Cellulose (2022) 29:7163–7176

https://doi.org/10.1007/s10570-022-04718-z

ORIGINAL RESEARCH

Plasma treatment of cellulose: investigation on molecular

changes using spectroscopic methods and chemical
derivatization
Sven Gerullis · Andreas Pfuch · Oliver Beier ·
Björn‑Sten‑Mark Kretzschmar · Mario Beyer · Steffen Fischer

Received: 15 February 2022 / Accepted: 17 June 2022 / Published online: 13 July 2022
© The Author(s), under exclusive licence to Springer Nature B.V. 2022

Abstract One interesting and sustainable method surface of the materials, specific derivatization reac-
for the treatment of a large variety of surfaces is the tions and subsequent analysis of the reaction products
atmospheric plasma technology. The use of plasma formed during derivatization are carried out in order
technologies can lead to physicochemical interactions to distinguish the different states of species that are
between the plasma and the substrate material. Such present. The formation of oxygen-containing, polar
interactions can result in a wide variety of molecu- functionalities on cellulose by plasma treatment could
lar (creation of functional groups, radical formation, be confirmed. Chemical derivatization of cellulose
microscopic (roughness, etching, reduction of wood in connection with subsequent XPS measurements
extractives and macroscopic (wettability, liquid pene- showed the formation of O = C bonds in connection
tration, surface energy) modifications of the substrate with the reduction of OH groups. In addition, the for-
surface. Aim of this study was to take a deeper look mation of carbonate groups on cellulose appears to be
on occurring molecular changes of cellulose by using possible as a treatment effect, for which a correspond-
spectroscopic methods. The treated materials were ing reaction mechanism is discussed.
analyzed by Attenuated Total Reflection – Fourier
Transform Infrared Spectroscopy (ATR-FTIR) and Keywords Atmospheric pressure plasma ·
X-ray Photoelectron Spectroscopy (XPS). The evalu- cellulose · XPS, Chemical derivatization · Ozone
ation of XPS detail-spectra and their separation into
sub-spectra can be used to describe general changes
in the chemical bonding states. For a better differen- Introduction
tiation of the species produced by the plasma at the
Natural cellulose, the main component substance of
all plant cells, is the main product of plant biosyn-
S. Gerullis (*) · A. Pfuch · O. Beier · B.-S. Kretzschmar
INNOVENT E.V. Technologieentwicklung Jena, Jena, thesis and is therefore the most abundant biopolymer
Germany in terms of quantity. It is by far the most important
e-mail: [email protected] industrially used renewable raw material and the
areas of application of cellulose-based raw materials
M. Beyer
Institut Für Holztechnologie Gemeinnützige GmbH, are correspondingly diverse. Cellulose and its deriva-
Dresden, Germany tives are used as technical and textile fibrous materi-
als, as well as technical tools and materials. They are
S. Fischer
mainly used in adhesives, emulsifiers and as thicken-
TU Dresden: Institut Für Holz- Und Pflanzenchemie,
Dresden, Germany ers or additives in food and pharmaceutical products.

Vol.: (0123456789)
13
 7164 Cellulose (2022) 29:7163–7176

Important properties of cellulose such as indus- composite. The improved mechanical properties of
trial processability, wettability, biodegradability and TPS/PHB composites with plasma-treated PHB are
solubility are strongly related to its chemical structure mainly due to the interactions of the two polymers
and the functional groups contained in the structure. which is evidenced in FTIR and by the changes in the
Because cellulose is the main component of plant crystallinity of the XRD caused by the process. Even
cells, cellulose also influences the surface proper- the production of nanoscale cellulose fibers from
ties (wettability and chemical compatibility) of sur- microfibers is possible through a plasma treatment
faces for a large variety of lignocellulosic (e.g. wood) (Fazeli and Simão 2019). In this case, treatment with
materials and can influence their glueability or coat- a low-pressure plasma of microfibers led to ablation
ability. In addition, the chemical surface condition of of nanofibers (by etching the amorphous regions).
such materials plays a decisive role in the production These nanofibers, in turn, were embedded in a ther-
of composite materials (e.g. natural fiber in polymer moplastic starch matrix and were able to improve the
matrix or hybrid material). The wettability is influ- mechanical properties of the composite material.
enced by the roughness of the surface and the poros- A few papers in the literature investigate the
ity of the material (capillary forces) and through, plasma treatment of cellulose with various technolo-
for example, acid–base, electrostatic and hydrogen gies. So far, the focus has been on cellulose treat-
bonding interactions (material surface-liquid interac- ment using low-pressure plasma processes (Hua et al.
tions). (Rowell 2005) Furthermore, the materials to 1997; Fras Zemljič et al. 2009). In this context, Hua
be bonded, painted or mixed are often not chemically et al. have investigated the influence of RF plasmas
compatible with the adhesive, varnish or matrix mate- and the use of oxygen and argon as working gas. An
rial, resulting in insufficient adhesion. In this context, increased oxygen concentration of the cellulose was
there is the possibility of adapting the adhesive or detected by investigations using XPS. In addition, a
joining partner, which often seems time-consuming significant influence of process parameters such as
and ecologically ineffective. In addition, such strate- plasma performance and process pressure could be
gies often involve adjusting the whole volume of the demonstrated (Hua et al. 1997). In addition to low-
materials. pressure plasmas, there are those that can be operated
One versatile method for the treatment of sur- under atmospheric pressure conditions. Investigations
faces is the atmospheric plasma technology. The use carried out with the help of atmospheric pressure
of plasma technologies can lead to physicochemical plasmas are, for example, those of Belgacem et al.
interactions between the plasma and the substrate (1995). Here, an increased oxygen concentration was
material. Such interactions can result in a wide vari- found after treatment with a corona-plasma. In these
ety of molecular (creation of functional groups, investigations, it becomes clear that the plasma sys-
radical formation (Busnel et al. 2010)), microscopic tem used, the operation of the plasma system at low
(roughness, etching, reduction of impurities Rehn and pressure or atmospheric pressure as well as the pro-
Viöl 2003; Avramidis et al. 2016; Nguyen et al. 2017) cess parameters used have a decisive influence on the
and macroscopic (wettability, liquid penetration, treatment effect. The literature also describes that the
surface energy Avramidis et al. 2010; Gerullis et al. initial state of the treated material, regard to purity,
2018)) modifications of the substrate surface. Due to lignin content and processing state, can have an
such changes, the wettability, the bonding strength of influence on the treatment. Furthermore, it is known
gluing or the adhesion of coatings to lignocellulosic (Sivachandiran and Khacef 2016; Reuter et al. 2018)
surface can be increased (Busnel et al. 2010; Riedl that ozone is generated when plasma systems operate
et al. 2014; Wolkenhauer et al. 2009). Such a plasma under normal pressure conditions and in the presence
treatment can also be useful for the production of of oxygen. An interaction of the generated ozone with
composite materials. Florez et al. (2019) discribed the cellulose and an associated influence on properties
preparation of polyhydroxybutyrate (PHB)-contain- such as wettability, structure and surface chemistry is
ing thermoplastic starch (TPS) composites, whereby therefore conceivable.
the PHB components were plasmatreated. The inves- The use of plasma sources working under atmos-
tigations showed that a plasma treatment can have a pheric pressure is still little considered for the treat-
positive impact on the mechanical properties of the ment of cellulose in the literature. The advantage

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7165

of such systems is in particular the renunciation of electrodes consist of 15 pairs of silver strips embed-
expensive vacuum technology and the inline capabil- ded 0.5 mm below the surface of a 96% A ­ l2O3
ity of the devices. However, the treatment results and ceramic. These are operated by a 14 kHz voltage sup-
effects cannot be transferred directly from the vacuum ply (Šimor et al. 2002; Lux et al. 2013). This system
process to the atmospheric process due to the oper- generates a plasma with the dimensions 200 × 80 ­mm2
ating mode and different process parameters (power, and an effective thickness of approximately 0.3 mm
treatment time, process gas). Even the comparability on the ceramic. This leads to an active plasma volume
of different sources of atmospheric pressure plasma of 4.8 c­ m3. The power was kept constant at 400 W,
with each other is often not possible and must be con- which resulted in a power density of approximately
sidered separately from each other. In order to better 83 W ­cm−3. The DCSBD was operated at ambient air.
assess and use the potential of this plasma technology Textile pulp (TP) (Buckeye Steinfurt GmbH, Stein-
for the treatment of lignocellulosic materials, a well- furt, Germany) was used as material. Initially, this
founded knowledge of possible influence and changes material was used as powder, but for better handling
to the materials surface is desirable. Aim of this they were pressed into tablets. The samples were kept
study was to take a deeper look on occurring molecu- at a distance of 0.1 mm from the ceramic electrodes
lar changes of the cellulose due to an atmospheric- during the treatment with the aid of cover glasses. It
pressure plasma treatment using a so called “diffuse is really important to mention that the gap between
coplanar surface barrier discharge” (DCSBD) plasma the ceramic and the substrate can have an influence on
system. To be able to show a possible influence of the treatment. This gap influences, for example, the
the ozone formed in the plasma process, the materi- accessibility of the process gas to the sample surface
als were also treated with ozone, examined, and com- and thus the plasma discharge with regard to reactive
pared in parallel with plasma-treated cellulose. The species. The distance of 0.1 mm was chosen because
plasma- or ozone-treated cellulose were analyzed it represents an optimum. Smaller distances can mean
by Attenuated Total Reflection – Fourier Transform that process gases (air) cannot be sufficiently trans-
Infrared Spectroscopy (ATR-FTIR) and X-ray Pho- ported to the ceramic (weaker to no discharge), but
toelectron Spectroscopy (XPS). The evaluation of at longer distances the treatment intensity decreases.
XPS detail-spectra and their separation into sub- Such investigations were not initiated in the context
spectra can be used to describe general changes in the of this work, but have been discussed in the litera-
chemical bonding states. However, an exact distinc- ture (Odrášková et al. 2008). To achieve a homoge-
tion is complex. For example, cellulose is composed neous treatment, the samples were passed over the
of C–OH, C–O–C, O-C-O and may contain C = O ceramic surface with a velocity of 100 mm ­s−1. The
bonds. The evaluation of newly formed chemical treatment time (cellulose/plasma interaction time)
groups after a plasma treatment becomes even more was varied between 0 and 20 s. For shorter interac-
complex (Gross et al. 2010). These groups show a tion times < 20 s, no significant difference could be
specific chemical reactivity and these properties can found using ATR-FTIR. For the XPS measurements
be used for chemical derivatization for XPS meas- an increasing Oxygen-content and decreasing Car-
urements. In doing so, specific derivatization reac- bon-content was observed with increasing treatment
tions and subsequent analysis of the reaction products durations. In order to be able to compare the methods
formed during derivatization are carried out in order and to present the results more clearly, examinations
to distinguish the different states of species that are on test specimens treated for 20 s are presented in the
present. following. The temperature of the ceramic surface
was measured to be in the range of 70 ± 5 °C using a
contact thermometer.
Experimental section In complementary experiments, an ozone gen-
erator COM-AD-02 (Anseros Klaus Nonnenmacher
In the course of this work, a non-thermal plasma GmbH, Tübingen, Germany) was used to generate
source was used: more precisely, a diffuse copla- the ozone. To generate the ozone, dry oxygen was
nar surface barrier discharge (DCSBD) "RPS400" introduced into the ozone generator via an SEP 100
(Roplass s.r.o., Brno, Czech Republic). The DCSBD oxygen generator (Anseros). The oxygen introduced

Vol.: (0123456789)
13
 7166 Cellulose (2022) 29:7163–7176

is converted into ozone by high-voltage corona dis- trifluoroacetic anhydride (TFAA) container closed.
charge tubes. In parallel, the ozone concentration (in After 30 min of evacuation, the line between the pump
­gO3 ­m−3) was measured with the Ozonomat GM- and the desiccator was closed. Afterwards the line
6000-OEM (Anseros). With the set current and a gas between the TFAA container and the desiccator was
flow of 50 standard liters per hour, an ozone con- opened, whereby the sample chamber was flooded with
centration of 50 ­gO3 ­m−3 results. The samples were gaseous TFAA. The samples to be derivatized were
placed in a glass flask and the gas was passed over exposed to the TFAA atmosphere for a total of 15 min,
them for 3, 6 and 15 min, which corresponds to an after which the line between the TFAA container and
ozone mass of approximately 0.125, 0.25 or 0.625 g, the desiccator was closed. The line between the pump
respectively. and the desiccator was then opened again and the
Cellulose samples were examined by scanning treatment chamber was again evacuated to < 8 mbar.
electron microscopy (SEM) with a Supra 55 VP (Carl This pressure was maintained for an additional 30 min
Zeiss NTS GmbH) using the in-lens detector and an to remove unreacted TFAA and trifluoroacetic acid
operating voltage of 5 kV in order to analyze the mor- formed during the reaction. Polyvincyl alcohol (PVA)
phology and changes due to the plasmatreatment. To was used as OH-group containing reference.
avoid charging effects, the samples were coated with Chemical derivatization of carbonyl groups was car-
carbon before the imaging. ried out using pentafluorophenylhydrazine (PFPH) in
During plasma treatment, there may also be a low the liquid phase (Hua; 1997). For this purpose, the test
temperature input into the cellulose (temperature of specimens to be examined were treated in a saturated,
the plasma source is about 70 °C). To investigate this absolutely ethanolic PFPH solution (20 ml solution in
effect, some samples were dried in an oven at 110 °C a sealed glass vessel) for 2 h at room temperature. The
for one hour and examined within the FTIR studies. samples were then washed 3 times with absolute etha-
FT-IR spectra of cellulose pellets were recorded nol and dried for 24 h under laboratory conditions. Pol-
with an FTIR-6100 (Jasco in the spectral range yvinyl methyl ketone (PVMK) was used as the carbonyl
between 7900 and 350 ­cm−1 using the diamond ATR group reference.
technique. The derivatization of carboxyl groups was carried
The XPS measurements were carried out using out with 2,2,2-trifluoroethanol (TFE) and N, N’-di-tert-
an AXIS ULTRA DLD (Kratos Analytical Ltd. A, butylcarbodiimide (Di-tBuC) following the protocol
Manchester, UK) hemispherical energy analyzer of Chilkoti et al. (1991) at room temperature. For this
with a mono-chromatic Al Kα (1486.6 eV) X-ray purpose, the samples to be derivatized were placed on
source. Overview spectra were recorded in 1.0 eV a glass slide inside a glass tube. TFE (0.9 ml), pyridine
steps with a pass energy of 160 eV and detail spec- (0.4 ml), and Di-tBuC (0.3 ml) were sequentially placed
tra in 0.1 eV steps with a pass energy of 40 eV on a below the slide at 15 min intervals, with the substances
measuring spot of 700 × 300 μm2. A charge neutral- not getting in contact. The glass tube was then closed
izer was used because wood and wood components with a PTFE stopper and stored for 12 h. Polyacrylic
are not electrically conductive. The binding energies acid (PAA) was used as the carboxyl group reference.
were corrected to 284.8 eV (aliphatic carbon) (Bock All chemicals were purchased from Sigma Aldrich
1984; Moulder et al. 1992; Ratner and Castner 2009). and were not further modified. It should also be men-
The spectra were recorded with a take-off angle of tioned that no visible changes in the samples were
Θ = 90°. All spectra were evaluated with the Casa observed. A schematic presentation of the derivatiza-
XPS software. The spectra were fitted with a Shirley tion reactions are shown in Fig. 1.
background and a Gauss-Lorentz model GL (30), i.e.
the sum of Gaussian (70%) and Lorentz curves (30%).
The FWHM was set to be 1.5 eV. Results and discussion
The chemical derivatization of the C–OH was car-
ried out at room temperature based on the protocol SEM
of Gross et al. (2010). The samples were placed in a
glass desiccator at room temperature. First the desic- Figure 2 shows scanning electron microscopic
cator was evacuated to a pressure below 8 mbar with (SEM) investigations on textile pulp test specimens.

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7167

Fig. 1  Schematic represen-

tation of the chemical deri-
vatization reactions; C–OH
(a), C = O (b), C-COOH (c)

Fig. 2  SEM investigation on textile pulp: untreated material in different magnifications (left); 20 s plasma-treated (middle) and 120 s
plasma-treated (right) material in different magnifications
On the one hand untreated materials and on the larger particles in the range of 1 µm can be iden-
other hand materials are shown, which were treated tified on the surface. Additionally to this, a large
for 20 or 120 s with the described DCSBD plasma number of nanoscale particles up to 100 nm can
system. The untreated surfaces appear mostly be seen. The increase in treatment time seems to
smooth. As a result of the plasma treatment, the increase this effect once again. Possibly, there is
roughness of the surfaces increases and numerous an etching effect on the material surface, which

Vol.: (0123456789)
13
 7168 Cellulose (2022) 29:7163–7176

primarily affects the amorphous areas of the treatment of the textile pulp with ozone, however, led
cellulose. only to a slight increase in the spectral range between
1780–1710 ­cm−1 (carbonyl range). DCSBD treat-
ment, led to a significant increase in the carbonyl area
ATR‑FTIR (1780 ­cm−1 – 1710 ­cm−1), which is clearly visible in
the difference spectrum. Furthermore, there is a slight
In Fig. 3, the ATR-FTIR spectra of untreated textile reduction in the OH area and in the "fingerprint" area
pulp, dried at 110 °C, 15 min ozone-treated (the 3 or of the cellulose after the DCSBD treatment, which,
6 min treated materials showed no significant measur- however, cannot be classified as significant. This is
able influence on cellulose) and 20 s DCSBD-treated due on the one hand to the high standard deviation in
are shown. In addition, the standard deviation (SD) of this area, and on the other hand to the fact that drying
the respective spectra is shown using dashed curves. (here the heat treatment at 110 °C for 1 h) also leads
Furthermore, untreated and DCSBD-treated textile to a reduction in these areas. In addition, drying often
pulp are compared in the right-hand area of this illus- leads to hornification, which changes the visibility of
tration and their difference spectrum is shown (spectra the OH groups (e.g. due to acetal formation).
of plasma-treated minus spectra of reference). Look- Similar observations were made by Kolářová et al.
ing at the difference spectrum, it becomes clear that (2013) after an Ar plasma treatment at low pressure.
a significant change in the bands of plasma-treated An increase in the region of 1714 ­cm−1 was attributed
cellulose occurs only in the range between 1780 and to the formation of COOH groups by the plasma treat-
1710 ­cm−1. The bands specific for cellulose can be ment. The formation of these groups could have been
detected in all spectra in the spectral range between caused by ex-situ oxidation, since no oxygen-contain-
4000–2700 ­cm−1 and between 1800–800 ­cm−1 (Adel ing process gases were used in the low-pressure pro-
et al. 2011). The drying treatment led to a reduction cess itself. Such groups (C-COOH) could have been
in all specific cellulose bands, but especially in the formed by dehydrogenation at the C6 carbon and oxi-
range between 3600–3000 ­cm−1 (different OH stretch- dation of the primary alcohols (Vaideki et al. 2008).
ing modes) and at ≈ 1650 ­cm−1 (adsorbed water). The DBD treatment under atmospheric pressure with He

Fig. 3  ATR-FTIR spectra of untreated textile pulp (reference), (left); untreated and plasma-treated textile pulp, including the
dried (for 1 h at 110 °C in oven), plasma and ozone-treated difference spectrum (spectra of plasma-treated minus spectra
textile pulp including the respective standard deviation (SD) of reference) (right)

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7169

as the working gas was performed by Kusano et al. spectra. The additionally used XPS analysis however
(2018) on cellulose nanofibers. In the course of this collects information from a depth of only 2–10 nm.
work, however, no change in the ATR-FTIR spectrum Due to the mode of action of this method (resulting
could be shown, which the authors justified with a photoelectrons only emerge from the uppermost sam-
possibly insufficient surface sensitivity. ple surface), it is therefore particularly suitable for
A 15-min ozone treatment of TP in the present detecting plasma-induced effects.
work only led to a slight increase in the spectral range For example, the XPS survey spectra of an
around 1730 ­cm−1, which may be due to the low reac- untreated and a plasma-treated cellulose surface are
tivity of pure cellulose towards ozone (Pouyet et al. shown in Fig. 4. The peak at about 533 eV is the
2014). Positive effects of ozone treatment on cellulose O1s peak. The peak at about 285 eV is the C1s peak.
with ATR-FTIR were mostly described in the litera- Since cellulose mainly contains oxygen and carbon,
ture when lignin and/or hemicelluloses were present these are the most dominant peaks. As a result of
in the cellulose material (Östenson and Gatenholm the plasma treatment, the intensities of these peaks
2005; Pouyet et al. 2014). change. The corresponding element concentrations
are determined from the peak surfaces. The deter-
mined chemical composition and the oxygen-to-
XPS carbon ratio (O/C-ratio) of the examined textile pulp
(TP) test specimens are shown in Table 1. Untreated
Elemental analysis TP has an O/C-ratio of around 0.64. After a 20 s
DCSBD treatment, the oxygen content increased and
The ATR-FTIR method discussed in the previous sec- the carbon content decreased, resulting in an O/C-
tion has an information depth of a few μm depend- ratio of approximately 0.84. In addition, smaller
ing on the material and the ATR equipment (Grdad- amounts of nitrogen of about 0.8 at.% were found.
olnik 2002). However, since it can be assumed that A slight increase in the oxygen content and a related
a plasma treatment dominantly affects the outer- decrease in the carbon content were also detect-
most surface areas, information from the untreated able for the ozone-treated surface—associated with a
substrate material also contributes to the obtained slight increase in the O/C-ratio.

Fig. 4  XPS survey spectrum of untreated and plasm-treated textile pulp (left); XPS C1s detail spectrum of untreated, plasma and
ozone-treated textile pulp, including difference spectrum (plasma-treated – untreated) (right)

Vol.: (0123456789)
13
 7170 Cellulose (2022) 29:7163–7176

Cellulose ­(C6O5) has a theoretical O/C-ratio of elements by evaluating detail spectra. In this con-
around 0.83. Belgacem et al. (1995) found an O/C- text, the respective C1s detail spectra were evaluated
ratio of 0.66 for untreated cellulose, which is in good in further studies. The evaluation of the C1s spectra
agreement with the O/C-ratio of 0.64 obtained in our was chosen for two reasons: on the one hand, carbon
own work. Depending on the processing of the cel- is available in sufficiently high concentrations in the
lulose, significantly lower and significantly higher sample and can therefore be evaluated, on the other
O/C ratios can also occur: Kolářová et al. (2013) hand, the binding energies to be evaluated in the car-
used untreated cotton and determined an O/C-ratio of bon are sufficiently far apart and can therefore be sep-
approx. 0.2. Bañuls-Ciscar et al. (2016) and Klarhöfer arated more easily from each other. The evaluation of
et al. (2010), however, found values of 0.78 and 0.83, the oxygen peak is not to be favored due to the more
which were closer to the theoretical value for cellu- difficult differentiation of the binding energies. The
lose. These differences can be traced back to different C1s-spectra of untreated and treated TP are shown
sample quality and preparation. in Fig. 4 for comparison. In addition, the difference
As was described above, DCSBD treatment of a spectra of plasma-treated and untreated test speci-
textile pulp led to an increase in the O/C ratio from mens are shown.
0.64 to 0.84, which is higher than the theoretical ratio The C1s detail spectra of TP showed the typical
for cellulose. After a 30 s ­O2 low-pressure plasma shape and distribution of the individual components
treatment of viscose cellulose (Cellulose II), Fras for cellulose. The areas around 286.4 eV (C–O–C,
Zemljič et al. (2009) showed an increase in the O/C- C–OH) and 288 eV (C = O, O-C-O) are particu-
ratio from approx. 0.73 to approx. 0.92. The works larly dominant for this material. In theory, the C1s
listed here therefore show comparable effects to our detail spectrum of cellulose is composed only of the
own work. In the present experiments, ozone treat- ­C2 (composed of C–OH and C–O–C) and C ­ 3 (com-
ment (15 min) of the cellulose showed only a slight posed of O-C-O) peak at approximately 286.5 eV and
increase in the O/C-ratio from 0.64 to 0.67. This is 288 eV, respectively. A peak area ratio of 5 (­C2) to
the case particularly at low temperatures (room tem- 1 ­(C3) would be expected. In reality, however, con-
perature in our own work) (Van Nifterik et al. 1993). stituents of C
­ 1 (C–C, C-H) at 284.8 eV and often ­C4
The increase in the O/C ratio may be due to two (e.g. COOH) at 289.4 eV will always be detected, as
possible causes. On the one hand, impurities pre- was also observed in this work. The main component
sent in the material (e.g. by lignin or contamination of the ­C1 peak is aliphatic carbon (contamination,
from the atmosphere) could have been removed or lignin), while the C
­ 4 peak is often assigned to COOH
oxidized. On the other hand, chemical changes in groups. Although pure cellulose does not have any
the cellulose itself and thus the increase in the oxy- COOH groups, the degradation of cellulose by X-ray
gen content are possible. The last variant will be dis- radiation and the associated formation of aldehyde
cussed further in the following text. groups can lead to the formation of COOH groups
(Bañuls-Ciscar et al. 2016).
General overview C1s detail spectrum After the plasma treatment, changes in the detail
spectrum can be clearly seen. The changes are
In addition to determining the elementary composi- stronger pronounced in the difference spectrum
tion, an XPS analysis offers the possibility of making obtained by subtracting the spectra if the plasma-
statements about the chemical binding states of the treated and untreated surfaces. A closer look on these
changes due to plasma and ozone treatment will be
presented later.
Table 1  Chemical composition of the surfaces of untreated,
plasma and ozone treated textile pulp and lignin test specimens Chemical Derivatization
O [at.%] C [at.%] O/C N [at.%]
The chemical compositions of derivatized (TFAA,
TP 38,8 ± 0,7 60,3 ± 0,4 0,64 ± 0,2 0
PFPH, TFE) and non-derivatized TP are summa-
TP + DCSBD 44,7 ± 0,6 53,4 ± 0,7 0,84 ± 0,02 0,8 ± 0,1
rized in Table 2. Additionally, the chemical com-
TP + Ozone 39,9 ± 0,9 59,8 ± 1,1 0,67 ± 0,03 0
position depending on the treatment (untreated,

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7171

plasma-treated, ozone-treated) is compared in this could be detected using XPS. In addition, 3.9 at.% N
table. In addition to the TP, corresponding reference was found after the derivatization. The presence of
materials with a defined chemical composition or these elements is due to the formation of hydrazones
functional species were examined. Polyvinyl alco- through the chemical reaction of the C = O function-
hol (PVA) was used as the hydroxyl group reference, alities of the PVMK and the hydrazine group of the
polyvinyl methyl ketone (PVMK) as the carbonyl PFPH. If an untreated TP test specimen was derivat-
group reference and polyacrylic acid (PAA) as the ized with PFPH, there was only a slight increase in
carboxyl group reference. In addition, the derivatiza- the F content to 0.5 at.%. PFPH derivatization of a
tion reactions were carried out as control on untreated plasma-treated TP surface significantly increased the
polypropylene (PP) in order to be able to estimate F and N content to 7.7 at.% and 3.0 at.%, respectively.
the effect of the derivatization procedure on surfaces After an ozone treatment with subsequent PFPH deri-
without the species to be investigated. As expected, vatization, it was also possible to find increased F and
Fluorine could not be detected on any of the derivat- N contents. However, they were lower (F: 3.5 at.%,
ized PP surfaces. and N: 1.4 at.%) compared to the plasma treatment.
Chemical derivatization of PVA with TFAA The derivatization for the identification of carboxyl
resulted in a significant increase in the F content from groups was carried out with TFE. PAA having a
0 to 31.2 at.%, which is due to the esterification of the defined carboxyl group concentration was used as ref-
OH groups with TFAA and the associated connec- erence material. After a TFE derivatization of PAA,
tion of F-containing reaction products to the surface. an F content of 29.2 at.% could be measured. The
After a derivatization of TP with TFAA, an F content TFE derivatization of the untreated TP gave a value
of 24.6 at.% was found. An analysis of the plasma- of only 0.3 at.% of F. After the plasma treatment with
treated and TFAA-derivatized sample showed a lower subsequent TFE derivatization, slightly increased F
F content of 11.8 at.% The ozone treatment also had a contents of 1.6 at.% were measured. After an ozone
lower F content of 19.0 at.%. treatment with subsequent TFE derivatization, an
PFPH was used for the derivatization of carbonyl increased F content of 2.3 at.% was determined.
groups. After PFPH derivatization of the carbonyl The derivatization experiments in this work
reference material PVMK, an F content of 10.4 at.% yielded the following findings:

Table 2  Composition of Material Treatment Derivatization C [at.%] O [at.%] N [at.%] F [at.%]

the atomic composition
of non-derivatized and PVA / / 65.2 ± 0.4 33.9 ± 0.9 0 0
derivatized materials
PVA / TFAA 49.8 ± 1.0 18.8 ± 0.5 0 31.2 ± 0.5
PVMK / / 84.1 ± 1.3 15.0 ± 1.2 0 0
PVMK / PFPH 75.1 ± 0.4 11.7 ± 0.1 3.9 ± 0.2 10.4 ± 0.3
PAA / / 65.7 ± 0.2 34.3 ± 0.2 0 0
PAA / TFE 52.8 ± 0.2 17.5 ± 0.1 0.5 ± 0.02 29.2 ± 0.2
TP / / 65.4 ± 0.7 33.4 ± 1.0 0 0
TP / TFAA 48.3 ± 0.8 26.9 ± 0.6 0 24.6 ± 1.2
TP / PFPH 59.6 ± 0.5 39.1 ± 0.3 0 0.5 ± 0.1
TP / TFE 64.1 ± 0.9 35.3 ± 1.1 0 0.3 ± 0.1
TP DCSBD / 53.4 ± 0.6 44.7 ± 0.6 0.9 ± 0.1 0
TP DCSBD TFAA 52.8 ± 1.0 34.6 ± 2.0 0 11.8 ± 2.0
TP DCSBD PFPH 55.8 ± 0.6 33.5 ± 1.2 3.0 ± 0.6 7.7 ± 1.3
TP DCSBD TFE 58.0 ± 0.1 39.9 ± 0.2 0.5 ± 0.1 1.6 ± 0.1
TP Ozone / 59.8 ± 1.1 39.9 ± 1.0 0 0
TP Ozone TFAA 55.5 ± 0.9 23.3 ± 0.4 0 19.0 ± 0.6
TP Ozone PFPH 59.9 ± 1.0 34.1 ± 2.5 1.4 ± 0.5 3.5 ± 1.4
TP Ozone TFE 63.8 ± 0.4 33.0 ± 0.8 0 2.3 ± 0.3

Vol.: (0123456789)
13
 7172 Cellulose (2022) 29:7163–7176

• A reduced OH group concentration could be gen- A slight decrease in the intensity of the ­C1 compo-
erated by the DCSBD treatment. nent, a reduction in the intensity of the ­C2 component
• In contrast, the C = O group concentration (56.6%) and a slight increase in the ­C3 component
increased significantly. (19.0%) compared to the untreated TP were observed.
• The formation of COOH groups by DCSBD treat- In addition, an increase in the ­C4 component from
ment was less pronounced. Therefore, the ­C4 peak approx. 0.9 to 5.0% and a significant increase in the
in the C1s detail-spectrum that occurs after the ­C5 component from 2.1% to 10.3% were observed.
plasma treatment cannot be exclusively attributed The ozone treatment (Fig. 6) showed a comparable,
to the formation of COOH groups (formation of but weaker trend.
groups that cannot be derivatized by TFE: forma- Especially the additional partition of the ­C4 peak
tion of carbonates, esters, lactones has to be con- at 298.3 eV into two separate peaks at 288.7 eV
sidered). and 289.6 eV proves the formation of COOH-type
• An ozone treatment of cellulose produced qualita- (288.7 eV) and other O-containing groups (289.6 eV).
tively similar conditions, but at lower concentra- After the DCSBD treatment, an increase in the com-
tion. ponent at 289.6 eV was found. Such a model was
also applied by Hua et al. (1997), which assigned the
fifth peak (­C5) to the formation of carbonate groups
C1s detail‑spectrum textile pulp caused by a low-pressure ­O2-plasma treatment. An
increase in the ­C4 and ­C5 fractions was also shown
The derivatization experiments showed the general in our work. But at the same time a decrease in the
trend that plasma and ozone treatment lower the ­C3 fractions (O-C-O) at 288.0 eV described by Hua
amount of OH groups, while the portions of C = O et al. (1997) remained almost unchanged in our own
groups and COOH groups increase. Moreover, the investigations. An increase in the ­C4 and ­C5 frac-
position of the associated peaks in the spectrum tions was attributed to a modification of the O-C-O
could be determined by using reference materials components. A possible formation of C = O groups
with defined functional groups. A binding energy (ketones or aldehydes) would also be associated
of 286.4 eV was found for the C–OH component of
PVA, a binding energy of 287.7 eV for the C = O
component of PVMK and a binding energy of 288.7
for the COOH component of PAA. These data as
well as the information from the previous derivatiza-
tion experiments, and literature references were used
to further underpin the C1s spectra with individual
spectra. A system consisting of five sub-spectra was
chosen for this. The maximum binding energy of the
individual peaks is 284.8 eV (­C1), 286.6 eV (­C2),
288.2 eV ­(C3), 288.7 eV (­ C4) and 289.6 eV (­ C5). The
­C5 peak can be assigned to carbonates (Hua et al.
1997).
For the untreated TP (see Fig. 5 for a detailed
spectrum and Fig. 7 for percentage distribution of the
components), the cellulose-specific components were
primarily ­C2 (combination of C–O–C and C–OH)
with around 65.5% and ­C3 (primarily O-C-O) domi-
nant at around 16.6%. The ­C1 component takes up a
share of around 14.6%. Small proportions of ­C4 and
­C5 were also assigned.
The C1s signal of the plasma-treated TP with its Fig. 5  C1s detail spectrum with additional splitting into sub-
associated sub-spectra is shown in Fig. 6 and Fig. 7. spectres (C1-C5) of untreated textile pulp

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7173

Fig. 6  C1s detail spectrum

with additional splitting
into subspectres (C1-C5)
of plasma-treated (left) and
ozone-treated (right) Textile
pulp

Fig. 7  Percentage distribution of the proportions (C1-C5) of the C1s detail spectra of untreated, DCSBD-treated and ozone-treated
textile pulp

with an increase in the ­C3 range. It would therefore One difference is the treatment temperature
be difficult to separate a modification/decrease in the (DCSBD: ≈ 70° C, ozone: ≈ 20° C). If the influ-
O-C-O proportions with the simultaneous forma- ence of temperature on the treatment effect on cel-
tion of ketone or aldehyde groups. Hua et al. (1997) lulose during ozone treatment is discussed, the work
pointed out in their work that the formation of car- of Van Nifterik et el. (1993) should be considered.
bonate fractions in cellulose is only possible through An increase in cellulose degradation or an increase
cleavage of the C ­ 1-C2 carbon atoms of the pyranose in corresponding breakdown products could be dem-
ring, followed by subsequent oxidation. onstrated with an increase in temperature (0 °C to
In summary, the studies on molecular changes in 65 °C). The authors justified the temperature-depend-
cellulose using ATR-FTIR and XPS showed that ent decay rate of ozone and the formation of OH radi-
DCSBD treatment caused significant changes. In cals associated with water. Blechschmidt (2013) also
contrast, ozone treatment caused only minor effects refers to the formation of OH radicals in the tempera-
for TP (cellulose). In this work, plasma treatment ture range between 40 °C to 50 °C of ozone bleach-
showed essential differences compared to the ozone ing. Pulp bleaching with ozone is always carried out
treatment. in an acidic medium. This serves to minimize the

Vol.: (0123456789)
13
 7174 Cellulose (2022) 29:7163–7176

self-decomposition of the ozone and the associated which a corresponding reaction mechanism was dis-
minimization of OH radical formation. Unlike ozone, cussed. Based on the derivatization experiments, a
OH radicals react non-selectively due to their high pentamodal model was selected for the evaluation of
oxidation potential and can cause chemical changes the C1s peaks.
in cellulose. An important reaction is the H abstrac- The investigations on molecular changes of this
tion from the anhydroglucose unit. This results in the wood component showed the high oxidation ability of
oxidation of the corresponding secondary OH group the plasma. A treatment with ozone was compared as
to the keto group. The active species generated in the an alternative pretreatment method, which had lower
plasma are also of crucial importance. In addition oxidative effects.
to ozone, OH radicals could also made an impor- As a result, the ozone formed by the plasma cannot
tant contribution to the treatment effect on wood and be solely responsible for the treatment effect and fur-
wood components, but especially in the oxidation of ther reaction mechanisms must be considered in the
cellulose. In DBD systems that are operated with air, interpretation of the results. The influence of temper-
OH radicals can form under certain conditions (Eli- ature and a possible OH radical generation during the
asson et al. 1994; Gentile and Kushner 1996). The operation of the plasma source was therefore taken
formation of such radicals takes place primarily when into account in the discussion of the results. The fact
water is present in the atmosphere (Čech et al. 2015). that plasma treatments had also clear effects on cel-
A certain amount of moisture in the substrate material lulose could accordingly lead to new applications in
could also have contributed to this. Čech et al. (2015) the processing and finishing of paper, thus expanding
used optical emission spectroscopy (OES) to investi- the range of implementations for these technologies.
gate the influence of the humidity of the plasma gases
used on the intensity of the corresponding OH signals Funding The authors have no financial or proprietary inter-
ests in any material discussed in this article.
from a DCSBD source. With an increase in the water
content in the working gas, an increase in the OH Declaration
signals was shown. Since the DCSBD source used in
our work was operated under atmospheric pressure Conflict of interest The authors have no competing interests
to declare that are relevant to the content of this article.
conditions and thus in the presence of air and water,
the presence of corresponding OH radicals can be
assumed. Various reaction mechanisms, including
the decomposition of water by electrons, radicals or References
metastable particles, can lead to the formation of OH
radicals (Bruggeman and Schram 2010). Adel AM, Abd El-Wahab ZH, Ibrahim AA, Al-Shemy MT
(2011) Characterization of microcrystalline cellulose
prepared from lignocellulosic materials. Part II: Phys-
icochemical Properties. Carbohydr Polym 83:676–687.
Conclusion https://​doi.​org/​10.​1016/j.​carbp​ol.​2010.​08.​039
Avramidis G, Wascher R, Militz H, Viöl W (2016) Impact of
air-plasma treatment at atmospheric pressure on wood and
The formation of oxygen-containing, polar function-
wood extractives. Int Wood Prod J 7:76–79. https://​doi.​
alities on cellulose by plasma and ozone treatment org/​10.​1080/​20426​445.​2016.​11624​29
could be confirmed using ATR-FTIR and XPS meas- Avramidis G, Tebbe B, Nothnick E, Militz H, Viöl W, Wolken-
urements. The data obtained by both methods are in hauer A (2010) Wood veneer modification by atmospheric
pressure plasma treatment for improved absorption char-
good agreement to each other. Furthermore, chemical
acteristics. Proceedings of the fifth European conference
derivatization proved to be an interesting and valu- on wood modification, Latvian State Institute of Wood
able tool for the more precise differentiation between Chemistry 365–372
the individual binding states. Chemical derivatiza- Bañuls-Ciscar J, Abel M, Watts JF (2016) Characterisation of
cellulose and hardwood organosolv lignin reference mate-
tion of cellulose in connection with subsequent XPS
rials by XPS. Surf Sci Spectra 23:1–8. https://​doi.​org/​10.​
measurements showed the formation of O = C bonds 1116/1.​49430​99
in connection with the reduction of OH groups. In Belgacem MN, Czeremuszkin G, Sapieha S, Gandini A (1995)
addition, the formation of carbonate groups on cellu- Surface characterization of cellulose fibres by XPS and
lose appears to be possible as a treatment effect, for

Vol:. (1234567890)
13
Cellulose (2022) 29:7163–7176 7175

inverse gas chromatography. Cellulose 2:145–157. https://​ Polym 7:494–503. https://​doi.​org/​10.​1002/​ppap.​20090​

doi.​org/​10.​1007/​BF008​13015 0142
Blechschmidt J (2013) Taschenbuch der Papiertechnik. Carl Hua ZQ, Sitaru R, Denes F, Young RA (1997) Mechanisms of
Hanser Verlag GmbH Co KG, München oxgen-and argon-RF-plasma-induced surface chemistry of
Bock RH (1984) Methoden der Analytischen Chemie: 2. cellulose. Plasmas Polym 2:199–224. https://​doi.​org/​10.​
Nachweis-und Bestimmungsmethoden. Verlag Chemie 1007/​BF027​66154
Weinheim Klarhöfer L, Viöl W, Maus-Friedrichs W (2010) Electron spec-
Bruggeman P, Schram DC (2010) On OH production in water troscopy on plasma treated lignin and cellulose. Holz-
containing atmospheric pressure plasmas. Plasma Sources forschung 64:331–336. https://​doi.​org/​10.​1515/​hf.​2010.​
Sci Technol 42:045025. https://​doi.​org/​10.​1088/​0963-​ 048
0252/​19/4/​045025 Kolářová K, Vosmanská V, Rimpelová S, Švorčík V (2013)
Busnel F, Blanchard V, Prégent J, Stafford L, Riedl B, Blan- Effect of plasma treatment on cellulose fiber. Cellulose
chet P, Sarkissian A (2010) Modification of sugar maple 20:953–961. https://​doi.​org/​10.​1007/​s10570-​013-​9863-0
(Acer saccharum) and black spruce (Picea mariana) wood Kusano Y, Madsen B, Berglund L, Aitomäki Y, Oksman K
surfaces in a dielectric barrier discharge (DBD) at atmos- (2018) Dielectric barrier discharge plasma treatment
pheric pressure. J Adhes Sci Technol 24:1401–1413. of cellulose nanofibre surfaces. Surf Eng 34:825–831.
https://​doi.​org/​10.​1163/​01694​2410X​501007 https://​doi.​org/​10.​1080/​02670​844.​2017.​13344​11
Čech J, Sťahel P, Dvořáková H, Černák M (2015) Diffuse Lux C, Szalay Z, Beikircher W, Kováčik D, Pulker HK (2013)
Coplanar Surface Barrier Discharge: Influence of Gas Investigation of the plasma effects on wood after activa-
Humidity on Plasma Parameters. Plasma Phys Technol tion by diffuse coplanar surface barrier discharge. Eur J
2:304–307 Wood Wood Prod 71:539–549. https://​doi.​org/​10.​1007/​
Chilkoti A, Ratner BD, Briggs D (1991) Plasma-deposited pol- s00107-​013-​0706-3
ymeric films prepared from carbonyl-containing volatile Moulder JF, Stickle WF, Sobol PE, Bomben KD (1992) Hand-
precursors: XPS chemical derivatization and static SIMS book of X-ray Photoelectron Spectroscopy: a reference
surface characterization and static SIMS surface charac- book of standard spectra for identification and interpreta-
terization. Chem Mater 3:51–61. https://​doi.​org/​10.​1021/​ tion of XPS data. Physical Electronics Divison, Perkin-
cm000​13a016 Elmer Corp, Eden Prairie, p 261
Eliasson B, Egli B, Kogelschatz U (1994) Modelling of die- Nguyen TT, Ji X, Van Ngyen TH, Guo M (2017) Wettabil-
lectric barrier discharge chemistry. Pure Appl Chem ity modification of heat-treated wood (HTW) via cold
66:1275–1286. https://​doi.​org/​10.​1351/​pac19​94660​61275 atmospheric-pressure nitrogen plasma jet (APPJ). Holz-
Fazeli M, Simão RA (2019) Preparation and characterization forschung 72:37–43. https://​doi.​org/​10.​1515/​hf-​2017-​0004
of starch composites with cellulose nanofibers obtained by Odrášková M, Ráhel J, Zahoranová A, Tiňo R, Černák
plasma treatment and ultrasonication. Plasma Processes M (2008) Plasma activation of wood surface by dif-
Polym 16:1800167. https://​doi.​org/​10.​1002/​ppap.​20180​ fuse coplanar surface barrier discharge. Plasma Chem
0167 Plasma Process 28:203–211. https://​doi.​org/​10.​1007/​
Florez JP, Fazeli M, Simão RA (2019) Preparation and char- s11090-​007-​9117-8
acterization of thermoplastic starch composite reinforced Östenson M, Gatenholm P (2005) Improvement of the wet-
by plasma-treated poly (hydroxybutyrate) PHB. Int J Biol ting and absorption properties of lignocellulosic fibers
Macromol 123:609–621. https://​doi.​org/​10.​1016/j.​ijbio​ by means of gas phase ozonation. Langmuir 21:160–165.
mac.​2018.​11.​070 https://​doi.​org/​10.​1021/​la040​066y
Fras Zemljič L, Peršin Z, Stenius P (2009) Improvement of chi- Pouyet F, Chirat C, Potthast A, Lachenal D (2014) Formation
tosan adsorption onto cellulosic fabrics by plasma treat- of carbonyl groups on cellulose during ozone treatment of
ment. Biomacromol 10:1181–1187. https://​doi.​org/​10.​ pulp: Consequences for pulp bleaching. Carbohyd Polym
1021/​bm801​483s 109:85–91. https://​doi.​org/​10.​1016/j.​carbp​ol.​2014.​02.​082
Gentile AC, Kushner MJ (1996) Microstreamer dynamics dur- Ratner BD, Castner DG (2009) Electron Spectoscopy for
ing plasma remediation of NO using atmospheric pressure Chemical Analysis. In: Surface Analyis - The Principal
dielectric barrier discharges. J Appl Phys 79:3877–3885. Techniques. John Wiley and Sons Ltd., pp 374–381
https://​doi.​org/​10.​1063/1.​361805 Rehn P, Viöl W (2003) Dielectric barrier discharge treatments
Gerullis S, Kretzschmar BM, Pfuch A, Beier O, Beyer M, at atmospheric pressure for wood surface modification.
Grünler B (2018) Influence of atmospheric pressure Holz Als Roh-Und Werkstoff 61:145–150. https://​doi.​org/​
plasma jet and diffuse coplanar surface barrier discharge 10.​1007/​s00107-​003-​0369-6
treatments on wood surface properties: A comparative Reuter S, von Woedtke T, Weltmann KD (2018) The kINPen—
study. Plasma Processes Polym 15:1800058. https://​doi.​ a review on physics and chemistry of the atmospheric
org/​10.​1002/​ppap.​20180​0058 pressure plasma jet and its applications. J Phys D Appl
Grdadolnik J (2002) ATR-FTIR spectroscopy: Its advantage Phys 51:233001
and limitations. Acta Chim Slov 49:631–642 Riedl B, Angel C, Prégent J, Blanchet P, Stafford L (2014)
Gross T, Pippig F, Merz B, Merz R, Vohrer U, Mix R, Unger Effect of wood surface modification by atmospheric-
WE (2010) Determination of OH Groups at Plasma Oxi- pressure plasma on waterborne coating adhesion. BioRe-
dised Poly(propylene) by TFAA Chemical Derivatisation sources 9:4908–4923
XPS: An Inter-laboratory Comparison. Plasma Processes Rowell RM (2005) Handbook of Wood Chemistry and Wood
Composites. Taylor & Francis Group, Boca Raton

Vol.: (0123456789)
13
 7176 Cellulose (2022) 29:7163–7176

Šimor M, Ráhel’ J, Vojtek P, Černák M, Brablec A (2002) by anion-exchange chromatography with pulsed ampero-
Atmospheric-pressure diffuse coplanar surface discharge metric detection. J Chromatogr A 640:335–343. https://​
for surface treatments. Appl Phys Lett 81:2716–2718. doi.​org/​10.​1016/​0021-​9673(93)​80199-I
https://​doi.​org/​10.​1063/1.​15131​85 Wolkenhauer A, Avramidis G, Hauswald E, Militz H, Viöl W
Sivachandiran L, Khacef A (2016) In situ and ex situ NO oxi- (2009) Sanding vs. plasma treatment of aged wood: A
dation assisted by sub-microsecond pulsed multi-pin-to- comparison with respect to surface energy. Int J Adhes
plane corona discharge: the effect of pin density. RSC Adv Adhes 29:18–22. https://​doi.​org/​10.​1016/j.​ijadh​adh.​2007.​
6:29983–29995. https://​doi.​org/​10.​1039/​C6RA0​2972H 11.​001
Vaideki K, Jayakumar S, Rajendran R, Thilagavathi G (2008)
Investigation on the effect of RF air plasma and neem Publisher’s note Springer Nature remains neutral with regard
leaf extract treatment on the surface modification and to jurisdictional claims in published maps and institutional
antimicrobial activity of cotton fabric. Appl Surf Sci affiliations.
254:2472–2478. https://​doi.​org/​10.​1016/j.​apsusc.​2007.​
09.​088
Van Nifterik L, Xu J, Laurent JL, Mathieu J, Rakoto C (1993)
Analysis of cellulose and kraft pulp ozonolysis products

Vol:. (1234567890)
13