Study of the reactivity of lignin model compounds to fluorobenzylation using 13C and 19F NMR: Application to lignin phenolic hydroxyl group quantification by 19F NMR

Article abstract of DOI:10.3390/molecules25143211

Lignin is an aromatic biopolymer derived from lignocellulosic biomass. Providing a comprehensive structural analysis of lignin is the primary motivation for the quantification of various functional groups, with a view to valorizing lignin in a wide range of applications. This study investigated the lignin fluorobenzylation reaction and performed a subsequent 19F-NMR analysis to quantify hydroxyl groups, based on a work developed two decades ago by Barrelle et al. The objectives were to check the assignments proposed in this previous study and to examine the reactivity of various types of lignin hydroxyls with the derivatization agent. Selected lignin model compounds containing phenolic and aliphatic hydroxyls were subjected to the fluorobenzylation reaction, and the obtained reaction medium was analyzed by 13C and 19F NMR spectroscopy. The model compound results showed that phenolic hydroxyls were totally derivatized, whereas aliphatic hydroxyls underwent minimal conversion. They also confirmed that 19F NMR chemical shifts from -115 ppm to -117.3 ppm corresponded to phenolic groups. Then, a F NMR analysis was successfully applied to Organosolv commercial lignin after fluorobenzylation in order to quantify its phenolic group content; the values were found to be in the range of the reported values using other analytical techniques after lignin acetylation.

Full text of DOI:10.3390/molecules25143211

molecules
Article
Study of the Reactivity of Lignin Model Compounds
1
3
19
to Fluorobenzylation Using C and F NMR:
Application to Lignin Phenolic Hydroxyl Group
1
9
Quantification by F NMR
Esakkiammal Sudha Esakkimuthu *, Nathalie Marlin, Marie-Christine Brochier-Salon
and Gérard Mortha
Univ. Grenoble Alpes, CNRS, Grenoble INP, Institute of Engineering Univ. Grenoble Alpes, LGP2,
F-38000 Grenoble, France; nathalie.marlin@grenoble-inp.fr (N.M.);
marie-christine.brochier-salon@lgp2.grenoble-inp.fr (M.-C.B.-S.); gerard.mortha@grenoble-inp.fr (G.M.)
*
Correspondence: esakkimuthu.sudha@gmail.com
ꢀ
ꢁꢂꢀꢃꢄꢅꢆꢇ
ꢀ
ꢁꢂꢃꢄꢅꢆ
Received: 26 May 2020; Accepted: 11 July 2020; Published: 14 July 2020
Abstract: Lignin is an aromatic biopolymer derived from lignocellulosic biomass. Providing a
comprehensive structural analysis of lignin is the primary motivation for the quantification of various
functional groups, with a view to valorizing lignin in a wide range of applications. This study
19
investigated the lignin fluorobenzylation reaction and performed a subsequent F-NMR analysis to
quantify hydroxyl groups, based on a work developed two decades ago by Barrelle et al. The objectives
were to check the assignments proposed in this previous study and to examine the reactivity of various
types of lignin hydroxyls with the derivatization agent. Selected lignin model compounds containing
phenolic and aliphatic hydroxyls were subjected to the fluorobenzylation reaction, and the obtained
reaction medium was analyzed by 13C and F NMR spectroscopy. The model compound results
showed that phenolic hydroxyls were totally derivatized, whereas aliphatic hydroxyls underwent
19
1
9
minimal conversion. They also confirmed that F NMR chemical shifts from
−
115 ppm to 117.3 ppm
−
corresponded to phenolic groups. Then, a 19F NMR analysis was successfully applied to Organosolv
commercial lignin after fluorobenzylation in order to quantify its phenolic group content; the
values were found to be in the range of the reported values using other analytical techniques after
lignin acetylation.
Keywords: lignin model compounds; organosolv lignin; 13C-NMR; 19F-NMR; fluorobenzylation;
etherification; hydroxyl group quantification
1
. Introduction
Lignin is an aromatic biomacromolecule which is highly branched and widely available, with a
large variety of functional groups such as hydroxyl, methoxyl, carbonyl and carboxyl [
1]. In particular,
the quantification of the hydroxyl group (aliphatic and phenolic) content in the lignin molecule is
of primary importance in analyses of reactivity, solubility and stability [ ], which are relevant to
2
applications such as additives, coatings, adhesives and polymer materials [3,4]. However, some of the
obstacles to the use of lignins as polymeric materials are heating deformation and phase separation.
These issues were successfully overcome by chemical modification of the hydroxyl groups, mainly via
esterification and etherification [5–8].
Classically, phenolic functional groups can be quantified after lignin acetylation followed by
aminolysis (de-acetylation) [9] using gas chromatography. However, this method has some limitations,
e.g., incomplete and time-consuming reaction, and interference by aliphatic hydroxyls in lignin
Molecules 2020, 25, 3211; doi:10.3390/molecules25143211
www.mdpi.com/journal/molecules

Molecules 2020, 25, 3211
2 of 15
and sugar impurities which lead to a deacetylation process; hence, the true quantification of free
phenolic groups is questionable [10]. At present, modern NMR spectroscopic techniques [11 14
such as 13C, F, P, and 2D H- C Heteronuclear Single-Quantum Coherence (HSQC) are the most
popular methods in the literature [15 17] for the quantification of lignin functionalities. However,
direct phenolic group quantification using C-NMR is relatively hard due to broad C-OH
–
]
19
31
1
13
–
13
13
C
chemical shifts and their overlap with other lignin signals. In such cases, acetylation derivatization
is mainly performed for the quantification of hydroxyl groups. Moreover, due to the low sensitivity
of the 13C nucleus, long acquisition times (i.e., more than 12 h), high amounts of derivatized
lignin (minimum of 300 mg) and advanced expertise are required for quantitative analysis [18
,19].
On the other hand, 31P-NMR spectroscopic analysis [16
,
20 25], developed by Argyropoulos and
–
collaborators to quantify phenolic, aliphatic hydroxyl groups and carboxyl groups of lignin in a
single analysis, has become a preferred method. Several phosphitylating agents have been used such
as 2-chloro-1,3,2-dioxaphospholane [20
,21] or 2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane
(TMDP), which give better signal separation [26], especially between syringyl phenolic hydroxyls and
aliphatic hydroxyl groups. The latter reagent also offers reduced reactivity, and thus, enhances the
stability of phosphitylated compounds and lignin. Argyropoulos et al. studied the chromophores
in mechanical pulps using trimethyl phosphite derivatives with 31P-NMR [23
,24]. Compared to
acetylation, the phosphitylation reaction is faster, since the reagent is introduced directly in the NMR
tube containing the lignin to be analyzed, thereby providing a clear description of syringyl and
guaiacyl units. However, in the case of using 2-chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane,
no distinction can be made between primary and secondary aliphatic hydroxyl groups [26], and the
hydroxyl group content tends to be slightly underestimated compared to the data obtained by other
analytical techniques [25]. Furthermore, phosphorylated lignin derivatives exhibit poor stability for a
31
longer period of time, and thus require instant acquisition for P-NMR [27].
Two decades ago, Barrelle et al. [28
30] developed 19F-NMR spectroscopic analysis after
–
lignin derivatization using fluorinated compounds to quantify phenolic and aliphatic hydroxyl
groups, as well as carbonyl groups (Figure 1). Fluorobenzylation and fluorobenzoylation methods
were applied to quantify phenolic hydroxyl groups and aliphatic hydroxyl groups, whereas
trifluoromethylphenylhydrazine derivatization enabled carbonyl groups determination [31
–33].
This particular study did not consider the influence of impurities, for instance, the sugar present
19
in the lignin. Furthermore, the analyses were performed in the early 1990s, and modern F-NMR
instruments can be different; therefore, the signal assignments need to be validated. 19F-NMR has some
advantages: the derivative is stable, so that it can be reused for other analyses; the natural abundance
of 100% and the high sensitivity of the 19F nucleus compared to C-NMR; the F-NMR acquisition
time is quite low (less than 30 min for a lignin sample); and quantitative analyses require only a small
amount of lignin derivative (15 mg). In contrast, 13C-NMR and P-NMR require about 100–300 mg
and 30 mg, respectively.
13
19
31
Figure 1. A schematic reaction of lignin fluorobenzylation (Cl-R-F—4-Fluorobenzylchloride,
Lig—Lignin).
The general 19F spectral width is very large (more than 300 ppm). In previous years, since it
was easier to detect small spectral widths (especially due to the frequencies filters), NMR instrument

Molecules 2020, 25, 3211
3 of 15
technology made it possible to work on spectral foldings, leading to incorrect chemical shift assignments,
and sometimes, to reverse shaped signals. Considering recent NMR technological developments, that
19
can now examine the whole spectral window, the present study proposes to re-investigate F-NMR
analysis in the context of a lignin characterization. First, the 19F-NMR chemical shift assignments given
by Barrelle were verified using various lignin model compounds (phenolic and nonphenolic ones;
see Figure 2) after fluorobenzylation (Figure 3). At the same time, the fluorobenzylation reactivities
of different hydroxyl were studied using a combination of 13C and F-NMR analyses performed on
lignin model compounds. Finally, the hydroxyl groups of a commercial lignin, i.e., Organosolv (ORG)
lignin, were quantified using 19F-NMR spectroscopy.
19
Figure 2. Studied model compounds: Vanillin (
Veratryl alcohol (5) and D(+) Cellobiose (6).
1), Acetovanillone (2), Guaiacol (3), Vanillyl alcohol (4),
Figure 3. Reaction of lignin fluorobenzylation.

Molecules 2020, 25, 3211
4 of 15
2
. Results and Discussion
Five lignin model compounds (Figure 2) were selected based on the nature of the hydroxyl groups
which can be found in lignins. Vanillin ( ) and acetovanillone ( ) contain a phenolic hydroxyl group with
a carbonyl function in the position (aldehyde and ketone respectively). Guaiacol (3) is one of the basic
phenolic units of lignin. Vanillyl alcohol ( ) consists of both aliphatic and aromatic hydroxyl groups,
while veratryl alcohol ( ) contains only one aliphatic hydroxyl group. A carbohydrate model, i.e., D(+)
Cellobiose ( ), was also studied, because commercial lignins are usually contaminated with sugars
1
2
α
4
5
6
containing hydroxyl groups (primary and secondary OH) which may also undergo derivatization. In
such cases, lignin 19F-NMR spectra will contain signals belonging to sugar carbohydrate contaminants;
this eventuality will also be examined.
NMR Analysis
First, the derivatization reagent FBC and all the fluorobenzylated model compounds and their
reaction products were quantitatively analyzed by both 19F and C-NMR. The fluorobenzylation
reaction is shown in Figure 3. To study this reaction, the following methodology was adopted:
The freeze-dried reaction products, recovered from the organic phase without any purification, were
13
first analyzed by 19F-NMR. After F signal detection, C-NMR analysis was used to confirm, or
19
13
not, that the F nucleus belongs to the lignin model derivative. This made it possible to elucidate the
structure of the derivatized compounds and detect all possible nonfluoro substituted products. The
n
chemical shifts and the specific hyperfine structures due to J couplings were used to determine
CF
the presence or absence of a F-
sequence indicates the unambiguous assignment of the
Φ
(para-fluoro phenyl ring originating from FBC) structure. The DEPT
0
CH2 carbons (Figure 4), and examination of
their specific chemical shifts makes it possible to ascertain whether fluoro derivatization has occurred.
Chemical shift assignments gave the following results (Table 1). The methylene group of FBC ( H Cl)
and that of N-Bu ( H N) were detected at 45.3 ppm and 57.54 ppm, respectively, in relation to the
nature of the heteroatom (Cl or N). The FBC reagent reacts with water (from moisture) and/or methanol
present in N-Bu reagent) to form FBOH and/or FBOMe. Their corresponding signals were seen in the
range of 62–63 ppm and 72.81 ppm respectively. Finally, the corresponding fluoroderivatized model
C
2
C
2
(
compound CH OR signals were in the range of 69–71 ppm.
2
Figure 4. Structure and number assignment of fluorobenzylated model compounds and lignin.

Molecules 2020, 25, 3211
5 of 15
Table 1. 13C-NMR chemical shifts of -CH2 groups of derivatives presented in Figure 4 in
0
different environments.
δc (in ppm)
45.3
57.54
62–63
69–71
72.81
Methylene Group
-CH2Cl
-CH2N
-CH2OH
-CH2OR
-CH2OMe
Aliphatic OH
FBOH (
FBOMe (b)
Compounds
FBC
NBu
F-derivatized compounds
a)
(a)
FBOH: resulting product from the reaction of FBC with water. (b) FBOMe: resulting product from the reaction of
FBC with MeOH (solvent of NBu).
The reactivity of the fluorobenzylation reaction on lignin model compounds was studied. However,
the obtained reaction products were not purified prior to the 13C-NMR analysis. The reaction medium
contained a mixture of products such as the unreacted starting model compounds, fluorobenzylated
model compounds (F-Compound), remaining pure reagent (FBC) and the byproducts FBOH (produced
during the reaction of FBC with H O) and FBOMe (produced during the reaction of FBC with MeOH).
2
To quantify each component after fluorobenzylation in the reaction medium, the total composition inside
the NMR tube was considered to be 100%. Then, a quantitative 13C-NMR analysis was performed on
the reaction medium which remained perfectly homogeneous throughout the analysis. In this way, the
ratio of each moiety formed during the reaction could be determined (Table 2). The fluorobenzylation
reagent FBC purity was controlled by taking a blank 13C-NMR spectrum. According to the C results,
FBC contained 94.1% pure FBC and 5.9% FBOH. Similarly, a “FBC blank reaction” was carried out
with all the reagents except the model compound. After the reaction, the spectrum indicated the
following reaction mixture composition: 7.9% FBOH and 92.1% NBu reagent. No starting FBC signal
was detected (see Table 2); this clearly shows that the FBC reagent was totally converted into FBOH.
13
Table 2. Mixture ratio (in %) of model compound fluorobenzylation, calculated from 13C-NMR data
(Organic part).
Mixture Composition (%)
FBC
Fluorobenzylation
Conversion (%)
Compounds
Unreacted Starting
Compound
F-Compound
FBOH
FBOMe
NBu
Reagent
1
2
3
4
100
100
100
100
9
-
93.6
81.3
67.6
34.0
7.1
2.7
-
6.4
-
-
-
5.2
3.8
-
-
-
5.2
-
-
8.3
24.5
4.1
-
53.9
Traces
Traces
Traces
74
10.1
5
1 day
71.5
27.2
-
-
-
21.4
7
5
3 days
8
32.8
30.3
6
Organic part
Aqueous part
0
-
-
-
-
-
-
-
-
26
90
-
6
0
10
94.1
-
-
-
FBC blank
-
5.9
FBC reacted
-
7.9
92.1
For lignin and carbohydrate model compounds, the fluorobenzylation conversion was calculated
as follows: Conversion = (F-Compound/(Unreacted pristine compound + F-Compound)) 100.
As seen in Table 2, it is interesting to note that the conversion achieved 100% with the lignin model
compounds ( ), ( ) and ( ), as they contain only phenolic hydroxyl groups. In the case of Vanillyl
alcohol ( ), both phenolic and aliphatic hydroxyl groups were present. The absence of a starting
×
1
2
3
4
model compound shows that its conversion was total (Table 2). Three different compounds were
found in the reaction mixture (Figure 5): FBC, with its methylene type carbon at 45.3 ppm, was the
most predominant product (53.9%); FBOMe, formed from the reaction of methanol with FBC and
its corresponding methyl and methylene group carbon at 57.40 ppm and 72.81 ppm, respectively,
represented the minor part (10%); and finally, the remaining approximately 34% was fluoroderivatized

Molecules 2020, 25, 3211
6 of 15
vanillyl alcohol. From these results, it may be concluded that the fluorobenzylation conversion of
lignin phenolic hydroxyl groups was nearly quantitative, as already seen with previous lignin model
compounds (1, 2 and 3). However, the remaining signal at 62.76 ppm corresponding to the methylene
carbon covalently bonded to the free aliphatic hydroxyl –CH OH group, and the absence of a signal at
2
7
1 ppm corresponding to the –CH OR aliphatic fluoroderivatized group, show that the conversion
2
was due solely to the phenolic hydroxyl (Table 3), with no reaction of the aliphatic hydroxyl group.
This was also confirmed by the integral measurements. The integrals of carbons belonging to the
model compound part displayed the following values: 0.34, 0.31 and 0.29, corresponding to –OCH3
(
55.39 ppm), C (135.75 ppm) and CH (62.76 ppm), respectively. Carbons belonging to the FBC part
1 2α
0
0
displayed the following values: 0.32, 0.32 and 0.77 corresponding to CH (69.33 ppm), C 4 (133.56 ppm)
2
0
0
and C 3 and C 5 (129.87 ppm), respectively. This confirmed unambiguously that there was only one
FBC moiety per model compound. So, the fluorobenzylation took place on the free phenolic group
only, without any reaction on the primary aliphatic hydroxyl function.
Table 3. 13C-NMR Chemical shifts (in ppm) of fluorobenzylated model compounds.
Compounds
1
2
3
4
5
C1
C2
C3
C4
C5
C6
129.8
109.9
149.4
153
130.14
110.4
120.59
112.22
149.23
147.66
113.88
121.25
135.75
110.76
149.05
146.38
113.66
118.45
130.61
110.51
148.65
148.30
111.51
120.04
151.88
148.79
112.27
122.92
112.6
125.8
Model Compounds
Part
5
5
5.38
5.46
OCH3
55.53
55.51
55.44
55.39
C=O
HC=O
CH2 (α)
CH3
-
191.34
-
196.2
-
-
-
-
-
-
-
-
-
62.76
-
71.36
-
-
26.31
69.15
132.74
-
0
C H2
69.28
132.5
69.15
133.44
2.84
129.92
9.2
69.33
133.54
70.33
134.71
0
C 4
4JCF (Hz)
0
0
C 3,C 5
130.2
115.3
130.14
115.29
161.88
129.85
115.15
161.72
129.57
115.40
161.56
FBC Part
3JCF (Hz)
0
0
C 2,C 6
115.15
21.45
161.7
243
2JCF (Hz)
0
C 1
161.98
1JCF (Hz)

Molecules 2020, 25, 3211
7 of 15
Figure 5. 13C-NMR spectra of the mixture issued from the fluorobenzylation of vanillyl alcohol (
4)
in DMSO-d6.
For veratryl alcohol (5), bearing only one aliphatic hydroxyl and no phenolic one, three products
were found (Table 2, Figure 6 and Table 3): unreacted starting veratryl alcohol, with–CH OH and C
2
1
signals respectively at 62.74 ppm and at 135.13 ppm, was recovered in the largest quantity (71.5%);
Next, 21.4% butyl ammonium was found with its characteristic signals at 13.46, 19.22, 23.09 and
57.54 ppm; This last signal, very close to the chemical shift of the FBOMe methyl group (57.39 ppm),
displayed a CH multiplicity (as given with the DEPT experiment in Figure 6). Nevertheless, CH
2
3
multiplicity was expected in the case of FBOMe, confirming the assignment of this signal to the butyl
ammonium moiety. Finally, fluoroderivatized veratryl alcohol was recovered as a minor compound
(7.1%). The existence of compound (5) was confirmed with the characteristic signal of the FBC part,
and the chemical shifts modifications of –CH -OR at 71.36 ppm and of C at 130.61 ppm. Thus, the
2
1
fluorobenzylation conversion of the veratryl alcohol (5), containing only one aliphatic hydroxyl group,
was found to be low, i.e., about 9%, and most of the starting model compound remained unchanged.
The typical results again suggest that aliphatic hydroxyls hardly react with FBC. Moreover, no more
FBC remained after the reaction. Trials were conducted for veratryl compound (5) with an increasing
FBC stoichiometric ratio and a longer reaction time, i.e., up to 3 days, to further study the reactivity of
aliphatic hydroxyls towards fluorobenzylation. It is clearly seen in Table 2 that new conditions did
not affect the fluorobenzylation conversion, that remained below 10%. In addition, a high quantity of
unreacted FBC was observed, and fluorobenzylation took place with a methanol solvent, resulting in
the formation of FBOMe. Therefore, it is evident that the fluorobenzylation was more efficient with
phenolic hydroxyl than with primary aliphatic hydroxyl groups.

Molecules 2020, 25, 3211
8 of 15
Figure 6. 13C-NMR spectra of the mixture resulting from the fluorobenzylation of veratryl alcohol (
5)
in DMSO-d6.
Similarly, in the case of cellobiose (6), the composition of both the aqueous and organic parts
recovered after the fluorobenzylation reaction revealed that no signal was obtained for the fluorinated
product, nor for the starting carbohydrate model compound (in Table 2). Only butyl ammonium (26%)
and FBOH (74%) were detected. The aqueous part results showed that the reaction mixture contained
only the starting cellobiose (10%) and NBu reagent (90%). Considering that the total composition
inside the NMR tube was 100%, the starting cellobiose (
6) remained unchanged after the reaction.
This clearly indicates that hydroxyls of carbohydrates cannot be fluorobenzylated. Therefore, lignin
19
contamination with sugar will not interfere with the F-NMR results for lignin derivatization.
The derivatized model compounds were also analyzed by 1 F-NMR (same NMR tube as for
9
13
C-NMR). According to 19F-NMR experiments, only fluorinated compounds can be detected, and
hence, no signal remained for the either starting material or for NBu. The reaction mixture compositions
expressed in %) showed some differences between the 13C and F-NMR data. From the F-NMR
Tables 4 and 5 and Figure 7), it is clear that all the phenolic hydroxyl groups of compounds ( ), (
) were quantitatively fluorobenzylated, showing signals at 116.36 ppm, 116.49 ppm and
116.89 ppm respectively. Concerning the model compound ( ), containing both aliphatic and phenolic
OH, F-NMR confirmed that the conversion took place only in the phenolic region ( 116.99 ppm),
19
19
(
(
1
2)
and (
−
3
−
−
4
19
−
since no peak was detected in the aliphatic region. The conversion of 98% comes from the derivatization
of the phenolic hydroxyl group alone. The aqueous part was also verified, but no 19F signal was
detected, confirming that only phenols could be fluorobenzylated under the experimental conditions.
For model compound (
the conversion was very low, i.e., around 16% (Table 4). The resulting signal of the fluoroderivatized
compound was detected at 117.50 ppm (Table 5). Even after a longer reaction time (Table 4),
F-compound conversion decreased, and simultaneously, the byproduct quantity increased, confirming
5), containing only one aliphatic hydroxyl group, after a reaction time of one day,
−
13
19
the C-NMR results. In the case of the cellobiose (6) (Table 4), no F signal was detected in either the

Molecules 2020, 25, 3211
9 of 15
aqueous or organic parts. Moreover, it can be seen that the organic part contained 100% of the FBOH
originating from the reaction between FBC and water. This result is consistent with the 13C-NMR data.
Table 4. Conversion and mixture ratio of model compound fluorobenzylation, calculated from
1
9F-NMR data.
Mixture Composition (%)
F-Compound
Compounds
Conversion (%)
F-Compound
FBC Reagent
FBOH
FBOMe
1
100
100
100
98
16
12
0
94.5
86.3
72.3
38.8
89.7
4.9
-
5.5
-
-
-
2
3
6.3
7.4
23.3
-
4.4
4
55.9
1.1
10.3
0.6
100
-
10.2
5 1 day
-
-
5 3 days
6 Organic part
6 Aqueous part
FBC
46.0
48.5
-
-
-
-
-
-
-
-
0
-
-
99.9
-
0.1
100
FBC reacted
-
Table 5. 19F-NMR chemical shifts (in ppm) of fluorobenzylated model compounds.
Compounds δF (ppm)
Nature
Φ-OH + Aldehyde (α)
Φ-OH + Ketone (α)
Φ-OH
1
−116.36
−116.49
−116.89
−116.99
−117.50
-
2
3
4
Φ-OH
5
6
CH2OH
CH2OH
MeOH
FBC (blank)
FBOH
−117.49
−115.85
−118.7
CH3OH
CH2Cl
CH2OH
After studying the fluorobenzylation conversion of lignin hydroxyl groups, the 19F-NMR chemical
shift assignments given by Barelle [28 30] were tested, since working on folding spectra may have
–
induced incorrect assignments. In the present study, the entire spectral window was examined.
The 19F-NMR chemical shifts of all investigated fluoroderivatized model compounds are given in
Table 5, and the spectra are illustrated in Figure 7. The general shape of the signal is in agreement
with the work of Barrelle. Chemical shifts from
group region, whereas primary aliphatic hydroxyl groups yielded a signal between
and 118.5 ppm. Moreover, in the phenol region, phenols with a C=O in the position could be
116.6 ppm, which is again in
−115 ppm to
−
117.3 ppm correspond to the phenol
−
117.3 ppm
−
α
distinguished, since they led to chemical shifts between
agreement with the observations of Barrelle.
−
116.2 ppm to
−
The 13C and 19F-NMR results confirmed that the conversion of the phenolic hydroxyls of (
1), (2),
(3), and (4) was fully achieved, whereas that of veratryl alcohol (5) was poor. Therefore, the results
clearly established that fluorobenzylation was efficient on phenolic hydroxyl groups, but not on
aliphatic hydroxyls, and that no reaction occurred with polysaccharides.

Molecules 2020, 25, 3211
10 of 15
Figure 7. Comparison of 19F-NMR spectra of fluorobenzylated ORG lignin (in blue) and derivatized
model compounds. (Note: Only phenol derivatization for 4).
Based on the model compounds’ reactivity towards fluorobenzylation, lignin derivatization was
tested. A commercial Organosolv lignin sample (ORG) from the CIMV (Compagnie Industrielle de
tale, France) was analyzed using 19F-NMR after fluorobenzylation. The F-NMR
19
la Matière V
ég
é
spectrum is given in Figure 7. Signal assignment was made with the help of the lignin model
compounds’ 19F-NMR spectra. The finest signal at
118.7 ppm was assigned to FBOH, originating
−
from the reaction of FBC with water. After 19F-NMR signal assignment, OH function quantification was
done by a comparison of the integrals with an internal standard, i.e., 2-fluoroacetophenone, exhibiting a
signal at
−
112.86 ppm. The integral of the sharp peak around 115.8 ppm for FBC was subtracted from
−
1
9
the total integral value of the aromatic range. According to F-NMR quantification, the ORG lignin
contained 1.71 mmol of phenolic hydroxyl groups per gram of lignin. The phenolic hydroxyl group
result was in the same order as those obtained using other tested methods (see Table 6), i.e., the fast
method developed in our lab (1.5 mmol/g), classical UV method (1.7 mmol/g), P-NMR (1.3 mmol/g)
and conventional 13C-NMR after lignin acetylation (2 mmol/g). Moreover, the aminolysis method
31
1
overestimated the phenolic hydroxyl group by a quantity of 2.4 mmol/g lignin, whereas H-NMR
underestimated the corresponding values (0.9 mmol/g) [10]. This confirmed that 19F-NMR is a robust
method for phenolic hydroxyl group quantification of the investigated ORG lignin.
Table 6. Comparison of phenolic hydroxyl groups (mmol/g) for Organosolv lignin using 19F-NMR and
other techniques [10].
19F-NMR
.7
Aminolysis UV 1H-NMR 13C-NMR 31P-NMR Fast Method
2.4 1.7 0.9 2.0 1.3 1.5
1
Several NMR techniques were used for lignin hydroxyl group determination. Table 7 provides
19
some information about the fundamental characteristics of the most common NMR techniques ( F,
13
31
C and P-NMR).

Molecules 2020, 25, 3211
11 of 15
Table 7. Summary of distinctive characteristics of 19F, 13C and 31P-NMR.
Characteristics
Natural abundance
Sample quantity
Acquisition time
Derivatization
19F
100%
13C [34]
1.108%
31P [27]
100%
15–20 mg
Up to 30 min
Fluorobenzylation
1 day
100–300 mg
Up to 36 h
Acetylation
1 day
30 mg
30–120 min
Phosphitylation
in-situ reaction
Reaction time
Stability of
derivatized sample
Not stable for a long period;
requires instant acquisition
Good stability
Good stability
Strong influence
(requires high
purity samples)
Strong influence (requires high
purity samples)
Sugar contaminants
No influence
Detailed chemical information
for phenolic hydroxyl groups,
primary and secondary aliphatic
hydroxyl groups,
Provides structural
information only for
phenolic hydroxyl
groups
Provides detailed
structural information;
Severe overlap for high
molecular weight lignin
Structural information
stereo-chemical information
Expensive phosphitylating
reagent (TMDP); however, it can
be synthesized easily by the
procedure described in
Reference [35]
Difficult to determine
side chain carbons in
different lignin
Poor reactive towards
aliphatic hydroxyl
groups
Limitations
substructures
3
. Materials and Methods
3
.1. Model Compounds
Five commercially available lignin model compounds were used: Vanillin (CAS: 121-33-5),
Acetovanillone (CAS: 498-02-2), Guaiacol (CAS: 90-05-1), Vanillyl alcohol (CAS: 498-00-0) and Veratryl
alcohol (CAS: 93-03-8), as well as one model of cellulose: cellobiose (CAS: 528-50-7). All were purchased
from Sigma-Aldrich (Merck, Darmstadt, Germany) (Figure 2).
3
.2. Commercial Lignin
A commercial Organosolv lignin, referred to as ORG in the present study, was purchased from the
CIMV (Compagnie Industrielle de la Mati
wheat straw using the CIMV process (formic acid/acetic acid/water at 185–210 C) and the procedure
è
re V
é
g
é
tale, Marne, France). The lignin was extracted from
◦
can be found in the Reference [36].
3
.3. Chemicals
Tetrabutylammonium hydroxide (1M in methanol), referred to as NBu in the present study,
CH CH CH CH ) N(OH) (CAS: 2052-49-5) and 4-fluorobenzyl chloride, referred to as FBC in the
(
3
2
2
2 4
present study, FC H CH Cl, (CAS: 352-11-4) were purchased from Sigma-Aldrich. Acetonitrile CH CN
6
4
2
3
9
9.95% (CAS: 75-05-8), ethyl acetate, referred to as EtOAc in the present study, CH COOC H 99.5%,
3 2 5
(
(
CAS: 141-78-6), sodium sulfate Na SO , 99%, (CAS: 7757-82-6) and diethyl ether (CH CH ) O, 99.5%,
2 4 3 2 2
CAS: 60-29-7) were purchased from Carl Roth (Karlsruhe, Germany) and sodium chloride NaCl (CAS:
7
647-14-5) and tetrahydrofuran, referred to as THF in the present study, (CH ) O (CAS: 109-99-9) was
2 4
obtained from Acros Organics (Geel, Belgium).
3
.4. Acetylation of the Lignin
The detailed procedure of acetylation can be found in the Reference [10].
3
.5. Fluorobenzylation of Model Compounds
First, 100 mg of model compounds (vanillin: 0.657 mmol; acetovanillone: 0.602 mmol; guaiacol:
.806 mmol; vanillyl alcohol: 0.649 mmol; veratryl alcohol: 0.595 mmol and cellobiose: 0.292 mmol)
0

Molecules 2020, 25, 3211
12 of 15
◦
were dissolved in 1 mL of NBu (3.198 mmol) and stirred for 1 h at 50 C. Then, 10 mL of acetonitrile
was added, followed by 300 mg of 4-fluorobenzyl chloride (2.076 mmol) (FBC), the derivatizing agent.
◦
The reaction mixture was stirred at 50 C for overnight. Next, distilled water (30 mL) and EtOAc
(30 mL) were added to the reaction mixture. The aqueous layer was separated and extracted with
EtOAc (2 30mL). The combined EtOAc layer was washed with distilled H O (2 30 mL) and
×
×
2
saturated sodium chloride solution (30 mL). The extracted EtOAc layer was dried with sodium sulfate,
filtered, evaporated and analyzed without any further purification.
3
.6. Fluorobenzylation of the Lignin
One hundred milligrams (0.5 mmol) of lignin was used for the fluorobenzylation reaction. Lignin
fluorobenzylation was performed using the same experimental conditions as for the model compounds.
Derivatized lignin recovery was performed by precipitation in diethyl ether. During this process,
diethyl ether might cause a loss of smaller molecular weight fragments, and especially impurities.
This organic part formed a viscous precipitate, which was precipitated again in ice distilled water.
This precipitate was filtered through a 0.45 µm PTFE filter, washed with distilled water several times
◦
and oven dried at 50 C [29,30].
3
.7. NMR
Spectroscopic measurements were recorded on a Bruker AVANCE400 spectrometer (Billerica,
13
MA, USA) equipped with a 5 mm BB/19F-1H/d Z-GRD probe operating at 100.612 MHz for C,
76.447 MHz for 19F and 400.130 MHz for H. Data acquisition treatment was done using the LINUX
1
3
TopSpin 3.2 software. Lignin model compound derivatives were dissolved in DMSO-d6 (40–50 mg/0.7
mL) using C F as a reference ( 164.90 ppm/CFCl ). Lignin derivatives were dissolved in DMSO-d6
−
6
6
3
(
15–20 mg/0.7 mL), and quantification was done using 2-Fluoroacetophenone (3 mg) as an internal
standard. The measurements were performed at 298 K. 3.75 mg of Chromium acetylacetonate was
dissolved in 0.075 mL of DMSO and used as relaxing agent in 5mm NMR tubes.
3
.8. 19F-NMR
The Bruker invgate sequence was used. The experiments were conducted with a 1.25 s acquisition
◦
time, 8.76 s relaxation delay and a 30 pulse using a 65 ppm spectral width. For data acquisition,
4k data points were used for model compounds. Prior to Fourier transformation, zero-filling at 64k
was applied, followed by apodization with a 0.3 Hz exponential. Chemical shifts are given relative
to CFCl ( = 0 ppm). The positions of the peaks were referred for C F as an internal reference at
6
δ
3
6 6
−
164.90 ppm. For lignin OH quantification, the experiments were conducted at 298 K, with 4.35 s
◦
acquisition time, 8.76 s relaxation delay and a 30 pulse using a 20 ppm spectral width. For data
acquisition, 64k data points were used. The quantification and chemical shifts of peaks were referenced
with 2-fluoroacetophenone as the internal reference at −112.86 ppm.
3
.9. 13C-NMR
Analyses were performed on derivatized model compounds only (the same tube was used for 19
F
and 13C experiments), at 298 K, using the Bruker invgate sequence. The experiments were conducted
◦
with a 0.648 s acquisition time, 20 s relaxation delay and a 45 pulse using a 250 ppm spectral width.
Proton broad band decoupling was applied only during acquisition time. For data acquisition, 32 k
data points were used. Prior to Fourier transformation, zero-filling at 64 k was applied, followed by
apodization with a 2 Hz exponential. Chemical shifts are given relative to TMS (tetramethylsilane,
δ = 0 ppm). The positions of the peaks were referred to the DMSO signal at 39.5 ppm.

Molecules 2020, 25, 3211
13 of 15
3
.10. 13C DEPT
The Bruker dept sequence was used. The experiments were conducted with a 0.648 s acquisition
◦
time, 3.0 s relaxation delay and a last pulse at 135 to select CH carbons reversed compared to CH and
2
CH , optimized for a 145 Hz coupling constant. (For the labelling atoms, see Figure 4).
3
4
. Conclusions
This study re-investigated the 19F-NMR signal assignment of fluorobenzylated lignin and corrected
the chemical shift values [31] using lignin-based model compounds. Fluorobenzylation reactivity
was also investigated. Complete derivatization was obtained with model compounds containing
only phenolic hydroxyl groups, whereas for model compounds containing both aromatic and
aliphatic hydroxyls, fluorobenzylation was partial for the aliphatic group. In the case of compounds
with only a primary aliphatic hydroxyl group, the derivatization was very slow and incomplete.
Lignin fluorobenzylation is thus fully efficient with phenolic hydroxyl groups, but quite inefficient
with aliphatic hydroxyls. To be successfully applied on lignin for aliphatic OH quantification,
the derivatization conditions still have to be improved. Moreover, no fluorobenzylation was observed
with a carbohydrate model compound (cellobiose), meaning that lignin sugar contamination should
not interfere with the analyses. Fluorobenzylation was also applied on commercial Organosolv lignin.
Based on the obtained results from the model compounds, the Organosolv lignin signal assignment
and chemical shifts were aligned precisely, and the phenolic hydroxyl groups were quantified using
19
F-NMR. The obtained phenolic hydroxyl content value was close to those acquired using other
proven methods [10].
Author Contributions: E.S.E., performed experiments and wrote the paper; M.-C.B.-S., analyzed 13C and 19F-NMR
analysis; M.-C.B.-S., N.M., and G.M., supervised the study and reviewed the manuscript. All authors have read
and agreed to the published version of the manuscript.
Funding: E.S.E would like to thank French Ministerial Fellowship for PhD funding (IMEP2 doctoral school)
(2016–2018).
Acknowledgments: E.S.E thanks Grenoble-INP PAGORA and LGP2 for the constant support and research. LGP2
◦
is part of the LabEx Tec 21 (Investissements d’Avenir—grant agreement n ANR-11-LABX-0030) and PolyNat
◦
Carnot Institute (Investissements d’Avenir—grant agreement n ANR-16-CARN-0025-01).
Conflicts of Interest: The authors declare no conflict of interest.
References
1
.
Li, C.; Zhao, X.; Wang, A.; Huber, G.W.; Zhang, T. Catalytic Transformation of Lignin for the Production of
Chemicals and Fuels. Chem. Rev. 2015, 115, 11559–11624. [CrossRef] [PubMed]
2
.
Baumberger, S.; Abaecherli, A.; Fasching, M.; Gellerstedt, G.; Gosselink, R.; Hortling, B.; Li, J.; Saake, B.;
de Jong, E. Molar mass determination of lignins by size-exclusion chromatography: Towards standardisation
of the method. Holzforschung 2007, 61, 459–468. [CrossRef]
3
.
.
Berghel, J.; Frodeson, S.; Granström, K.; Renström, R.; Ståhl, M.; Nordgren, D.; Tomani, P. The effects of kraft
lignin additives on wood fuel pellet quality, energy use and shelf life. Fuel Process. Technol. 2013, 112, 64–69.
[CrossRef]
4
Calvo-Flores, F.G.; Dobado, J.A.; Isac-GarcÃa, J.; MartÃn-MartÃnez, F.J. Lignin and Lignans as Renewable
Raw Materials: Chemistry, Technology and Applications; John Wiley & Sons: Hoboken, NJ, USA, 2015;
ISBN 978-1-118-68351-4.
5
6
7
.
.
.
Glasser, W.G. Cross-linking Options for Lignins. In Adhesives from Renewable Resources; ACS Symposium
Series; American Chemical Society: Washington, DC, USA, 1989; pp. 43–54, ISBN 978-0-8412-1562-7.
Laurichesse, S.; Avérous, L. Chemical modification of lignins: Towards biobased polymers. Prog. Polym. Sci.
2
014, 39, 1266–1290. [CrossRef]
Kazzaz, A.E.; Feizi, Z.H.; Fatehi, P. Grafting strategies for hydroxy groups of lignin for producing materials.
Green Chem. 2019, 21, 5714–5752. [CrossRef]

Molecules 2020, 25, 3211
14 of 15
8
.
Figueiredo, P.; Lintinen, K.; Hirvonen, J.T.; Kostiainen, M.A.; Santos, H.A. Properties and chemical
modifications of lignin: Towards lignin-based nanomaterials for biomedical applications. Prog. Mater. Sci.
2
018, 93, 233–269. [CrossRef]
9
1
.
Lin, S.Y.; Dence, C.W. Methods in Lignin Chemistry; Springer Science & Business Media: Berlin, Germany, 1992.
0. Serrano Cantador, L.; Esakkimuthu, E.S.; Marlin, N.; Brochier-Salon, M.-C.; Mortha, G.; Bertaud, F. Fast, easy
and economical quantification of lignin phenolic hydroxyl groups. Comparison with classical techniques.
Energy Fuels 2018, 32, 5969–5977. [CrossRef]
1
1
1
1. Zakis, G.F. Functional Analysis of Lignins and Their Derivatives; TAPPI Press: Atlanta, GA, USA, 1994; ISBN
9
78-0-89852-258-7.
2. Faix, O. Fourier Transform Infrared Spectroscopy. In Methods in Lignin Chemistry; Springer Series in Wood
Science; Springer: Berlin/Heidelberg, Germany, 1992; pp. 233–241, ISBN 978-3-642-74067-1.
3. Robert, D.R.; Brunow, G. Quantitative estimation of hydroxyl groups in milled wood lignin from spruce and
in a dehydrogenation polymer from coniferyl alcohol using 13C NMR spectroscopy. Holzforsch. Int. J. Biol.
Chem. Phys. Technol. Wood 1984, 38, 85–90.
1
4. Lundquist, K. Proton 1H NMR Spectroscopy. In Methods in Lignin Chemistry; Springer Series in Wood Science;
Springer: Berlin/Heidelberg, Germany, 1992; pp. 242–249, ISBN 978-3-642-74067-1.
5. Cateto, C.A.; Barreiro, M.F.; Rodrigues, A.E.; Brochier-Salon, M.C.; Thielemans, W.; Belgacem, M.N. Lignins
as macromonomers for polyurethane synthesis: A comparative study on hydroxyl group determination.
J. Appl. Polym. Sci. 2008, 109, 3008–3017. [CrossRef]
1
1
6. Hoareau, W.; Trindade, W.G.; Siegmund, B.; Castellan, A.; Frollini, E. Sugar cane bagasse and curaua lignins
oxidized by chlorine dioxide and reacted with furfuryl alcohol: Characterization and stability. Polym. Degrad.
Stab. 2004, 86, 567–576. [CrossRef]
1
1
1
2
7. Gellerstedt, G.; Robert, D.; Parker, V.; Oivanen, M.; Eberson, L. Quantitativea 13C NMR analysis of Kraft
lignins. ACTA Chem. Scand. 1987, 41b, 541–546. [CrossRef]
8. Balakshin, M.; Capanema, E. On the quantification of lignin hydroxyl groups with 31P and 13C NMR
spectroscopy. J. Wood Chem. Technol. 2015, 35, 220–237. [CrossRef]
9. Xia, Z.; Akim, L.G.; Argyropoulos, D.S. Quantitative 13C NMR analysis of lignins with internal standards.
J. Agric. Food Chem. 2001, 49, 3573–3578. [CrossRef] [PubMed]
0. Argyropoulos, D.S.; Bolker, H.I.; Heitner, C.; Archipov, Y. 31P NMR Spectroscopy in wood chemistry. Part IV.
Lignin models: Spin lattice relaxation times and solvent effects in 31P NMR. Holzforsch. Int. J. Biol. Chem.
Phys. Technol. Wood 1993, 47, 50–56. [CrossRef]
2
2
2
1. Argyropoulos, D.S. Quantitative Phosphorus-31 NMR analysis of six soluble lignins. J. Wood Chem. Technol.
1
994, 14, 65–82. [CrossRef]
2. Zawadzki, M.; Ragauskas, A. N-hydroxy compounds as new internal standards for the 31P-NMR
determination of lignin hydroxy functional groups. Holzforschung 2001, 55, 283–285. [CrossRef]
3. Argyropoulos, D.S.; Heitner, C.; Morin, F.G. P NMR Spectroscopy in wood chemistry-Part III. Solid state 31P
NMR of trimethyl phosphite derivatives of chromophores in mechanical pulp. Holzforsch. Int. J. Biol. Chem.
Phys. Technol. Wood 1992, 46, 211–218. [CrossRef]
2
2
2
4. Argyropoulos, D.S.; Zhang, L. Semiquantitative Determination of quinonoid structures in isolated lignins by
3
1P nuclear magnetic resonance. J. Agric. Food Chem. 1998, 46, 4628–4634. [CrossRef]
5. Berlin, A.; Balakshin, M. Chapter 18-Industrial lignins: Analysis, properties, and applications. In Bioenergy Research:
Advances and Applications; Elsevier: Amsterdam, The Netherlands, 2014; pp. 315–336, ISBN 978-0-444-59561-4.
6. Granata, A.; Argyropoulos, D.S. 2-Chloro-4,4,5,5-tetramethyl-1,3,2-dioxaphospholane, a reagent for the
accurate determination of the uncondensed and condensed phenolic moieties in lignins. J. Agric. Food Chem.
1
995, 43, 1538–1544. [CrossRef]
2
7. Meng, X.; Crestini, C.; Ben, H.; Hao, N.; Pu, Y.; Ragauskas, A.J.; Argyropoulos, D.S. Determination of
hydroxyl groups in biorefinery resources via quantitative 31 P NMR spectroscopy. Nat. Protoc. 2019, 14,
2
627–2647. [CrossRef]
2
8. Barrelle, M.; Fernandes, J.C.; Froment, P.; Lachenal, D. An approach to the determination of functional groups
in oxidized lignins by 19F NMR. J. Wood Chem. Technol. 1992, 12, 413–424. [CrossRef]
2
9. Barrelle, M. A new method for the quantitative 19F NMR spectroscopic analysis of hydroxyl groups in lignins.
Holzforsch. Int. J. Biol. Chem. Phys. Technol. Wood 1993, 47, 261–267. [CrossRef]

Molecules 2020, 25, 3211
15 of 15
3
3
3
3
3
0. Barrelle, M. Improvements in the structural investigation of lignins by 19F NMR spectroscopy. J. Wood Chem.
Technol. 1995, 15, 179–188. [CrossRef]
1. Sevillano, R.M.; Mortha, G.; Barrelle, M.; Lachenal, D. 19F NMR Spectroscopy for the quantitative analysis of
carbonyl groups in lignins. Holzforschung 2001, 55, 286–295. [CrossRef]
2. Huang, F.; Pan, S.; Pu, Y.; Ben, H.; Ragauskas, J.A. 19F NMR spectroscopy for the quantitative analysis of
carbonyl groups in bio-oils. RSC Adv. 2014, 4, 17743–17747. [CrossRef]
3. Constant, S.; Lancefield, C.S.; Weckhuysen, B.M.; Bruijnincx, P.C.A. Quantification and classification of
carbonyls in industrial humins and lignins by 19F NMR. ACS Sustain. Chem. Eng. 2017, 5, 965–972. [CrossRef]
4. Adeosun, S.O.; Gbenebor, O.P. Characterization techniques and quality assessment of lignin and lignin
carbon materials. In Sustainable Lignin for Carbon Fibers: Principles, Techniques, and Applications; Akpan, E.I.,
Adeosun, S.O., Eds.; Springer International Publishing: Cham, Switzerland, 2019; pp. 193–279.
5. Savy, D.; Mazzei, P.; Drosos, M.; Cozzolino, V.; Lama, L.; Piccolo, A. Molecular characterization of extracts
from biorefinery wastes and evaluation of their plant biostimulation. ACS Sustain. Chem. Eng. 2017, 5,
3
3
9
023–9031. [CrossRef]
6. Delmas, G.-H.; Benjelloun-Mlayah, B.; Bigot, Y.L.; Delmas, M. Functionality of wheat straw lignin extracted
in organic acid media. J. Appl. Polym. Sci. 2011, 121, 491–501. [CrossRef]
Sample Availability: Samples of the compounds are not available from the authors.
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).