relatively bulky pivalic
acid.
EXPERIMENTAL
Spectral measurements. The electronic spectra were measured in 40 vol.% aqueous ethanol using
a Specord M-40 apparatus. The mass spectra were measured with a Kratos MS 25RSA apparatus
1
13
using the EI technique in the GC/MS arrangement. The H NMR and C NMR spectra of the sub-
stances were measured in deuteriochloroform using a JEOL JNM-FX 100 or a Bruker AM-400 ap-
paratus at 300 and 330 K, respectively.
1
-Phenyl-3,3-dialkyltriazenes were prepared by the reaction of benzenediazonium chloride with
2
4
the corresponding secondary amine . A solution of 0.3 mol amine in 250 ml water with ice (the
preparation procedure A) was treated with 75 g (ca 0.8 mol) sodium carbonate and the suspension
obtained was then treated with cold solution of 0.3 mol benzenediazonium chloride added during
2
0 min. The diazonium solution was obtained by diazotizing 28 g (0.3 mol) aniline with 21 g
(
0.3 mol) sodium nitrite in 200 ml 20% hydrochloric acid. The reaction mixture was stirred for an-
other 30 min at 5 – 10 °C. In the cases of water insoluble amines (the preparation procedure B), the
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
403
solution of benzenediazonium chloride (0.3 mol) was mixed with a suspension or solution of hydro-
chloride of the corresponding amine (0.3 mol). The mixture obtained was treated with a cold solution
of 90 g sodium carbonate (ca 1 mol) in 250 ml water added with constant stirring, and the reaction
suspension was then stirred for another 30 min. The products were purified by chromatography on
alumina column with ethyl acetate–chloroform 10 : 1 as the mobile phase. The preparation procedure
of the individual model triazenes, yields, and elemental analyses are given in Table I.
Determination of concentration and pKHA of pivalic acid. The concentration of catalyst acid and
its pKHA in 40 vol.% aqueous ethanol were determined by potentiometric titration using an automatic
−3
titrator Radiometer RTS 622 and methanolic tetrabutylammonium hydroxide (0.1 mol dm ) as the
2
5
titrant. The electrodes used for the measurements were identical with those in our earlier works .
2
6
Benzoic acid was chosen for the standard (pKHA 5.76 in 40% ethanol ). The solutions with the same
concentrations of proton at different concentrations of pivalic acid (used in the case of dicyclohexyl
derivative) were prepared by titrating the solutions of acid with a 20% solution of NaOH in 40%
ethanol to the pH (mV) value of the least concentrated solution used.
Kinetic measurements. The solution of pivalic acid of definite concentration was thermostatted in
a quartz cell at 25 ± 0.1 °C for 10 min. Then, using a Hamilton syringe, a solution of respective
−3
−3
triazene in dioxane (6 µl ca 5 . 10 mol dm ) was added into the cell and the solution was mixed.
The concentration decrease of starting triazene was monitored at the wavelength of the longestwave
maximum (Table II) during a period of three halflives.
Treatment of results. The kinetic curves were evaluated by means of our own optimization pro-
2
7
gram . The statistical treatment of the results by nonlinear regression, principal components ana-
lysis, and linear regression was carried with the use of our own programs on a PC-AT.
TABLE I
Preparation methods, yields, and elemental analyses of synthesized 1-phenyl-3,3-dialkyltriazenes
C H N=N−NR
2
6
5
Calculated/Found
% H
Formula
M.w.)
Method
Yield, %
R
M.p., °C
(
%
C
% N
Methyl (Me)
Ethyl (Et)
C8H11N3
A
64.75
64.43
7.65
7.38
27.91
28.18
oil
oil
(
149.2)
60
C10H15N3
A
67.63
67.79
8.69
8.47
23.64
23.73
(
177.2)
73
Isopropyl (i-Pr)
Butyl (Bu)
C12H29N3
A
69.93
70.24
9.23
9.27
20.22
20.49
39 – 40
oil
(
215.4)
48
C14H23N3
A
71.81
72.10
9.98
9.87
18.22
18.03
(
233.4)
Cyclohexyl (cHx) C18H27N3
285.4)
74
B
75.41
75.79
9.83
9.47
14.67
14.73
133 – 134
(
27
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

4
04
Ludwig, Valaskova, Pytela:
RESULTS AND DISCUSSION
Spectral data (UV-VIS, MS, H NMR, 13C NMR) of the 1-phenyl-3,3-dialkyltriazenes
prepared are presented in Table II. The mass spectra showed a similar fragmentation in
all the model substrates. In all the cases the most intensive fragment is C H (m/z = 77),
1
6
3
5
2
most spectra exhibit the fragment C H N (m/z = 105) formed by N −N bond rupture.
6
5
2
1
2
This bond splitting also produces the fragments R R N (m/z = 44, 72, 100, 128, 180)
which further decompose to give the RNH fragments (m/z = 30, 44, 58, 72, 98) and R
fragments (m/z = 15, 29, 43, 57, 83). The NMR spectra proved the presumed structure
1
of triazenes. In addition, the H NMR spectra of the dimethyl and dibutyl derivatives
showed the equivalence of both alkyl groups with regard to chemical shifts, which
2
3
corresponds to rapid rotation around the N −N bond. On the other hand, the CH signals
of the isopropyl derivative are broadened at 300 K and those of the cyclohexyl deriva-
tive are very broad. At 330 K they become more narrow, which confirms hindered
TABLE II
Spectral characteristics (UV-VIS, MS, NMR) of synthesized 1-phenyl-3,3-dialkyltriazenes
C H N=N−NR
6
5
2
R
λmax, nm
m/z, %
δ, ppm (T = 300 K)
1
Me
Et
310
312
318
44 (8), 51 (32), 77 (100), 78 (40),
H arom: 7.51 – 7.00 m, 5 H
aliph: 3.35, 6 H
1
05 (90), 120 (8), 149 (63)
72 (8), 77 (100), 78 (12), 105 (42),
1
77 (15)
1
i-Pr
43 (32), 51 (20), 58 (44), 77 (100), H arom: 7.47 – 7.00 m, 5 H
7
1
8 (25), 80 (12), 100(40), 105 (63),
20 (8), 205 (25)
aliph: 4.58, 2 H (2 CH), 1.29, 12 H
(4 CH3, J(H,H) = 6.8)
C arom: 151.7 s, 128.7 d, 124.7 s
3
1
3
1
20.3 s
aliph: 77.0 m
1
Bu
313
317
51 (12), 57 (16), 77 (100), 78 (40), H arom: 7.42 – 7.00 m, 5 H
8
1
4 (14), 86 (26), 93 (11), 105 (92),
20 (8), 128 (28), 148 (8), 233 (39)
aliph: 3.68, 4 H (2 CH2), 1.65, 4 H
(2 CH2), 1.10, 4 H (CH2), 0.93,
1
2 H (2 CH3)
1
cHx
55 (18), 77 (100), 83 (18), 83 (18), H arom: 7.46 – 7.00 m, 5 H
9
8 (25), 138 (15), 180 (38), 285
aliph: 4.1 2 H (2 CH), 1.90 – 1.00,
20 H (10 CH2)
C arom: 151.7 s, 128.6 d,
(
12)
1
3
1
2
24.5 s, 120.3 s alif: 77.0 m,
5.8 d
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
405
rotation with bulky substituents. These conclusions are also confirmed by the signals in
1
3
C NMR spectra of isopropyl and cyclohexyl derivatives at 300 and 330 K.
Table III gives the observed rate constants kobs of the acid-catalyzed decomposition
of the model triazenes in 40 vol.% aqueous ethanol at 25 °C. The experimental values
obtained exhibit diminishing increases in the observed rate constant with increasing
concentration of pivalic acid, and, beginning from the acid concentration of 0.34 – 0.56
−
3
mol dm (depending on the substituents of substrate) kobs even decreases. The facts
mentioned can be seen in Fig. 1 (full circles).
From the survey in the introduction of this paper it follows that the decomposition of
triazene can take place by several basic mechanisms. In the first one the triazene (T) is
split into the products (P) by the A-S 2 mechanism in a reaction of the triazene with
E
TABLE III
−1
Observed rate constants (kobs, s ) of decomposition of 1-phenyl-3,3-dialkyltriazenes C H N=N−NR
2
6
5
in 40 vol.% aqueous ethanol at 25 °C at various concentrations of catalyst acid
3
−1
kobs . 10 , s
3
c . 10
mol dm
−3
Me
Et
i-Pr
Bu
cHx
0
0
.0837
.551
0.795
5.868
1.042
5.583
0.736
3.575
0.704
3.153
0.635
2.458
2
.451
3.302
.290
3.194
.378
0
0
.796
.923
7.742
8.606
8.254
8.567
13.39
4.130
6.471
3.840
3.986
3
3
1
1
2
3
.69
.97
.46
.32
12.64
15.21
14.35
18.34
7.571
9.376
9.853
9.873
5.093
5.299
5.380
5.531
4.844
5.283
15.54
18.09
5.782
5
.704
3
5
.56
.01
19.09
18.72
17.60
18.14
10.60
11.16
5.508
5.015
5.571
5.318
5
.297
5
5
.56
.57
17.52
17.14
11.57
10.86
4.750
4.639
5.234
4.876
18.44
17.82
5
.130
6
.74
17.26
10.97
4.635
4.841
4.785
17.85
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

4
06
Ludwig, Valaskova, Pytela:
+
19 – 21
either the proton (H ) or the undissociated acid (HA)
, Eqs (A), (B). The existence
1
9 – 21
of dimer of the catalyst acid is not likely in the medium studied by us
.
kH
T + H+
P
(A)
(B)
kHA
T + HA
P
In the mechanism describing the specific acid catalysis, the protonated triazene is
1
4
formed and then decomposed into products in the rate-determining step (C).
KTH
k′H
T + H+
TH+
P
(C)
However, the schemes given cannot interpret the decrease in kobs at higher concentra-
tions of the catalyst acid. This decrease is probably caused by the decrease of concen-
tration of reactive triazene in the reaction mixture due to formation of some less
reactive complex involving the triazene molecule, e.g. a 1 : 1 (or 1 : 2) associate of the
2
3
triazene with the undissociated acid . Such complexes can be formed in fast pre-
equilibria followed by very slow processes producing the products P (schemes (D), (E)).
KTK1
k′HA
T + HA
TK1
P
(D)
(E)
KTK2
k′′HA
TK1 + HA
TK2
P
The associates of TK1 or TK2 types can be formed unless the catalyst acid is strong
enough to transfer the proton to the substrate. The existence of the complexes of
triazene with general acid was proved by measuring the observed rate constants of the
decomposition of dicyclohexyl derivative at a constant proton concentration (c
+
=
H
−
4
−3
1
0
.05 . 10 mol dm ) and variable concentration of undissociated acid (for c = 0.061,
HA
−3
−3
−3
−3 −1
.305, and 0.610 mol dm it was kobs = 3.038 . 10 , 2.458 . 10 , and 1.584 . 10 s ,
respectively). The observed rate constants decrease with increasing concentration of
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
407
pivalic acid and exhibit nearly the same decreasing trend as the kobs values at higher
concentrations of catalyst acid in the case of this substrate.
The analytical concentrations of acid and triazene in (A) – (E) and the corresponding
equilibrium constants can then be expressed by Eqs (1) – (6).
+
+
cHA = [H ] + [HA] + [TK1] + [TK2] + [TH ]
cT = [T] + [TK1] + [TK2] + [TH+]
KTK1 = [TK1]/[HA][T]
(1)
(2)
(3)
(4)
(5)
(6)
2
KTK2 = [TK2]/[HA][TK1] = [TK2]/KTK1[HA] [T]
+
+
KTH = [TH ]/[T][H ]
+
−
+ 2
KHA
= [H ][A ]/[HA] = [H ] /[HA]
Using Eqs (1) – (5) and the equation for the reaction rate one can suggest a summary
relationship (7) for the dependence of observed rate constant upon the concentrations of
catalyst pivalic acid and proton. When deriving this relationship we neglected, on the
basis of the dissociation constant determined by us for pivalic acid in 40% ethanol
−
7
(
KHA = 1.8 . 10 ) and on the basis of the ratio of triazene to catalyst acid in the reac-
FIG. 1
−1
Observed rate constants (kobs, s ) of acid-cata-
lyzed decomposition of 3,3-dicyclohexyl-1-
phenyltriazene in 40 vol.% aqueous ethanol
(
empty circles correspond to samples titrated to
−4
−3
proton concentration cH+ = 1.05 . 10 mol dm ;
see Experimental). Curve represents depend-
ence of kobs vs catalyst acid concentration cal-
culated according to Eq. (10)
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

4
08
Ludwig, Valaskova, Pytela:
tion medium, the concentrations of all the particles at the right-hand side of Eq. (1)
except for [HA].
′
′
2
′ ′
+ (k KTK1 + kHA)cHA + (k K + k )[H ]
HA HA H TH H
+
k
KTK2 KTK1
c
HA
kobs
=
(7)
2
+ KTK1 cHA + KTH[H+]
1
+ KTK2 KTK1
c
HA
Using Eq. (6) and experimental dissociation constant of pivalic acid in the given me-
dium (pKHA = 6.74) one can write Eq. (8) for the proton concentration.
+
H ] = 0.000424 c H1 ⁄A2
[
(8)
The equation (7) obtained is relatively complex but provides the possibility of arrang-
ing a series of models with various powers of the concentration of catalyst acid both in
the numerator and in the denominator. The various models were tested with the use of
the data for the dicyclohexyl derivative which was measured in detail. For this deriva-
tive we also measured the decomposition kobs values at the same proton concentration
but various concentrations of the catalyst acid. The decrease in the observed rate con-
stant with increasing concentration of the catalyst acid at the same proton concentration
can be explained by the reaction course according to (D) and (E); in addition to it, the
decrease indicates that in the final relationship there must be, side by side, a term
containing the proton concentration and a term containing the acid concentration.
The decrease in kobs at higher acid concentrations can be caused by the existence of
the complexes TK1 or TK2 only if the complexes do not react further or react much
′
′′
more slowly than the other particles (i.e. k
= k
= 0). As a consequence, the reac-
HA
HA
tion according to scheme (B) will not take place either, hence it is kHA = 0. On the basis
of these facts, Eq. (7) can be simplified to Eq. (9).
+
2
+
kobs
=
k [H ]/(1 + K
KTK1 c + KTK1 cHA + KTH[H ])
HA
(9)
H
TK2
−4
The solution of Eq. (9) by nonlinear regression (dicyclohexyl derivative: s = 2.03 . 10 )
gave the four parameters, out of which the KTK1 parameter was statistically insignifi-
cant (t = 1.68, tcrit = 2.10) and the KTH parameter was negative (without physical
meaning) and, in addition to it, had the lowest t test value (t = 0.18). These results
excluded the term containing the KTH parameter, hence the mechanism of specific acid
catalysis described in scheme (C) is not operating.
With regard to the lowered dimension of problem, the following calculations were
carried for all the derivatives studied. In all the cases, the term involving the KTK1
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
409
parameter turned out statistically insignificant (the results for the dicyclohexyl deriva-
−
4
tive are given as an example: s = 2.21 . 10 , t(KTK1) = 0.0). The statistical insignific-
ance of the term involving the KTK1 parameter indicates its negligible magnitude as
compared with the other terms in the denominator as well as the validity of the in-
equality KTK2 > KTK1. The resulting model can be expressed by Eq. (10) for the dicy-
clohexyl derivative (because Eq. (8) does not apply to solutions with the same proton
concentration and different acid concentrations) or by simplified Eq. (11) for the other
substrates:
+
2
kobs = k [H ]/(1 + K c ),
(10)
(11)
H
K
HA
kobs
=
0.000424 kH c1 ⁄ 2/(1 + KK c2 ),
HA HA
where K = K
KTK2. The k , K values obtained for the dicyclohexyl derivative are
H K
K
TK1
identical whether or not the points obtained with the same proton and different acid
concentrations have been involved in the calculations.
Table IV gives the calculated values of catalytic rate constant k and the stability
H
constants K of the complex. Figure 1 presents the interpretation of the experimental
K
dependence by a curve according to the model (10) for the dicyclohexyl derivative. The
points denoted by empty circles represent the kobs values for the titrated samples with
the same proton concentration (see Experimental), the other points correspond to the
values of Table III. Figure 1 represents only one section of the three-dimensional de-
+
pendence of kobs upon cHA and [H ].
The catalytic rate constant k decreases with increasing size of alkyl, which unequi-
H
3
vocally confirms the effect of substituents at N atom upon the rate of acid catalyzed
decomposition of 1-phenyl-3,3-dialkyltriazenes. The effect observed can be explained
TABLE IV
Optimized values of parameters in Eq. (11), their standard deviations and residual standard deviations s
a
b
3
R
kH
s(k )
KK
s(KK)
s . 10
n
H
Me
Et
77.35
75.90
45.84
31.47
28.17
3.87
1.02
1.12
0.947
3.41
2.30
0.248
0.249
0.226
0.223
0.157
1.38
1.17
13
12
13
13
22
3.43
i-Pr
Bu
1.93
0.735
0.208
0.235
0.693
0.597
c
cHx
a
3
−1 −1
b
6
−2 c
k , dm mol s ; K , dm mol ; according to Eq. (10).
H
K
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

4
10
Ludwig, Valaskova, Pytela:
3
by hindered approach of solvated proton to the N atom, which supports the statement
of bimolecular reaction mechanism with general acid catalysis. The substituents at the
3
N atom could, however, simultaneously deform the triazene skeleton and thereby – on
the other hand – increase its reactivity. In the final effect the steric effect of substituents
can be a combination of both the influences discussed. In addition to it, at higher con-
centrations of acid the observed rate constant is diminished by the formation of the
unreactive complexes TK1 and TK2. The combined stability constant K of these com-
K
plexes has a higher value with the triazenes containing bulkier alkyl groups (butyl,
cyclohexyl). Although the differences are not too high, the magnitude of stability con-
stant of complex seems to be connected with the size of alkyl groups. The derivatives
with bulkier alkyls probably undergo a deformation of triazene skeleton connected with
lowering of conjugation between the free electron pairs at nitrogen atoms and the aro-
matic system of phenyl group. The thereby increased basicity of nitrogen atoms can
facilitate the formation of the said complexes. Obviously, the hindered rotation (found
with the dicyclohexyl and diisopropyl derivatives) has no effect on this process. The
whole process is undoubtedly complicated by the participation of solvating molecules
of water and ethanol. The unreactive complex TK not only makes the attack by the
proton difficult, but the solvation of triazenes with water and ethanol molecules is made
difficult too. Hence the steric effect is manifested in both the process (A) and, probably,
the equilibrium (E), which indicates a relatively complex reaction course involving
several equilibria not only between the reacting components but also with the participa-
tion of molecules of both solvents.
The kinetic data of Table III were also treated by the method of principal compo-
nents28 (PCA). The source matrix was arranged in such a way that the rows and col-
umns corresponded to the individual substrates and the catalyst acid concentrations,
respectively. This matrix was standardized before the calculation. The first three col-
umns of the data matrix (i.e. the first three rows in Table III) were excluded from the
calculation because of their lower modelling power. From the PCA method it follows
that 98% of the variability is described by a single significant latent variable which is
at the same time significant for all the columns. The concentrations participate uni-
formly in creating the model. As the score vector describes the quality of substrates, the
latent variable is a new set of constants describing the substituent effects. In this case
the dominant effect of substituents is their steric effect. The first principal component
from the PCA calculation (PCA : Me 3.39, Et 3.28, i-Pr −0.239, Bu −2.93, cHx −2.93)
1
2
9
was used, in analogy to the classic steric constants E and ν , as the vector explaining
s
ef
the variables for the simple linear regression of the dependence of the catalytic rate
constant k upon the steric parameters of the model substrates. Logically, the results of
H
this calculation show that the effect of substituents is best described by the parameter
obtained from the PCA calculation (r = 0.995). With the application of E parameters
s
the correlation coefficient was r = 0.908, and with νef the correlation was statistically
Collect. Czech. Chem. Commun. (Vol. 59) (1994)

Decomposition of 1-Phenyl-3,3-dialkyltriazenes
411
insignificant. The correlation between the K values and the steric constants given was
K
insignificant too. Hence the substituent effect on the course of the model reaction in-
vestigated is more complex and cannot be described by the substituent steric effects
alone.
REFERENCES
1
2
3
4
5
6
7
8
9
. Hughes E. D., Ingold C. K.: Q. Rev., Chem. Soc. 6, 34 (1952).
. Shine H. Y.: Aromatic Rearrangements, p. 212. Elsevier, Amsterdam 1967.
. Isaacs N. S., Rannala E.: J. Chem. Soc., Perkin Trans. 2 1974, 899.
. Isaacs N. S., Rannala E.: J. Chem. Soc., Perkin Trans. 2 1974, 902.
. Benes J., Beranek V., Zimprich J., Vetesnik P.: Collect. Czech. Chem. Commun. 42, 702 (1977).
. Pytela O., Vecera M., Vetesnik P.: Collect. Czech. Chem. Commun. 45, 1269 (1980).
. Pytela O., Vecera M., Vetesnik P.: Collect. Czech. Chem. Commun. 45, 2108 (1980).
. Pytela O., Vecera M., Vetesnik P.: Collect. Czech. Chem. Commun. 46, 898 (1981).
. Pytela O., Svoboda P., Vecera M.: Collect. Czech. Chem. Commun. 46, 2091 (1981).
1
1
0. Ludwig M., Pytela O., Vecera M.: Collect. Czech. Chem. Commun. 46, 3104 (1981).
1. Jones C. C., Kelly M. A., Sinnott M. L., Smith P. J., Tzotzos G. T.: J. Chem. Soc., Perkin Trans.
2
1982, 1655.
1
1
1
2. Sieh D. H., Michejda C. J.: J. Am. Chem. Soc. 103, 442 (1982).
3. Pytela O., Pilny M., Vecera M.: Collect. Czech. Chem. Commun. 49, 1173 (1984).
4. Smith R. H., Dehlinger C. L., Kupper R., Koepke S. R., Michejda C. J.: J. Am. Chem. Soc. 106,
1
056 (1984).
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
5. Svoboda P., Pytela O., Vecera M.: Collect. Czech. Chem. Commun. 51, 553 (1986).
6. Svoboda P., Pytela O., Vecera M.: Collect. Czech. Chem. Commun. 52, 2217 (1987).
7. Pytela O., Svoboda P., Vecera M.: Collect. Czech. Chem. Commun. 52, 2492 (1987).
8. Iley J., Moreira R., Rosa E.: J. Chem. Soc., Perkin Trans 2 1987, 1503.
9. Nevecna T., Pytela O., Ludwig M., Kavalek J.: Collect. Czech. Chem. Commun. 55, 147 (1990).
0. Pytela O., Nevecna T., Ludwig M.: Collect. Czech. Chem. Commun. 55, 156 (1990).
1. Pytela O., Nevecna T., Kavalek J.: Collect. Czech. Chem. Commun. 55, 2701 (1990).
2. Schmiedekamp A., Smith R. H., Michejda C. J.: J. Org. Chem. 53, 3433 (1989).
3. Nduen M. T., Hoesch L.: Helv. Chim. Acta 69, 1627 (1986).
4. Rodnestvedt C. S., Davis S. J.: J. Org. Chem. 22, 200 (1957).
5. Ludwig M., Pytela O., Kalfus K., Vecera M.: Collect. Czech. Chem. Commun. 49, 1182 (1984).
6. Taft R. W.: Prog. Phys. Org. Chem. 13, 135 (1981).
7. Pytela O., Vecera M., Vetesnik P.: Chem. Listy 73, 754 (1979).
8. Wold S., Esbensen K., Geladi P.: Chemometrics and Intel. Chem. Lab. Syst. 2, 37 (1987).
9. Exner O.: Korelacni vztahy v organicke chemii. SNTL, Praha 1981.
Translated by J. Panchartek.
Collect. Czech. Chem. Commun. (Vol. 59) (1994)