Electrochemical Lignin Degradation in Ionic Liquids on Ternary Mixed Metal Electrodes

Article abstract of DOI:10.1515/zpch-2017-0951

Lignin is the second most abundant natural polymer and a promissing feedstock for the generation of renewable aromatic chemicals. We present an fundamental approach for the electrocatalytic cleavage of lignin dissolved in a recoverable, inexpensive ionic liquid using mixed metal oxide electrodes of different compositions. The distribution of depolymerization products generated by electrochemical oxidation were analyzed by means of mass spectrometry. The distribution and yield of the cracked species was found to depended strongly on the implemented metal catalyst and therefore offers the potential to tailor the amount and composition of the low molecular weight cleavage products. This approach could help to provide a more sustainable valorization of lignin for the potential production of high value aromatic compounds due to synergistic effects.

Full text of DOI:10.1515/zpch-2017-0951

Z. Phys. Chem. 2017; aop
Daniel Rauber, Tobias K.F. Dier, Dietrich A. Volmer and
Rolf Hempelmann*
Electrochemical Lignin Degradation in Ionic
Liquids on Ternary Mixed Metal Electrodes
https://doi.org/10.1515/zpch-2017-0951
Received February 28, 2017; accepted August 21, 2017
Abstract: Lignin is the second most abundant natural polymer and a promiss-
ing feedstock for the generation of renewable aromatic chemicals. We present an
fundamental approach for the electrocatalytic cleavage of lignin dissolved in a
recoverable, inexpensive ionic liquid using mixed metal oxide electrodes of dif-
ferent compositions. The distribution of depolymerization products generated by
electrochemical oxidation were analyzed by means of mass spectrometry. The
distribution and yield of the cracked species was found to depended strongly on
the implemented metal catalyst and therefore offers the potential to tailor the
amount and composition of the low molecular weight cleavage products. This
approach could help to provide a more sustainable valorization of lignin for the
potential production of high value aromatic compounds due to synergistic effects.
Keywords: electrochemistry; ionic liquids; lignin; sustainable.
1 Introduction
Lignin is second to cellulose the most abundant natural polymer and a major com-
ponent of woody plants [1, 2]. The phenolic macromolecule is a large scale byprod-
uct of the sugar cane milling and wood pulp industries and mainly used as low-grad
fuel in pulping processes [3]. The usage of this renewable feedstock for sustainable
production of aromatic chemicals is of major interest in green chemistry applica-
tions and highly desirable from economic and ecological points of view [4]. There-
*Corresponding author: Rolf Hempelmann, Institute of Physical Chemistry, Saarland University,
Campus B2.2, 66123 Saarbrücken, Germany; and Transfercentre Sustainable Electrochemistry,
Saarland University and KIST Europe, 66125 Saarbrücken, Germany,
e-mail: r.hempelmann@mx.uni-saarland.de
Daniel Rauber: Institute of Physical Chemistry, Saarland University, Campus B2.2, 66123 Saar-
brücken, Germany; and Transfercentre Sustainable Electrochemistry, Saarland University and
KIST Europe, 66125 Saarbrücken, Germany
Tobias K.F. Dier and Dietrich A. Volmer: Institute of Bioanalytical Chemistry, Saarland
University, Campus B2.2, 66123 Saarbrücken, Germany
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
2ꢀ
Fig. 1:ꢀStructure of the monolignols (left) and resulting complex heterogenous structure of
lignin (right). Indicated linkages are biphenyl (1), diphenly ether (2), dibenzyl ether (3), β–O–4
(4), β–5 (5), β–β (6) and β–1 (7).
fore the conversion of lignin into high value chemicals [5] has, in theory, enormous
potential for sustainable development, as it is not competing with food production
for cropland [6]. Lignin has a complex heterogenous polymeric macrostructure,
which is derived from the three common monolignols 4-coumaryl alcohol, coniferyl
alcohol and sinapyl alcohol [7]. Due to the particular nature of the lignin biosyn-
thesis, there is a variety of different linkages between the monomers in the lignin
structure [8]. The molecular structures of the monolignols and lignin structure,
including different possible linkages, are shown in Figure 1.
Among the different linkages, the β–O–4 linkage was found to be the most dom-
inant bond, constituing approximately 50% of all lignin linkages [9]. In theory, con-
version of lignin to high-value low-molecular weight chemicals has a great portential,
if the whole process can be designed in a sustainable way, as it is the only renewable
source for aromatic compounds available on a large scale. The valorization of lignin
is complicated by the chemical and biochemical resistance of lignin and poor solu-
bility in most common organic solvents [10]. Ionic liquids (ILs) are a relatively new,
diverse class of solvents that consist of salts with melting points below 100ꢀ°C. They
exhibit a unique combination of properties such as non-volatility and conductivity,
along with thermal, chemical and electrochemical stability. Their properties can be
modified over a wide range by varying the anion and cation combination or by incor-
porating different chemical functionalities. ILs were found to be powerful solvents
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
ꢂ3
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
for biomass and can dissolve various types of lignins and even untreated woods,
in large quantities [11, 12]. Furthermore, these neoteric solvents can be designed to
selectively dissolve cellulose or lignin from biomass, and therefore permit sustain-
able processing and utilization of this renewable feedstocks [13]. Along with their
intrinsic ionic conductivity originating from their composition of ions and their wide
electrochemical window, ILs offer the ability to act as solvent-free electrolytes for
electrochemical degradation of dissolved lignin [14, 15]. Other electrochemical lignin
depolymerization approaches, that were mainly performed in aqueous media, suffer
from several disadvantages such as the limitation to hydrophilic lignins, limited
electrochemical window for water or the need for large amounts of additives such
as corrosive sodium hydroxide [16] that in most cases also needs to be neutralized
before work-up. Electrochemical lignin depolymerization potentially offers a less
energy-consuming and better controllable method than competing technologies
such as pyrolysis or hydrogenation reactions [17] as the electrode potential and com-
position could be easily controlled or modified. Together with the recoverability of
the used hydrophilic IL via aqueous extraction, this could in theory lead to a more
environmentally friendly and economic process that does not require harsh reaction
conditions, further additives or non-recyclable organometallic catalysts [18]. The
electrochemical oxidation requires electrodes being stable towards anodic corrosion
and can act as catalyst for the depolymerization reactions. Examples are electrodes
based on RuO2 for generation of chlorine gas in the chlor-alkali process [19]. These
types of electrodes show high conductivity, catalytic activity along with mechanical
and chemical stability, even under harsh conditions. Therefore, they are referred to
as dimensionally stable electrodes (DSA). In the present study, we prepared a series
of DSA electrodes based on the combined use of ruthenium and iridium oxides as
binary DSA electrodes, as well as with other transition metals as ternary electrodes.
Although electrodes based on these transition metals are quite expensive the overall
cost of electrochemical processes implementing these types of electrodes could
benefit from their stability and catalytic activity. These mixed metal oxide (MMO)
electrodes were investigated with respect to their performance for electrochemical
lignin degradation in a protic ionic liquid. Binary MMO electrodes based on ruthe-
nium and iridium were previously utilized for lignin conversion in aqueous media,
where the electrochemical window is much narrower due to the electrolyte decom-
position [20]. The composition of the investigated binary coating for the MMO was
Ru0.4Ir0.6Ox; as ternary oxide coatings, Ru0.2Mn0.2Ir0.6Ox, Ru0.2Pd0.2Ir0.6Ox, Ru0.2V0.2Ir0.6Ox
and Ru0.2Ti0.2Ir0.6Ox were chosen. The solvent for lignin was the IL triethylammonium
methanesulfonate [Et3NH][MeSO3]. This protic IL was prepared by acid-base neu-
tralization from readily available precursors and was found to exhibit high solubility
for different types of lignins. The use of this IL ensured quantitative yield of the syn-
thesis, absence of halides and comparable inexpensive synthesis; this is important
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
4ꢀ
for potential future use in applications, where larger quantities are needed. Protic
ILs are generally favorable solvents for biomass, since the synthesis is environmen-
tally friendly as a result of the quantitative turnover and the absence of side reaction
[21]. This particular IL was found to exhibit a high electrochemical window, which
is essential for the electrochemical lignin degradation at higher voltages, and it
showed no decomposition with the investigated MMOs. Furthermore, the IL is able
to dissolve very large amounts of different lignins forming solutions that remained
stable in the liquid state, even below the melting point of the IL. Recovery of the
IL and separation from the hydrophobic extractable aromatic cleavage product and
untreated lignin was achieved by addition of water due to the hydrophilic nature
of the IL. Analysis of the multiple degradation products was performed using gas
chromatography-mass spectrometry (GC-MS) and direct infusion high resolution
mass spectrometry (DI-HRMS). The degradation products were found to maintain
many of their original chemical functionalities, which in general is favorable for the
generation of biomass-derived fine chemicals. The results indicated pronounced
differences of cleavage products as a function of the MMO composition and could
therefore be of interest for further developments.
2 Results
2.1 Cyclovoltammetric measurements
Cyclic voltammetry (CV) of pure, dried IL was performed to investigate the elec-
trochemical stability of anion and cation on the coated MMO electrodes. Both
the triethylammonium cation and the methanesulfonate anion showed high
electrochemical stability on the five measured electrodes. The cathodic stabil-
ity for reduction of the cation was found to be in the range of −2.2 to −2.0 V vs.
Ag-quasi-reference electrode, i.e. an higher stability than reported for Au (−0.90 V)
and Pt electrodes (−0.42 V), but a lower stability than on glassy carbon electrodes
(−2.70 V) [22]. Values for the anodic limit of [Et3NH][MeSO3] were in the region of
1.9–2.1 V vs. Ag-quasi reference, and were in the same range as reported for glassy
carbon (2.03 V), Au (1.93 V) and Pt electrodes (2.05 V) [22]. The overall electro-
chemical window for [Et3NH][MeSO3] for the investigated MMO electrodes was
between 3.9 and 4.3 V, which is comparably high for a protic IL [23]. The use of
the methanesulfonate anion was particularly beneficial for application to electro-
chemical oxidation reactions, since it showed a much higher anodic stability than
other anions incorporated in ILs for lignin dissolution, such as acetate, chloride or
dialkylphosphates. Other ILs consisting of the low coordinating anions tetrafluor-
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ ꢂ5
Fig. 2:ꢀCyclic voltammogram of [Et3NH][MeSO3] at Ru02Mn0.2Ir0.6Ox electrodes performed at 40ꢁ°C.
−
−
oborate BF4 , hexafluorophosphate PF6 or bis(trifluoromethanesulfonyl)imide
NTf2− exhibit higher anodic stabilities, but only negligible dissolution ability for
different types of lignins, even at elevated temperatures [24]; they were therefore
not applicable in our study. The comparison of anodic stabilities of the MMO elec-
trodes with those of other electrodes indicated a much stronger dependence of
the electrode material on the stability of the cation than of the anion. For electro-
chemical reactions it was therefore important to determine the combined stability
of electrode material and ionic liquid, to ensure that there is no decomposition
resulting in reduced recovery yield, impurities in the product from IL degradation,
passivation or degradation of the electrode. Figure 2 shows a representative cyclic
voltammogram of the IL on Ru02Mn0.2Ir0.6Ox electrodes. As the CV measurements
showed an anodic stability of the IL on all MMO-electrodes for applied voltages
below 1.9 V, we performed the electrochemical oxidation of lignin at 1.8 V. In this
way, we ensured that no degradation of the solvent takes place and the maximum
applicable voltage for the lignin degradation experiments was used; higher volt-
ages were found to result in higher degradation rates [14].
2.2 Thermal properties of the lignin solutions
The thermal behavior of pure [Et3NH][MeSO3] and solutions of ILs containing differ-
ent amounts of lignin was investigated by differential scanning calorimetry (DSC), as
the thermal behavior is very important in potential technical applications. Low scan-
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
6ꢀ
Tab. 1:ꢀCrystallization (TC), glass transition (Tg), cold crystallization (Tcc) and melting points for
pure IL and lignin solutions containing different weight percentages of lignin in the IL.
wt% Lignin
Tc (°C)
Tg (°C)
Tcc (°C)
Tm (°C)
0
−3
−22
–
–
–
−71
−76
–
–
−23
−6
36
33
32
31
5
10
20
–
ning rates of 1ꢀ°C were applied to ensure correct determinations of the phase transi-
tion temperatures. The detected transition temperatures are given in Table 1, and the
resulting thermograms are displayed in Figure 3. Pure [Et3NH][MeSO3] exhibited a
very sharp crystallization point (Tc) at −3ꢀ°C in the cooling cycle and a melting point
(Tm) of 36ꢀ°C in the heating step, showing that this IL can be distinctly supercooled in
the pure state. The occurrence of supercooling is a widely observed phenomenon in
ILs [25], which results from their molecular design to destabilize crystal structures
and to remain in the liquid state. The 5 wt% lignin solution showed a much broader
crystallization point at −22ꢀ°C and a slightly decreased melting point of 33ꢀ°C.
Fig. 3:ꢀNormalized thermograms of pure IL and solutions of the IL containing 5 and 10 wt%
of lignin with indicated transition temperatures. Exothermal events are plotted as positive
peaks.
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
ꢂ7
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
For the IL containing 10 wt% of lignin, the thermogram was different, showing
a glass transition (Tg) point at −71ꢀ°C upon cooling instead of a crystallization
point. Upon heating, a cold crystallization transition (Tcc) occurred at −23ꢀ°C fol-
lowed by a melting point at 32ꢀ°C. This indicated a stability of the obtained 10 wt%
lignin solution at 20ꢀ°C temperature without crystallization or phase separation,
although the IL has a melting point above ambient temperature. For the 20 wt%
solution, the situation was virtually identical, with the cold crystallization tem-
perature shifted to −6ꢀ°C. The 10 wt% solutions were also found to be stable in
the liquid state, even for months, whereas for the pure IL spontaneous crystal-
lization occurred. This demonstrated that even ILs that are not room temperature
ionic liquids (RTILs) could be used for the dissolution and treatment of biomass
at 20ꢀ°C in pure form, as they remain in their liquid form. This concept therefore
combines the advantages of RTILs for processing at ambient temperature with the
inexpensive synthesis of protic ILs from readily available and cheap precursors,
which have the drawback of the elevated melting point in pure state.
2.3 Electrochemical oxidation of lignin using MMO electrodes
For electrochemical lignin cleavage in protic ionic liquids we used solutions of
alkali lignin obtained from the Kraft process, which is the most commonly gen-
erated form of lignin, corresponding to about 85% of the worldwide lignin pro-
duction [26]. We chose this particular lignin, as it has a low content of sulfur from
modifications, and neglegible water solubility because of the highly hydropho-
bic nature of the unmodified aromatic lignin structure. The low sulfur content
is advantageous for production of fine chemicals, as any unwanted products do
not need to be separated nor desulfurization processes to be applied [27]. We
also tested the ability of the IL to dissolve other hydrophobic lignins, namely
from organosolv or ligninoboost processes. It was found that the IL was also
able to dissolve at least 20 wt% of theses lignins. Hydrophobic lignin with sul-
fonate groups can also be dissolved in triethylammonium methanesulfonate
but it additionally exhibits high water solubility, which prevents recovery of
the IL after electrochemical degradation. Concentration of the lignin in the
IL was 10 wt% for all electrocatalytic oxidations, as this allowed relative high
lignin loadings, moderate viscosity and stability of the solution at ambient tem-
perature. To compensate effects of increased viscosity on mass transfer to the
electrode, which would limit the current density, the reactions were performed
at 70ꢀ°C. The applied potential was set to 1.8 V to ensure that only electrolytic
lignin degradation occures instead of IL decomposition. Electrocatalytic oxida-
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
8ꢀ
Tab. 2:ꢀYield of isolated lignin degradation products and recovered lignin after electrochemical
oxidation for different MMO electrodes.
Electrode
composition
ꢁ
Yield of isolatedꢁ
cracking products (%)
Yield of residualꢁ
treated ligin (%)
Yield of
recovered IL (%)
Ru0.4Ir0.6Ox
ꢃ
7.5ꢃ
11.5ꢃ
10.7ꢃ
10.1ꢃ
9.5ꢃ
92.1ꢃ
88.3ꢃ
89.1ꢃ
89.5ꢃ
90.3ꢃ
99.0
98.9
98.5
99.2
99.1
Ru0.2Mn0.2Ir0.6Ox ꢃ
Ru0.2Pd0.2Ir0.6Ox
Ru0.2V0.2Ir0.6
Ru0.2Ti0.2Ir0.6Ox
ꢃ
ꢃ
ꢃ
O
x
tions were performed for 3 h monitored by chronoamperometry. Longer reaction
times showed only a neglegible increase in the yield of the cracking products.
To ensure that no thermal degradation occures at the slightly elevated tem-
peratures, an experiment without electrochemical treatment was performed as
control. Only trace amounts of organic compounds were found in the organic
phase of the control experiment, and the lignin could be recovered in nearly
quantitative yield. Isolated oxidation products from the electrochemical treat-
ment were obtained as yellow to brown slightly waxy solids. The obtained yield
of lignin degradation products along with the amount of recovered lignin and
[Et3NH][MeSO3], relativ to the initial weight, are summarized in Table 2.
All investigated coated electrodes showed electrochemical activity and
ability to degradate the dissolved lignin into low molecular weight products.
The lowest activity and cracking yield of 7.5% was found for the binary MMO
electrode consisting solely of ruthenium and iridium. Importantly, this value
already exceeded the yield reported for Ru0.25V0.05Ti0.7Ox electrodes, which was
only 6.0% at 1.7 V [14], indicating that MMO incorporating iridium may have a
higher activity for electrochemical lignin depolymerization. The electro-catalytic
activity of the coated electrodes with ternary composition was found to be the
highest for Ru0.2Mn0.2Ir0.6Ox, exhibiting 11.5% yield of degradation-products. The
activity of the electrodes with composition Ru0.2M0.2Ir0.6Ox decreased in the order
Mꢀ=ꢀMnꢀ>ꢀPdꢀ>ꢀVꢀ>ꢀTi. The differences of yield for lignin depolymerization showed
a significant dependence on the added transition metal in the MMO electrodes,
which highlights the potential for advanced electrodes, that could allow more
efficient and thereby more sustainable lignin cracking. The balance of observed
products and recovered treated lignin were always close to 100%, indicating only
small losses from formation of low molecular weight compounds with high water
solubility or volatility, such CO2, water or aliphatic alcohols and aldehydes, which
are irreversibly removed in the separation process. The electrode reactions at the
cathode will be investigated in future studies. One possibility may, for example,
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ ꢂ9
be the breaking of linkages similar to the catalytic reduction approaches leading
to alcohol functionalities [28].
2.4 GC-MS analysis of the degradation products
To get further insight into the degradation products, the isolated products of the
ternary mixed metal oxide electrodes were analyzed by GC-MS, as this method
allows simultaneous separation and identification of low molecular compounds.
Product identification was performed by comparison to commercial standards
and matching their retention times and mass spectra. If compounds were not
available as standards, identification was achieved via NIST database search. The
resulting chromatograms with assigned products for the ternary MMO electrodes
are shown in Figure 4. The corresponding retention times and molecular struc-
tures of identified cleavage products are given in Table 3.
All gas chromatograms of the electrochemical oxidation showed a large
number of peaks resulting from the highly heterogeneous lignin structure as
it was reported earlier [29]. Therefore, only a selected range of the most impor-
tant and unambiguously identified products are given here. In particular, for
Fig. 4:ꢀGas chromatograms for the ternary mixed metal oxide electrodes with identified cleav-
age products assigned (see Table 2 for assignment of the compounds).
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
10ꢀ
Tab. 3:ꢀRetention times and molecular structures of obtained lignin degradation products.
Numberꢁ
Retention time (min)ꢁ
Compound
ꢁ
Moleculare structure
1
ꢃ
4.2ꢃ
3-Methylbenzaldehyde
ꢃ
2
3
4
5
ꢃ
ꢃ
ꢃ
ꢃ
4.4ꢃ
4.9ꢃ
5.7ꢃ
6.5ꢃ
4-Methylbenzaldehyde
ꢃ
Cyclohexanecarboxylic acid ꢃ
Benzoic acid
ꢃ
ꢃ
p-Coumaric acid
6
7
ꢃ
ꢃ
7.2ꢃ
7.7ꢃ
3-Methylbenzoic acid
ꢃ
ꢃ
Ferulic acid
8
9
ꢃ
ꢃ
8.5ꢃ
8.8ꢃ
4-Hydroxybenzaldehyd
ꢃ
ꢃ
Vanillin
10
11
12
ꢃ
ꢃ
ꢃ
9.9ꢃ
10.3ꢃ
10.9ꢃ
4-Hydroxy-2-methoxy
ꢃ
ꢃ
ꢃ
acetophenone
4-Hydroxybenzoic acid
Vanillic acid
13
14
ꢃ
ꢃ
11.9ꢃ
12.1ꢃ
Homovanillic acid
Synapylaldehyde
ꢃ
ꢃ
15
ꢃ
12.4ꢃ
Vanillic acid ethyl ether
ꢃ
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
ꢂ11
Tab. 3ꢀ(continued)
Numberꢁ Retention time (min)ꢁ
Compound
ꢁ
Moleculare structure
16
ꢃ
12.9ꢃ
4-Hydroxy-3,5-dimethoxy
acetophenone
ꢃ
17
18
ꢃ
ꢃ
14.3ꢃ
15.2ꢃ
4-Hydroxy-3-methoxy-
ꢃ
ꢃ
cinnamic acid
4-Hydroxy-3,5-dimethoxy
cinnamaldehyde
retention times >15 min, belonging to products of higher molecular weights,
identification becomes more difficult because of the lack of standards and
reference spectra.
All identified depolymerization products possess functional groups such as
carbonyl, ester or carboxylic acid groups that are the result of oxidation of the
alcohol functionalities in the lignin linkages. This was the expected result for oxi-
dative depolymerization and could be a possible advantage of this approach for
the generation of fine chemicals if the selectivity and upscaling problems can be
solved. Waldvogel and coworkers, e.g. already published a method for the iso-
lation of low molecular weight compounds from the lignin electrolysis [30] as
well as isolation techniques for organic anions utilizing basic anion exchange
resigns [31, 32]. In contrast, reductive approaches such as hydrolysis lead to
the removal of functionalities, which for the production of high value aromatic
chemicals makes these methods less attractive than oxidative approaches [28,
33]. Furthermore, nearly all identified lignin degradation products were aromat-
ics with aromatic core and a high degree of substitution, with hydroxy and meth-
oxy-substituents the most abundant ones. Most of the cleavage products were
derivatives of monolignols; no aliphatic degradation products except cyclohex-
ane carboxylic acid were found. This demonstrates that oxidative electrochemi-
cal lignin depolymerization is breaking the linkages in the lignin structure, thus
leaving the aromatic cores intact. As most of the lignin cleavage products bear
hydroxyl or methoxy groups at the benzene core, these functionalities also seem
to remain intact under the chosen reaction conditions. For the electrochemical
depolymerization of lignin using protic ionic liquids, cleavage of the dominating
β–O–4 linkages was proposed [14]. This was also rationalized from the identified
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
12ꢀ
GC-MS products bearing 4-hydroxy functionalities, which are believed to be
the result of cleavage of the β–O–4 linkages [34]. The ternary MMO electrodes
exhibited distinctly variable chromatograms, showing strong dependence on the
metal composition used. The main degradation product for Ru0.2Mn0.2Ir0.6Ox was
p-coumaric acid, for Ru0.2Pd0.2Ir0.6Ox 4-hydroxy-3,5-dimethoxy cinnamaldehyde,
for Ru0.2V0.2Ir0.6Ox 4-hydroxy-3,5-dimethoxy acetophenone, and for Ru0.2Ti0.2Ir0.6Ox
vanillin. As vanillin is a comparably high-value product and of significant indus-
trial interest, the incorporation of titanium into MMO electrodes might be an eco-
nomically attractive way for the generation of vanillin from lignin, when focusing
on a special molecule instead of the overall cracking yield. The tailoring of the
electrode composition towards particularly desired electrolysis products could
therefore readily be possible with this approach. The lignin cleavage products
from electrodes coated with Ru0.2Mn0.2Ir0.6Ox showed the largest number of peaks
in the chromatograms, as well as the largest number of products at lower reten-
tion times. This indicates overall more degradation products, but also a higher
degree of low molecular weight compounds. This electrode therefore is particu-
larly efficient for depolymerization, which is in agreement with the higher yields
seen for this coated MMO electrode.
However, low molecular weight cracking products only constituted a small
amount of all detectable degradation products. Furthermore, GC-MS results of
Ru0.2Ti0.2Ir0.6Ox exhibited far fewer low molecular weight cracking products than
Ru0.2Mn0.2Ir0.6Ox, indicating that the degradation process preferably formed higher
molecular weight degradation products. The detection of these higher molecu-
lar weight fractions is difficult or impossible via GC-MS, necessitating the use of
atmospheric pressure ionization techniques, which allow soft ionization of much
larger, non-volatile, polar and ionic molecules.
2.5 High resolution mass spectrometry
HRMS studies on complex electrochemically degraded lignin have previously
revealed the presence of a large number of higher molecular weight degrada-
tion products [35–37]. The combination of atmospheric pressure photoioniza-
tion (APPI) and high resolution mass spectrometry using two-dimensional mass
defect filtering (MDF) data simplification strategies have been shown to readily
visualize higher mass oligomers and allow the classification of measured m/z
features into common lignin-related compound classes. Plotting Kendrick mass
defects (KMD) for [CH2], after transformation to the Kendrick m/z scale (KM)
(CH2ꢀ=ꢀ14.00000 u) versus KMD [C7H7O2] (C7H7O2ꢀ=ꢀ123.00000 u) readily revealed
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
ꢂ13
Fig. 5:ꢀ2D mass defect plots of the degradation products obtained from the different ternary
mixed metal oxide electrodes.
structural correlations between the measured m/z features. A more detailed
description of the 2D mass defect plots has been previously reported [35–37]. The
2D mass defect plots for each coated MMO electrode system also illustrated sig-
nificant differences of the compound distribution after decomposition (Figure 5).
The electrochemical conversion of lignin using Ru0.2Mn0.2Ir0.6Ox coated MMO
electrodes resulted in the primary formation of products with low KMD [CH2]
values (≤300). Such low KMD [CH2] values are usually seen for low molecular
weight products, indicating that the Ru0.2Mn0.2Ir0.6Ox electrode system primarily
forms low molecular weight degradation products, consistent with the results
from GC-MS (vide supra). The mass defect plots for the Ru0.2Pd0.2Ir0.6Ox and the
Ru0.2V0.2Ir0.6Ox electrode systems show fewer products compared to Ru0.2Mn0.2Ir0.6Ox,
indicating less efficient lignin degradation. We speculate that Ru0.2Pd0.2Ir0.6Ox
and Ru0.2V0.2Ir0.6Ox were only able to cleave specific chemical bonds of the lignin
polymer, whereas Ru0.2Mn0.2Ir0.6Ox was able to cover a wider range of bond species.
The composition of the degradation products for the Ru0.2Ti0.2Ir0.6Ox electrode
system contained, as expected, higher molecular weight degradation products
(KMD [CH2]ꢀ≥ꢀ350). Table 4 summarizes the percent distributions of substance
classes. Compound classes were assigned as described in [35]. In agreement with
the yields and GC-MS results, Ru0.2Mn0.2Ir0.6Ox coated electrodes provided higher
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
14ꢀ
Tab. 4:ꢀPercent distributions of substance classes for the degradation products after electro-
chemical treatment with the different coated MMO electrodes.
Substance class
ꢁ
Ru0.2Mn0.2Ir0.6Oxꢁ
Ru0.2Pd0.2Ir0.6Oxꢁ Ru0.2V0.2Ir0.6Oxꢁ
Ru0.2Ti0.2Ir0.6Ox
Lignins
ꢃ
2.3ꢃ
1.6ꢃ
20.5ꢃ
6.7ꢃ
3.6ꢃ
23.3ꢃ
42.0ꢃ
100ꢃ
0ꢃ
0ꢃ
9.3ꢃ
0ꢃ
9.8ꢃ
14.2ꢃ
66.7ꢃ
100ꢃ
0ꢃ
1.1ꢃ
0
2.0
4.0
0
Unsaturated unknownsꢃ
Alkylphenolics
Saturated unknowns
Cycloalkanes
Not classifiable
>450 g/mol
ꢃ
ꢃ
ꢃ
ꢃ
ꢃ
ꢃ
5.5ꢃ
1.2ꢃ
0.8ꢃ
0
6.6ꢃ
4.0
90
100
84.8ꢃ
100ꢃ
amounts of low molecular weight, with approximately 60% having molecular
weights <450 g/mol. The other electrodes generated predominantly products with
molecular weights >450 g/mol. In particular, alkylphenolics reached the highest
values for the Ru0.2Mn0.2Ir0.6Ox cleavage products, which therefore are important as
a possible source for fine chemicals.
The HRMS results were consistent with the cracking yields as a function of
the applied electrodes, following the trend of electrode activity for Ru0.2M0.2Ir0.6Ox-
type electrodes in the order Mꢀ=ꢀMnꢀ>ꢀPdꢀ>ꢀVꢀ>ꢀTi.
2.6 Recovery of the ionic liquid
To provide a more ecologically friendly process for the usage of lignin, the recov-
ery of the solvent was highly desirable. With the applied protocol, the recovery
of [Et3NH][MeSO3] and separation of the cleavage products and residual treated
lignin was simple. The hydrophilic IL could be dissolved in the aqueous phase,
while the organic phase showed no detectable amounts of [Et3NH][MeSO3] by
HRMS. After drying the IL, 1H-NMR spectra were recorded to ensure that no impu-
rities or degraded species were present in the recovered IL. The proton NMR
spectra of the IL (a) before dissolving lignin and electro-oxidation and (b) after
recovery are shown in Figure 6.
No visible contaminants were detected in the recovered IL after electrochemi-
cal oxidation reaction. Recovery rates ranged from 98.5 to 99.2% (Table 1), which
was nearly quantitative and showed the electrochemical stability of the IL in the
performed experiments. The purity and quantitiy of the recovered IL demon-
strated the efficient separation and facile recovery, which is important for the
use in potential technical applications. As ILs are rather expensive solvents, the
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
ꢂ15
Fig. 6:ꢀNMR spectra of (a) pure [Et3NH][MeSO3] before the dissolution of lignin and (b) recov-
ered [Et3NH][MeSO3] after electrolysis and work-up.
facile recovery makes the process economically and ecologically interesting, in
comparison to other approaches.
3 Conclusion
Electrochemical oxidation of lignin in ionic liquids using mixed metal oxide
(MMO) electrodes could potentially be a sustainable approach for the future
generation of high value aromatic compounds as demonstrated in this fun-
damental study. The use of the IL [Et3NH][MeSO3] as solvent for lignin might
offer a number of synergistic effects that increases the ecological and economic
benefits of this approach. The ionic liquid is able to dissolve large amounts of
different lignins, acts as solvent-free electrolyte without the need of additives,
and exhibits high chemical and electrochemical stability. Furthermore, it can
be separated and regenerated from the reaction products by facile aqueous
extraction. The rather inexpensive halide-free IL could also be applied to larger
scale processes or for other lignin depolymerization approaches. Nevertheless,
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
16ꢀ
upscaling of such a process is not straightforward and hampered by the high cost
of the incorporated transition metals and large amounts of solvents that need
to be removed and recovered. These problems might be faced using alternative
MMO electrodes based on cheaper metals and advanced separation techniques.
Although all investigated MMO electrodes showed the ability to degenerate
lignin, the ternary oxide Ru0.2Mn0.2Ir0.6Ox exhibited the best performance, as
seen by the yields of the cracking products and the amount of low molecular
weight compounds in GC-MS and HRMS analyses. The electrocatalytic oxida-
tion on the MMO electrodes in general leads to a large number of substituted
aromatic molecules, which could be of interest for the production of chemicals
from this renewable feedstock if the reactions could be tailored for higher effi-
ciency and selectivity. We also showed that it is possible to vary the amount
and composition of the observed cracking products by changing the electrode
composition, which can be brought about by incorporating transition metals
that increase the electro-catalytic activity. This could be of particular interest
for the selective production of specific chemicals, e.g. vanillin. Further investi-
gations on other MMO electrodes that can act as active and selective catalysts
for the electrochemical lignin degradation, with emphasis on their long-term
stability, will be performed in the future. Furthermore we will try to improve
the isolation of the cracked species by testing approaches with solvents that are
more sustainable, safer and more suitable for upscaling than the diethyl ether
used for product separation here: those features would be essential for techni-
cal applications. Also the electrode reactions will be investigated in more detail
to gain more insight in the degradation mechanisms.
4 Experimental
4.1 Materials and methods
Seventy Percent aqueous methanesulfonic acid solution was obtained from Carl
Roth (Karlsruhe, Germany) and used as received. Triethylamine (>99%) from
Sigma-Aldrich (St. Louis, MO, USA) was destilled prior to use. Alkali lignin from
Sigma-Aldrich was dried in oil pump vaccuum for 3 days prior to use. The lignin
solutions in the ionic liquid were prepared by dissolving weighted samples of the
dried lignin powder in pure [Et3NH][MeSO3] at 60ꢀ°C for 1 day. The homogeneity
of the lignin solutions in the IL was checked by visible light microcopy. DMSO-d6
from Sigma-Aldrich was dried over molecular sieves. The actylacetonate salts
Ru(acac)3, Mn(acac)3 (n. a.), Ir(acac)3 (99.95%), VO(acac)2 (>95%) along with PdCl2
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
ꢂ17
(98%) and Ti(OiPr)4 (>96%) were obtained from Chempur (Karlsruhe, Germany).
NMR measurements were performed on an AV II 400 (Bruker, Billerica, MA, USA)
and referred to the residual solvent signal. DSC measurements were performed
using sealed aluminum cruicibles on a STARe DSC (Mettler Toledo, Gießen,
Germany) equipped with liquid nitrogen cooling. The temperature program
started at 25ꢀ°C with a dynamic step heating at 5ꢀ°C/min to 100ꢀ°C, and an iso-
thermal step at 100ꢀ°C for 10 min. Subsequently, cooling at −1ꢀ°C/min to −120ꢀ°C,
followed by an isothermal step for 10 min was applied. Finally, the samples were
heated to 100ꢀ°C at a rate of 1ꢀ°C/min.
4.2 Synthesis of the ionic liquid triethylammonium methane-
sulfonate [Et3NH][MeSO3]
An aqueous solution of 1.00 eq. methanesulfonic acide was cooled with an ice
bath and neutralized with 1.15 eq. of triethylamine by dropwise addition under
argon. The resulting solution was stirred for 2 h and the excess triethylamine and
majority of water were removed by rotary evaporation. The residual amount of
water was removed by heating the IL to 70ꢀ°C in oil pump vacuum with intensive
stirring for 2 days. Triethylammonium methanesulfonate was obtained as a color-
less, hygroscopic solid in quantitative yield.
1H NMR (400 MHz, DMSO-d6): δ[ppm]ꢀ=ꢀ9.18 (s, 1H, N–H), 3.32–2.85 (m, 6H,
NCH2), 2.38 (s, 3H, SO3–CH3) 1.30 (t, Jꢀ=ꢀ7.3 Hz, 9H, CH2CH3).
13C NMR (101 MHz, DMSO-d6): δ[ppm]ꢀ=ꢀ45.72, 39.74, 8.56.
4.3 Fabrication of the mixed metal oxide coated electrodes
The mixed metal oxide electrodes with different metal compositions for the lignin
electrolysis were prepared by a sol–gel-based coating method [38]. A mesh of tita-
nium wire was first sandblasted and then heated in 10% aqueous oxalilic acid
solution at 70ꢀ°C for 2 h to remove contaminants and obtain a roughened surface.
The titanium mesh was then washed with excess isopropanol and dried in
vacuum. The solution for the coating was prepared by adding dissolved stoichio-
metric amounts of the metal salts in a 1:1 mixture of isopropanol and propionic
acid followed by refluxing under intensive stirring for 40 min. The prepared tita-
nium mesh was dip-coated with the solution of the metal salts, dried in air and
then sintered at 250ꢀ°C for 10 min (oven heating rate 1ꢀ°C/min). The dip-coating
and sintering processes were repeated 10 times and afterwards the coated mesh
was heated for 2 h at 450ꢀ°C (heating rate, 1ꢀ°C/min).
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
18ꢀ
4.4 Electrochemical experiments
Allelectrochemicalexperimentswereperformedindrynitrogenatmosphereusing
a VMP3 potentiostat (BioLogic, Seyssinet-Pariset, France). Before each measure-
ment, a stream of dry nitrogen was bubbled through the solutions to ensure the
absence of oxygen. For cyclic voltammetry measurements, a three-electrode set
up consisting of MMO electrode, platinum counter electrode and silver-quasi-
reference electrode was used. The Ag-quasi-reference electrode was fabricated by
immersing an Ag-wire in a glass tube with porous plug in dried [Et3NH][MeSO3].
The CV-cell was held constant at 40ꢀ°C to prevent the IL from solidification. For the
electrolysis reactions, the same set up was upscaled to 100 mL volume and a mag-
netic stirrer was added. For the electrochemical oxidation, the cell was heated to
70ꢀ°C, to lower the viscosity of the lignin solution and to reduce mass transport
limitations for the depolymerization reaction. The potential of the working elec-
trode was set to 1.8 V for 3 h. Previous studies showed that longer times for the
electrolysis did not further increase the yield of lignin depolymerization.
4.5 Separation of product-mixtures and analytical
characterization
After the electrochemical oxidation, the reaction mixture was cooled to room
temperature and transferred to a beaker. Water (500 mL) was added to dissolve
the IL and precipitate the treated lignin. Afterwards, 300 mL of diethyl ether were
added and the reaction mixture intensively stirred to extract the low molecular
organic compounds. The diethyl ether phase was decanted and the reaction
mixture extracted two further times with 150 mL diethyl ether. The combined
organic phases were dried over MgSO4 and filtered; subsequently the solvent was
removed by rotary evaporation and the yellow-brown residue was weighted. To
determine the yield of treated lignin, the precipitate was separated by filtering,
washed with water several times and dried under oil pump vacuum for 2 days. The
1
weight was measured and compared with the initially dissolved lignin. H-NMR
was used to check the purity of the regenerated ionic liquid.
GC-MS measurements were performed in duplicate using a Thermo (Bremen,
Germany) DSQ II system with a TR WAX GC column. The temperature program
started at 70ꢀ°C for 1 min, followed by heating to 260ꢀ°C at 10ꢀ°C/min, isothermal
step at 260ꢀ°C for 1 min, and cooling to the initial conditions. For HRMS measure-
ments, a sample of 100 μg was dissolved in 1 mL of methanol/toluene (90:10 v/v).
Direct injection experiments were performed in negative ionization mode on a
SolariX 7 Tesla FTICR mass spectrometer (Bruker, Bremen, Germany), equipped
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢁꢀꢀꢀ
Electrochemical Lignin Degradation in Ionic Liquidsꢂ
ꢂ19
with an atmospheric pressure photoionization (APPI) source and infinity cell.
For each spectrum, 16 transients with estimated resolving power of 280 000 at
m/z 400 were collected and co-added. Mass spectra were externally calibrated
using a commercially available APPI tune mix. Elemental restrictions for chemi-
cal compositions as well as data analyses using Data Analysis 4.2 from Bruker
were adopted from a previous study [35–37].
Acknowledgements: This work was supported by the German Research Founda-
tion (DFG) under grant number HE 2403/19-1/VO 1355/4-1. DAV acknowledges gen-
eral research support by the Alfried Krupp von Bohlen und Halbach-Stiftung. We
thank Reiner Wintringer and Verlaine Fossog for technical support.
References
1. B. M. Upton, A. M. Kasko, Chem. Rev. 116 (2016) 2275.
2. S. Laurichesse, L. Avérous, Prog. Polym. Sci. 39 (2014) 1266.
3. A. J. Ragauskas, G. T. Beckham, M. J. Biddy, R. Chandra, F. Chen, M. F. Davis, B. H. Davison,
R. A. Dixon, P. Gilna, M. Keller, P. Langan, A. K. Naskar, J. N. Saddler, T. J. Tschaplinski,
G. A. Tuskan, C. E. Wyman, Science 344 (2014) 1246843.
4. C. Li, X. Zhao, A. Wang, G. W. Huber, T. Zhang, Chem. Rev. 115 (2015) 11559.
5. V. L. Pardini, C. Z. Smith, J. H. P. Utley, R. R. Vargas, H. Viertler, J. Org. Chem. 56 (1991) 7305.
6. K. Stärk, N. Taccardi, A. Bösmann, P. Wasserscheid, ChemSusChem. 3 (2010) 719.
7. J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius, B. M. Weckhuysen, Chem. Rev. 110 (2010)
3552.
8. S.-H. Li, S. Liu, J. C. Colmenares, Y.-J. Xu, Green Chem. 18 (2016) 594.
9. S. Jia, B. J. Cox, X. Guo, Z. C. Zhang, J. G. Ekerdt, ChemSusChem. 3 (2010) 1078.
10. D. Glas, C. Van Doorslaer, D. Depuydt, F. Liebner, T. Rosenau, K. Binnemans, D. E. De Vos,
J. Chem. Technol. Biotechnol. 90 (2015) 1821.
11. W. E. S. Hart, J. B. Harper, L. Aldous, Green Chem. 17 (2015) 214.
12. N. Sun, M. Rahman, Y. Qin, M. L. Maxim, H. Rodríguez, R. D. Rogers, Green Chem. 11 (2009)
646.
13. A. Brandt, J. Gräsvik, J. P. Hallett, T. Welton, Green Chem. 15 (2013) 550.
14. E. Reichert, R. Wintringer, D. A. Volmer, R. Hempelmann, Phys. Chem. Chem. Phys. 14 (2012)
5214.
15. T. K. F. Dier, D. Rauber, D. Durneata, R. Hempelmann, D. A. Volmer, Sci. Rep. 7 (2017) 5041.
16. S. Stiefel, A. Schmitz, J. Peters, D. Di Marino, M. Wessling, Green Chem. 18 (2016) 4999.
17. O. Movil-Cabrera, A. Rodriguez-Silva, C. Arroyo-Torres, J. A. Staser, Biomass Bioenergy 88
(2016) 89.
18. S. K. Hanson, R. T. Baker, Acc. Chem. Res. 48 (2015) 2037.
19. S. Trasatti, Electrochim. Acta. 45 (2000) 2377.
20. H. Zhu, L. Wang, Y. Chen, G. Li, H. Li, Y. Tang, P. Wan, RSC Adv. 4 (2014) 29917.
21. E. C. Achinivu, R. M. Howard, G. Li, H. Gracz, W. A. Henderson, Green Chem. 16 (2014) 1114.
22. X. Lu, G. Burrell, F. Separovic, C. Zhao, J. Phys. Chem. B. 116 (2012) 9160.
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM

ꢀꢀꢀꢁ
ꢀD. Rauber et al.
20ꢀ
23. T. L. Greaves, C. J. Drummond, Chem. Rev. 115 (2015) 11379.
24. M. M. Hossain, L. Aldous, Aust. J. Chem. 65 (2012) 1465.
25. D. J. G. P. van Osch, L. J. B. M. Kollau, A. van den Bruinhorst, S. Asikainen, M. A. A. Rocha,
M. C. Kroon, Phys. Chem. Chem. Phys. 19 (2017) 2636.
26. M. V. Galkin, J. S. M. Samec, ChemSusChem. 9 (2016) 1544.
27. R. Rinaldi, R. Jastrzebski, M. T. Clough, J. Ralph, M. Kennema, P. C. A. Bruijnincx,
B. M. Weckhuysen, Angew. Chemie Int. Ed. 55 (2016) 8164.
28. G. Chatel, R. D. Rogers, ACS Sustain. Chem. Eng. 2 (2014) 322.
29. T. K. F. Dier, D. Rauber, J. Jauch, R. Hempelmann, D. A. Volmer, ChemistrySelect. 2
(2017) 779.
30. D. Schmitt, C. Regenbrecht, M. Hartmer, F. Stecker, S. R. Waldvogel, Beilstein J. Org. Chem.
11 (2015) 473.
31. D. Schmitt, C. Regenbrecht, M. Schubert, D. Schollmeyer, S. R. Waldvogel, Holzforschung 71
(2017) 35.
32. D. Schmitt, N. Beiser, C. Regenbrecht, M. Zirbes, S. R. Waldvogel, Sep. Purif. Technol. 181
(2017) 8.
33. R. Behling, S. Valange, G. Chatel, Green Chem. 18 (2016) 1839.
34. G. F. De Gregorio, C. C. Weber, J. Gräsvik, T. Welton, A. Brandt, J. P. Hallett, Green Chem. 18
(2016) 5456.
35. T. K. F. Dier, K. Egele, V. Fossog, R. Hempelmann, D. A. Volmer, Anal. Chem. 88 (2016) 1328.
36. Y. Qi, R. Hempelmann, D. A. Volmer, Anal. Bioanal. Chem. 408 (2016) 4835.
37. Y. Qi, R. Hempelmann, D. A. Volmer, Anal. Bioanal. Chem. 408 (2016) 8203.
38. R. Chen, V. Trieu, H. Natter, K. Stöwe, W. F. Maier, R. Hempelmann, A. Bulan, J. Kintrup,
R. Weber, Chem. Mater. 22 (2010) 6215.
Brought to you by | University of Virginia
Authenticated
Download Date | 10/9/17 10:24 PM