N.I. Kuznetsova et al.
Catalysis Communications 149 (2021) 106218
gradually increasing quantity of Fe(acac)3 and unchanged NHPI and
Phen, conversion of IPB increased, at least as long as Fe(acac)3 was
equimolar or slightly exceeded Phen (#11, #9 and #2). Since generation
of PINO is a slow step in the mechanism of NHPI catalyzed oxidation of
IPB (route A, Scheme 1) [26], observed acceleration was connected with
rapid interaction of NHPI with Fe(III) to form PINO (route A in Scheme
2). The fact that this interaction was controlled by Phen could find an
explanation in the formation of highly reactive complexes of Fe(III)-
Phen under the reaction conditions. Indeed, ligating with Phen en-
hances reduction potential of Fe(III)/Fe(II) redox couple from 0 0.771 to
1.14 V (vs. SHE) [27]. Well-known is a high oxidizing ability of binu-
clear iron-Phen complexes readily formed from Fe(III) nitrate, chloride
or perchlorate in water-acetonitrile solvent [28].
temperature of catalysis by metal ions [3]. The low-temperature reac-
tion was apparently caused by the above-discussed strong increase in the
oxidizing ability of Fe(acac)3 in the presence of Phen. A similar phe-
nomenon was reported for Cu(II)/Phen catalyst [29,30]. Low rate of
reaction and, accordingly, small conversion of IPB in #1corresponded to
a low concentration of the active components. Addition of a solid NHPI
caused a more intense oxidation of IPB under the same conditions. In the
presence of NHPI/Fe(acac)3/Phen, oxygen consumption by the reaction
mixture became noticeable 30 min after start of heating of the reactor
and continued at a constant rate until the reaction began to slow down at
a high conversion of IPB (Fig.2).
Variation of the reaction time in several experiments showed that
formation of the oxidation products grew with time, and the selectivity
for IPBHP was over 82% in experiments #2–5 (Table 2) with a slight
increase when conversion of IPB approached 37% in #5. The oxidation
was certainly participated by NHPI. HPLC and NMR analysis showed no
dissolution of NHPI in pure IPB, but NHPI did dissolve in IPB in the
presence of Phen at room temperature. Addition of a solid NHPI into
0.34٠
10ꢀ 3 M Phen solution in IPB led to appearance of approximately
equimolar amount of NHPI in solution. Associated with Phen, the sol-
ubility of NHPI suggested an interaction of two components, which was
confirmed by 13С and 1Н NMR. Indeed, introduction of Phen caused a
shift in NMR signals of NHPI in the acetonitrile solution. The spectral
characteristics corresponded to formation of the hydrogen-bonded
adduct of Phen and NHPI (Supplementary 3), for which the DFT calcu-
lation provided a structure shown in Fig.3. A similar phenomenon has
already been established for the binding of lipophilic NHPI to acetoni-
trile [14].
Contrary to our expectations, NMR spectroscopy proved inertness of
Fe(acac)3 with respect to Phen. Indeed, a lack of the ligand exchange
was evidenced from the absence of signals from free acetylacetone when
Phen was added to solution of Fe(acac)3. Rapid changes in the spectrum
began after the subsequent addition of NHPI, which manifested itself in
weakening the signals from Fe(acac)3 (broad signals at ꢀ 26.2 ppm
(width 1240 Hz, 1H, CH) and 22.3 ppm (width 700 Hz, 6H, 2CH3)) and
appearance of signals from free acetylacetone (δ 2.14 (s, 6H, 2CH3), 3.60
(s, 2H, CH2) of keto and 2.01 (s, 6H, 2CH3) 5.60 (s, 1H, CH), 15.6 (br.s,
1H, OH) of enol tautomers). Initiated by NHPI (probably, phthalimide-
N-oxyl anion), the removal of the ligand from Fe(acac)3 was followed
by the binding of Phen and inner-sphere charge transfer to form Fe
(Phen)23+ complex (δ 7.60 (m, 2H, H3,8), 7.67 (d, 2H, H2,9), 8.24 (s, 2H,
H5,6), 8.60 (d, 2H, 4,7)) and PINO. (More NMR data are in Supple-
mentary 2).
Thus encouraged by Phen, the intense redox-interaction of NHPI
with Fe(III) (route A in Scheme 2) was the main driving force for the
rapid oxidation of IPB in the system NHPI/Fe(acac)3/Phen. Re-oxidation
of Fe(II) occurred under the action of isopropylbenzene peroxy radicals
or IPBHP to form IPBHP in the first case and alkoxy radicals, then PP in
the second case (routes B in Scheme 2). The oxidation of Fe(II) with
molecular oxygen did not take place, in contrast to catalysis by NHPI/Co
(II). At a small amounts of Fe(acac)3 and Phen, PP appeared also in small
amounts, which indicated reaction with ROO⋅ to be predominant way of
Fe(II) oxidation. In addition, low concentrations of Fe(acac)3 did not
cause significant decomposition of IPBHP (route C in Scheme 2), which
comprised 98% of the oxidation products (#12, Table 1).
Assisted by Phen, reaction of dissolved NHPI and Fe(acac)3 (route A
in Scheme 2) afforded PINO which, in turn, initiated oxidation of IPB.
The same triple interaction is very likely to occur under the contact of Fe
(acac)3 solution with surface of the solid NHPI containing adsorbed
Phen. The Phen-NHPI interaction detected in acetonitrile solution was
responsible for the adsorption of Phen on solid NHPI providing an
additional pathway to generation of PINO. Thus, the interaction of Fe
(III) with Phen-associated NHPI in solution and on a solid surface
created necessary concentration of NHPI/PINO to start the oxidation,
and the concentration increased in time due to dissolution of NHPI in the
products, mainly IPBHP (#2–5, Table 2). When the reaction was
quenched after 160 min, only 5% of the added solid NHPI (0.04 mol%
relative to IPB) was detected in the solution (#2), but after 570 min of
the reaction already 50% NHPI was dissolved (#5). Note that the con-
centrations of NHPI were measured after the solutions were cooled to
15◦C, and these quantities should have been greater at the reaction
3.2. Solvent-free oxidation of IPB mediated by NHPI/Fe(acac)3/Phen
In previous experiments, the reaction mixture contained 14%
acetonitrile, which ensured almost complete dissolution of NHPI at
◦
temperature of 60 C. The gradual increase in concentration of NHPI
contributed to a slight increase in selectivity for IPBHP, as would be
expected according to Scheme 1.
◦
60 C. Subsequently, the oxidation of IPB mediated by the system Fe
(acac)3/NHPI/Phen was tried in the absence of the polar solvent. First, a
combination of IPB-soluble Fe(acac)3 and Phen was tested. It was
observed that the oxidation of IPB to predominantly IPBHP occurred
without NHPI at 60 ◦C (#1, Table 2) which is below the typical
As it turned out, the oxidation proceeded more slowly in pure IPB
than in dilute one. A twofold difference in the rate was estimated from
the linear part of the kinetic curves (Fig.2). It took more time, for
example, to achieve an IBP conversion of 17% and lower selectivity for
IPBHP was obtained in the absence than in presence of acetonitrile (#3
and #6 in Table 2). This was apparently due to the difference in con-
centration of NHPI in the reaction solution. Similar to the oxidation in
acetonitrile solution, the oxidation without solvent showed the sensi-
tivity of the products composition to the quantity of Fe(acac)3 and Phen.
This phenomenon is applicable to control the selectivity for each of the
major products. Indeed, the formation of PP became more selective as
the content of Fe(acac)3 and Phen increased (#7 and #8, Table 2), and,
conversely, IPBHP was practically the only product when using trace
amount of Fe(acac)3 (#9). Note that high selectivity for IPBHP has to be
interesting in the case of targeted synthesis of hydroperoxide.
It was also important for synthetic purpose to ensure sufficient sta-
bility of NHPI. The estimates made so far have shown that more than
95% of the starting NHPI remained intact in solution after the oxidation
of pure and acetonitrile diluted IPB (Supplementary 4).
Scheme 2. Mediator transformations during oxidation of IPB.
3
Babushkin, D. E.
