Full Papers
ed over time, possibly through agglomeration or a resting
state of the catalyst. Vanadium(IV) could represent such a rest-
ing state because it is only present in very low concentrations
at the beginning of the reaction when the homogeneous cata-
lyst is most active and increases in quantity over the course of
the reaction as the reactivity of the system decreases. This also
means that the catalytically active species does not contain
a vanadium(IV) center.
further proven by MALDI MS measurements (see Supporting
Information).
With the optimized reaction conditions in hand, the general-
ity of the approach was demonstrated by the conversions of
other lignin samples that had undergone different pretreat-
ment conditions (Sigma–Aldrich kraft lignin #471003 and
Sigma–Aldrich kraft lignin #370959). In all cases, the main
products had masses of approximately 300 Da or lower (see
Supporting Information). Notably, the impurities in both kraft
lignin sources did not affect the activity of the catalyst, which
underlines the robustness of the system. A control experiment
with organosolv beech 7 confirmed the key role of the cata-
lyst. Without catalyst, no depolymerization occurred, and the
formation of products of 4000 Da and greater was enhanced
significantly.
The fundamental understanding of the reaction and catalyst
behavior in the conversion of the model compound 1 contain-
ing a b-O-4 linkage was essential for the subsequent studies of
lignin degradations. Lignin model compounds represent a very
valuable tool to assess the preliminary catalytic activity and se-
lectivity of reaction systems for prominent bonds within lignin.
However, a direct correlation or comparison of the results ob-
tained with model compounds and actual lignin is difficult, as
the higher complexity and the presence of impurities in natu-
ral and extracted lignin frequently lead to a reduction in cata-
lyst activity.[3c] The utilized lignin samples were analyzed and
characterized by calibrated gel permeation chromatography
(GPC). Initial experiments with organsolv beech lignin 7 and 8
(for the pretreatment conditions of the lignin sources, see Sup-
porting Information) showed that more dilute reaction solu-
tions than those used previously for the cleavage of the model
compound were necessary to improve the solubility of the
lignin. Furthermore, longer reaction times and higher dioxygen
pressures were needed. Gratifyingly, with a reaction time of
40 h and 10 bar of dioxygen pressure, up to 86% of the origi-
nal mass of lignin was isolated after work-up, and more than
80% of this lignin sample was degraded to products of a re-
duced size. Excellent selectivities were observed, and the maxi-
mum of the mass peak shifted from approximately 1400 Da
before the reaction to 300 Da after the reaction, and this mass
corresponds to dimeric lignin fragments (Figure 4).[22] In addi-
tion to the depolymerization, small amounts of lignin were
converted to masses of 4000 to 6000 Da. Whether these were
polymerization products or agglomerates of products with
lower masses could not be determined with absolute certainty.
The depolymerization of lignin to lower molecular masses was
Finally, with 5 wt% of V(acac)3 and Cu(NO3)2·3H2O as the cat-
alyst, similar results were obtained, and cleavage products
with a mass of less than 300 Da were formed. To the best of
our knowledge, this is the first example in which the size re-
duction of lignin to such a high degree through selective cata-
lytic oxidative depolymerization is reported and has been
quantified by both calibrated GPC and MALDI MS measure-
ments.
To determine the extent of the cleavage of the b-O-4 and
other characteristic lignin linkages, 2D NMR (HSQC) spectrosco-
py measurements before and after the reaction with deuterat-
ed pyridine as the solvent were performed (Figure 5). The in-
terpretation of the spectra followed the method outlined by
Sun and co-workers.[23] Accordingly, in both organosolv beech
lignins 7 and 8 as well as Sigma–Aldrich kraft lignin #370959,
the b-O-4 linkages, acetylated b-O-4 linkages, resinol structures
and p-hydroxycinnamyl alcohols identified in the starting ma-
terial were completely degraded during the reaction. Especially
the cleavage of the resinol structures is noteworthy because
previous NMR-monitored lignin studies by us[5c] and Westwood
and co-workers[24] have shown that the resinol structure is ac-
tually more prominent in pretreated lignin sources than its nat-
ural abundance in native lignin would suggest. To the best of
our knowledge we report the first transition metal system
using dioxygen as oxidant that is able to cleave the resinol
structure in its entirety independently of which pretreatment
technology was used. Other structural modifications of lignin
could not be specified by this NMR technique although they
must have occurred as revealed by the degradation of Sigma–
Aldrich kraft lignin #471003, where b-O-4 linkages, for example,
are absent owing to the pretreatment process. Consequently,
it can be concluded that the newly devised catalyst systems
are active for the cleavage of not only b-O-4 linkages but also
other significant bonds that contribute to the original molecu-
lar size of lignin.
Conclusions
We have investigated the oxidative cleavage of lignin with di-
oxygen as the oxidant and inexpensive transition-metal-con-
taining hydrotalcites (HTc) or combinations of V(acac)3 and
Cu(NO3)2·3H2O as catalysts. In both the conversion of a model
Figure 4. GPC mass distributions of organosolv beech lignin 8. The mass dis-
tribution of 8 before the reaction is shown in gray, and that after the reac-
tion is shown in black. For additional information on the use of different
lignin sources and catalysts, see Supporting Information.
ChemSusChem 2015, 8, 2106 – 2113
2110
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim