M. Linhares et al. / Applied Catalysis A: General 470 (2014) 427–433
431
resulting from indole ring cleavage. It is worth to refer that dur-
ing the reaction monitoring by TLC, the blue spot due to indigo
was detected but at the end of the reaction the pigment could not
be isolated, indicating its degradation in these more severe con-
ditions (longer reaction times and higher excess of oxidant). The
yield on indirubin (I.3) decreased to 3% instead of 10%, when com-
pared with the previous conditions. The yield of 2-oxoindole (I.2)
also decreased to 39%, while the yield of trimer (I.4) was nearly
maintained (13%); however, the yields of isatin (I.5) and of the frac-
tion identified as “others” increased to 20% and 24%, respectively.
Probably, isatin and the fraction of “other products” result from
further oxidation of the initially formed compounds in the reaction
conditions used, and consequently indigo pigment I.1 could not be
obtained; therefore, a careful control of reaction time and amount
of oxidant added is required in order to avoid over-oxidation of the
former derivatives.
Some literature results showed that indigo is readily oxidized
in solution even in soft oxidizing media, while once precipitated it
behaves as a quite stable compound toward oxidation [38]. Thus,
the production of pigments should be achieved with shorter reac-
tion times and considering their precipitation or removal from the
oxidizing media in the reaction mixture before further oxidation. In
this context, in order to have a precise quantification of the indigo
produced, the subsequent reactions were performed and stopped,
at appropriate time, by passing the reaction mixtures through a
small plug of silica gel in order to retain the excess of catalyst and
oxidant. The compounds were then eluted using a mixture of chlo-
roform and ethyl acetate and the solvents removed. It should be
noted that after elution from the silica, the eluate showed a clear
green color, while after solvent evaporation, the formation of pig-
ments could be detected in the residue of the reaction mixture,
once the solid residue turned into dark blue or violet color, depend-
ing on the relative amounts of indigo and indirubin present. Indigo
is highly insoluble in almost all the common solvents, although
solutions can be obtained in DMSO or DMF. In order to obtain
an accurate quantification of the reaction products, each reaction
was performed in duplicate and the residue of the first assay was
dissolved in DMSO-d6 for 1H NMR analysis of the total reaction
mixture and quantification of product yields (S.I.3); the residue of
the second assay was dissolved in a precise volume of DMSO for
UV–vis analysis and quantification of indigo formed, by compari-
yields obtained by 1H NMR. These procedures were used to obtain
the results in all the subsequent studies, shown in Table 1, entries
3–8.
Further reactions were performed at shorter times, 2 h and 1 h
(Table 1, entries 3 and 4), and keeping the addition rate of H2O2
at 10 mol min−1, the total amount of oxidant added was reduced
to 4 equiv. and 2 equiv., respectively (Table 1, entries 3 and 4). The
observed conversions were 84% and 65%, respectively, lower than
in the initially performed reactions, but it was possible to isolate
indigo in 2% yield in the last reaction (entry 4).
In another experiment, the total addition of 4 mol equiv. of oxi-
higher rates than in the previous reactions (60 mol min−1 instead
of 10 mol min−1) in order to further diminish the reaction time.
intervals and the spectra are shown in Fig. 3. After 20 min, a total
amount of 4 equiv. has been added and the monitoring proceeded
were observed and a total reaction time of 30 min was considered.
Changes in the spectral pattern in Fig. 3 with increasing reaction
time can be noticed. The Soret band at ꢁ = 478 nm is slightly shifted
to lower wavelengths and its intensity decreases as expected from
bleaching of the metalloporphyrin [26]. Concomitantly, the band
at near ꢁ = 375 nm showed a significant and progressive increase in
Fig. 3. UV–vis spectra of the reaction mixture in conditions of Table 1, entry 5,
recorded at 3 min intervals.
intensity, which cannot be related with the original metal–ligand
charge transfer band of the metalloporphyrin, since as mentioned,
ceeded. In accordance with reported data, the formed band at
[39], which might be in equilibrium with their enol forms, the
indoxyls (hydroxyindoles). However, as can be seen in the spectra
of Fig. 3, the pigments are not detected during the reaction time,
since no bands are observed for indigo (ꢁmax = 603 nm) and indiru-
bin (ꢁmax = 550 nm, Fig. 2a). Only when the reaction mixture was
applied on TLC silica plates and after solvent drying, the blue and
red spots become visible during elution. This indicates that indigo
or indirubin are not significantly formed during the short catalytic
reaction of 30 min, but are formed during the work-up.
Since indoxyls are the main resulting products of the catalytic
reaction and are the key precursors of other reaction products in
indole oxidations, specifically 3-indoxyl for indigo, Fig. 3 indicates
the kinetic profile of the catalytic reaction of indole oxidation.
Furthermore, the possibility of separating the fraction of
indoxyls from the reaction media (catalysts and oxidant) can be
advantageous since it may allow the isolation of the pigments pre-
cursors from the oxidizing media, and performing the coupling
reactions in a further stage, thus avoiding their over-oxidation in
the reaction mixture.
An additional reaction was also performed in the last conditions
but the addition of hydrogen peroxide was performed in the form
of adduct with urea (UHP), in order to check if this anhydrous oxi-
dant could have a positive effect. No significant differences were
observed in the yields of the reaction products relatively to the use
of aqueous hydrogen peroxide, showing that this cheapest oxidant,
H2O2, is appropriate.
rins, [Mn(-NO2TDCPP)Cl] and [Mn(TF4NMe2PP)Cl] were tested in
indole oxidation using the conditions described in Table 1, entry
5, for [Mn(TDCPP)Cl]: reaction time of 30 min, with H2O2 addi-
tion at a rate of 60 mol min−1 in a total amount of 4 mol equiv.
of H2O2 relatively to the quantity of substrate (Table 1, entries
6 and 7). The increase of the electron-withdrawing characteris-
tics of the metalloporphyrin nucleus is acceded by the increase
of the chemical deviation observed in the 1H NMR spectra for
-protons: ı = 8.67 ppm for H2TDCPP < ı = 8.72–8.79 ppm for H2-
NO2TDCPP < ı = 8.92 ppm for H2TF5PP.
A decrease in conversion was observed for the more elec-
tron withdrawing metalloporphyrins, namely from 86% with