Rate-limiting steps in bromide-free TEMPO-mediated oxidation of cellulose - Quantification of the N-Oxoammonium cation by iodometric titration and UV-vis spectroscopy

Article abstract of DOI:10.1016/j.apcata.2015.07.024

A iodometric titration method was introduced to study the conversion of 2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO) to the corresponding N-oxoammonium cation (TEMPO⁺) by hypochlorite in the absence and presence of bromide ion. The validity of the titration was verified with UV-vis spectroscopy combined with a multivariate curve resolution (MCR) algorithm to calculate the concentrations and spectral signatures of the pure components (i.e., TEMPO, Cl(+1) and TEMPO⁺). The formation of the oxoammonium cation was successfully followed during the activation of TEMPO by HOCl and HOBr. It was found that HOBr is a more effective activator for TEMPO than HOCl is. Moreover, the importance of a separate activation step for TEMPO with bromide-free TEMPO oxidations could be identified with this titration method. The content of TEMPO⁺ was also monitored during the TEMPO-mediated oxidation of a cellulosic pulp by hypochlorite in the absence and presence of bromide. It was found that the oxidation of the alcoholic groups by TEMPO⁺ was generally the rate-determining step and much slower than the regeneration of TEMPO⁺ through oxidation of the hydroxylamine by HOCl and HOBr. However, at high pH the latter reaction became rate-limiting.

Full text of DOI:10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
Applied Catalysis A: General xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Rate-limiting steps in bromide-free TEMPO-mediated oxidation of
cellulose—Quantification of the N-Oxoammonium cation by
iodometric titration and UV–vis spectroscopy
Timo Pääkkönena,∗, Carlo Bertinettoa, Raili Pönnia, Gopi Krishna Tummalaa,1
,
Markus Nuopponenb, Tapani Vuorinena
a Aalto University, School of Chemical Technology, Department of Forest Products Technology, P.O. Box 16300, 00076 Espoo, Finland
b UPM, Tekniikantie 2 C, 02150 Espoo, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
A iodometric titration method was introduced to study the conversion of 2,2,6,6-tetramethyl-1-
piperidinyloxy (TEMPO) to the corresponding N-oxoammonium cation (TEMPO+) by hypochlorite in
the absence and presence of bromide ion. The validity of the titration was verified with UV–vis spec-
troscopy combined with a multivariate curve resolution (MCR) algorithm to calculate the concentrations
and spectral signatures of the pure components (i.e., TEMPO, Cl(+1) and TEMPO+). The formation of the
oxoammonium cation was successfully followed during the activation of TEMPO by HOCl and HOBr. It
was found that HOBr is a more effective activator for TEMPO than HOCl is. Moreover, the importance of
a separate activation step for TEMPO with bromide-free TEMPO oxidations could be identified with this
titration method. The content of TEMPO+ was also monitored during the TEMPO-mediated oxidation of
a cellulosic pulp by hypochlorite in the absence and presence of bromide. It was found that the oxidation
of the alcoholic groups by TEMPO+ was generally the rate-determining step and much slower than the
regeneration of TEMPO+ through oxidation of the hydroxylamine by HOCl and HOBr. However, at high
pH the latter reaction became rate-limiting.
Received 24 March 2015
Received in revised form 23 June 2015
Accepted 17 July 2015
Available online xxx
Keywords:
Birch pulp
Iodometric titration
Multivariate curve resolution
Oxoammonium cation
TEMPO-mediated oxidation
UV–vis spectrometry
© 2015 Elsevier B.V. All rights reserved.
1. Introduction
lulose production via a TEMPO-mediated oxidation can be reduced
from 700 to 1400 MJ kg−1 (microfibrillated cellulose (MFC) process
The chemistry of catalytic oxidation of alcohols with oxoam-
monium ion in aqueous media was reported decades ago [1–4].
The fast and selective oxidation of primary alcohols to carboxy-
lates via aldehydes by oxoammonium ion was introduced later [5].
More recently, this method was applied to the oxidation of cellu-
lose [6] and to the preparation of nanofibrillated cellulose (NFC)
[7,8]. Altogether, this oxidation has gained a vast research inter-
est [9–21]. The TEMPO-mediated oxidation reduces drastically the
energy consumption during the pulp disintegration to NFC, which
is a transparent gel consisting of individual cellulose microfibrils
and microfibril bundles. The energy consumption of the nanocel-
with a high pressure homogenizer treatment) to less than 7 MJ kg−1
(TEMPO-oxidized NFC) [22]. NFC can be utilized, for example, as
a reinforcing agent in composites [22,23]. Other applications of
NCF include gas-barrier films, electronics, cosmetics, and flame-
resistant materials [22].
The most widely studied TEMPO-oxidation process is based on
the use of the NaBr/TEMPO/NaOCl system. The pH level during this
oxidation is typically 10 or higher [10,12,13,19,22]. The amount of
bromide is commonly higher than the amount of TEMPO (mass ratio
10:1) during the oxidation [7]. Moreover, some TEMPO-oxidation
methods without the application of bromide have been reported
in the literature [20,24]. For example, acid-neutral conditions (pH
3.5–6.8) have been applied to oxidize regenerated cellulose apply-
ing a TEMPO/NaOCl/NaClO2 procedure without the use of bromide
[24]. With this low pH range, high carboxylate contents for the oxi-
dized pulps (4 mmol COOH/g) have been reported after an extended
reaction time of almost 80 h. In addition, an electro-mediated oxi-
dation has been used to oxidize the primary hydroxyl groups of
cellulose to carboxylates without the addition of either NaOCl or
∗
Corresponding author. Tel: +358 405339885.
E-mail addresses: timo.paakkonen@aalto.fi (T. Pääkkönen),
carlo.bertinetto@aalto.fi (C. Bertinetto), raili.ponni@aalto.fi (R. Pönni),
gopi.tummala@angstrom.uu.se (G.K. Tummala), markus.nuopponen@upm.com
(M. Nuopponen), tapani.vuorinen@aalto.fi (T. Vuorinen).
1
Present address: Uppsala University, Division of Nanotechnology and Functional
Materials, Department of Engineering Sciences, Box 534, 75121 Uppsala, Sweden.
http://dx.doi.org/10.1016/j.apcata.2015.07.024
0926-860X/© 2015 Elsevier B.V. All rights reserved.
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
2
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
TEMPO+
X(+1), Cl(+3), amperometric system,
ClO2, Cu(+2)
TEMPO
+ 2OH- + (H)X
+ H2O
2
(H)OX + 2
A
B
C
X(+1)
..
..
Scheme 1. Modified activation mechanism of TEMPO radical by HOX [11].
RCH2OH /
RCHO
NaBr. In this oxidation, TEMPO or a TEMPO derivative is used with
an amperometric system for the oxidation of the pulp. However,
achieving a carboxylate content of 1 mmol g−1 of pulp requires a
48 h oxidation, which indicates an extremely low reaction rate [20].
As a conclusion, the TEMPO-mediated oxidations without bromide
are slower compared to the ones that apply bromide. However, a
process without bromide would be of interest since its presence in
the waste water streams is highly undesired [11].
F
D
E
..
X(+1),
amperometric
system
R
R
R
[O]H
[O]H
[O]H
The role of bromide as a radical TEMPO activator has not yet
been clearly elaborated. The lack of an analysis method for TEMPO+
impedes the study of the catalytic cycle of the TEMPO-mediated
oxidation. On one hand, it has been proposed that the primary
oxidant, e.g., NaOBr when NaOCl is present (Scheme 1), oxidizes
the hydroxyl amine to oxidized TEMPO via a radical intermediate
[12,25]. On the other hand, NaOCl has been proposed to activate
TEMPO at pH 10 prior to the NaOBr/TEMPO oxidation without a rad-
ical TEMPO intermediate [22]. Both NaClO2 and NaOCl have been
proposed as TEMPO activators during the TEMPO oxidation under
the low pH conditions (pH 3.5–6.8) [21,24,26]. HOBr has been pro-
posed to be the activator for TEMPO during the NaBr/TEMPO/NaOCl
oxidation at pH 10.8 where the formation of HOBr in the presence
of ClO− and Br− promotes the conversion of TEMPO to TEMPO+ in
a reaction which is similar to the one shown in Scheme 1 [14]. A
similar reaction was proposed earlier by de Nooy et al. [9].
The pKa values of HOCl (7.5) and HOBr (8.7) define the applicable
pH level during the TEMPO activation. At pH > pKa the hypohalous
acids exist increasingly as hypohalites that are inactive in the cat-
alytic oxidation. Accordingly, the NaBr/TEMPO/NaOCl oxidations
take place under higher pH level than the NaOCl/TEMPO oxida-
tions [11]. The complexity of the system and the number of the
reactions involved are substantial when both HOCl and HOBr are
present in the solution [27]. Interestingly, the NaBr/TEMPO/NaOCl
(pH 10) oxidation of starch without a separate TEMPO activation
step is reported to be three times faster than the NaOCl/TEMPO oxi-
dation [11]. However, the reaction rates of the initial conversion of
TEMPO to TEMPO+ and its catalytic reaction with starch were not
studied separately [11].
X(+1),
amperometric
system
H
G
..
RCHO / RCOOH
Scheme 2. Proposed reaction routes for TEMPO-mediated oxidation of primary and
secondary alcohols.
The TEMPO-mediated oxidation of alcoholic groups has been
suggested to proceed via an A → C → F → H route under alkaline
conditions and via an A → C → E → H route under acidic conditions,
where the formation of a reactive complex (E and F) is generated
by a nucleophilic attack of an alcoholate anion on the nitrogen
atom of TEMPO+ [5,10,12,22,33,34]. Additionally, an A → B → D → H
route has been proposed to occur under alkaline conditions, where
the formation of a reactive complex (D) is generated by a nucle-
ophilic attack of an alcoholate anion on the oxygen atom of TEMPO
[34]. Moreover, hydroxylamine has been proposed to shift to the
protonated hydroxylamine H → G under low pH [12,35]. Further-
more, the conversion of TEMPOH to TEMPO+ (H → C) is proposed
to take place under alkaline conditions with [7,12,36] and with-
out [11,22,26,29,37–39] the radical intermediate by a reaction with
X(+1) or by applying an amperometric system.
Several reaction paths have been proposed for the TEMPO-
mediated oxidation of the primary and secondary hydroxyl groups.
However, some of the elementary reactions, e.g., the conver-
sion of the hydroxylamine (TEMPOH) to TEMPO+ and the initial
conversion of TEMPO to TEMPO+, are lacking plausible reaction
mechanisms. Scheme 2 summarizes the most commonly proposed
reaction routes and mechanisms published so far. The conversion
of TEMPO to TEMPO+ (A → C) has been quite often left without
explanation in the early studies [4,5]. Nevertheless, some routes
for this conversion have been proposed, even though the reaction
mechanisms have not been addressed. TEMPO+ has been proposed
to form during a reaction between CuCl2 and TEMPO [3]. Simi-
larly, the formation of TEMPO+ during a reaction between Br(+1)
and TEMPO [14,18,25,28] as well as between Cl(+1) and TEMPO
[11,16,24,26,29] or Cl2 and TEMPO [30] have been reported in
numerous studies. In addition, the conversion of TEMPO to TEMPO+
has been proposed to take place during a reaction with NaClO2
under neutral and acidic conditions [26] and during a reaction
between ClO2 and TEMPO via a transition complex intermediate
[31,32]. Despite the vast research interest on the TEMPO-mediated
oxidation, the reaction mechanisms for the formation of TEMPO+
still remain unrevealed.
For the present, only few elementary reaction rate constants
for the TEMPO-mediated oxidation process have been determined.
Some kinetic studies have been published with primary and sec-
ondary alcohols. The equilibrium constant (K) for the formation of a
complex (F) with MeO− and TEMPO+ (1.3 × 1013
M
−1) is reported to
be 106 times larger than that for the formation of the corresponding
i-PrO− complex (1.1 × 107 M−1) [34]. In addition, the rate con-
stants for the reactions between aldehydes/alcohols and TEMPO+
under alkaline conditions have been studied [9]. Those reactions
were followed by monitoring the consumption of hypochlorite [9].
However, measuring hypochlorite consumption does not allow dif-
ferentiating between the reaction of TEMPO+ with the alcohol and
the conversion of TEMPOH to TEMPO+. Thus specific quantification
of TEMPO+ would be useful for the determination of the rate con-
stants for the individual reactions in the catalytic cycle. In addition,
it is obvious that the mechanism of the conversion of TEMPO or
TEMPOH to TEMPO+ is not clearly defined due to the rival reaction
routes which have been proposed in the earlier studies. Therefore,
a method for the determination of TEMPO+ would be a useful tool
within this field of research.
In this study, we introduce a iodometric titration method to
quantitatively monitor TEMPO+ during TEMPO-mediated oxida-
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
3
tion of alcohols. The validity of the iodometric titration method
is verified with UV–vis absorption spectroscopy by fitting with the
spectra of the pure components, i.e., TEMPO, Cl(+1) and TEMPO+,
as calculated by a multivariate curve resolution algorithm (MCR).
Moreover, the method with UV–vis spectroscopy and MCR algo-
rithm can be applied separately to quantify both TEMPO+ and
TEMPO. The iodometric titration method [1,2,30,40] is applied to
study a separate activation step of TEMPO prior to the TEMPO-
mediated oxidation. Moreover, the role of HOX (e.g., HOCl and
HOBr) as an activator of TEMPO is studied by the titration method.
Finally, the TEMPO-mediated oxidation with a separate activation
step of TEMPO and the titration method to determine TEMPO+ are
applied during an oxidation of a cellulosic pulp, a process com-
monly applied for the preparation of NFC.
tor was not used for pH adjustment after NaOCl addition due to
buffered conditions.
2.4. Conversion of residual aldehydes to carboxylates with
chlorous acid
TEMPO-oxidized pulp suspension was acidified to pH 3 with HCl
and NaClO2 Was added (10 mM initial concentration). The suspen-
sion was mixed in the Büchi reactor for 2 h at 50 ◦C. Finally, the
pulp was washed with pure water. CED-viscosity of the pulp was
analyzed according to the standard method SCAN-CM 15.99 prior
to the calculation of DP [41].
2.5. Analyzing of carboxylate and aldehyde contents of pulps
2. Experimental
The carboxylate content of the pulps were determined by con-
ductometric titration (SCAN-CM 65:02) using Methrohm 751 GPD
Titrino automatic titrator and Tiamo 1.2.1. software. The aldehyde
contents of the pulps were calculated as the difference in their car-
boxylate contents after and before the post-oxidation with HClO2.
2.1. Materials
Industrially dried fully-bleached birch kraft pulp (Finland),
fully-bleached eucalyptus kraft pulp (Brazil) and eucalyptus pre-
hydrolysis kraft pulp (Brazil) were used as the raw materials for
the TEMPO-mediated oxidation of cellulose. Xylitol (Sigma–Aldrich
(St Louis)) was used as the raw material for the TEMPO-mediated
oxidation of alcohol. TEMPO (Sigma–Aldrich (St Louis)) was used
as a catalyst. A 13% NaOCl solution (Merck (Darmstadt, Germany))
was the primary oxidant in the TEMPO-oxidations. 22 g of ortho-
boric acid (VWR (Leuven, Belgium)) and 1.8 g of NaOH pellets (VWR
(Leuven, Belgium)) were diluted to 2000 ml of distilled water to
prepare a borate buffer (pH 8.3) in situ. 1 M NaOH (Merck (Darm-
stadt, Germany)), 1 M HCl (Merck (Darmstadt, Germany)), Büchi
reactor (volume 1.6 dm3), and Metrohm 718 Stat Titrino titra-
tor with pH adjustment were applied during the pulp oxidations.
UV–vis absorption spectra were measured with a Shimadzu UV-
2550 spectrophotometer (Shimadzu Corporation (Kyoto, Japan)).
Ion-exchanged water was used in pulp washings.
2.6. Iodometric titration of (Cl(+1), Br(+1)) and TEMPO+
The applied titration sequence is based on the method of Wartio-
vaara [42] excluding the analysis of TEMPO+. Wartiovaara describes
a three point titration of ClO2 and hypochlorite at pH 8.3, fur-
ther titration of chlorite at pH below 2, and finally, the titration
of chlorate at pH below 1. Only the alkaline (pH 8.3) titrations to
determine the formed TEMPO+ and Cl(+1) were conducted. The
liberated iodine was titrated against Na2S2O3 using starch as an
indicator. The titration of iodine with sodium thiosulfate is based
on the following reaction (Eq. (1)):
I2 + 2S2O32− → S4O62− + 2I−
(1)
First, 25 ml of a borate buffer (pH 8.3) was added to two sample
containers. Then, 0.5 ml of DMSO, which can be used as a masking
agent for HOCl and HOBr, was added to one of the sample containers
[43–47]. A known amount of the sample solutions together with an
excess amount of KI was added to both of the sample containers.
When bromine was present, the solution required stabilization for
a few minutes prior to the addition of KI, since HOBr is trapped by
DMSO more slowly than HOCl. All samples were titrated against
Na2S2O3 using starch as an indicator. The following reactions (Eqs.
(2) and (3)) occur in the mildly alkaline medium:
2.2. Activation of TEMPO by NaOCl
NaOCl and TEMPO were mixed in a buffer solution (pH 8.3)
at room temperature. The consumption of Cl(+1) and the forma-
tion of TEMPO+ were monitored by iodometric titration. In parallel
experiments absorption spectra of the samples withdrawn were
measured in the UV–vis region with a Shimadzu UV-2550 spec-
trophotometer. Reference spectra of the buffer, 2 mM TEMPO in
the buffer and 4 mM NaOCl were also measured. All solutions were
diluted with the buffer solution in a 1:5 ratio prior to the mea-
surements. The concentrations of Cl(+1), TEMPO and TEMPO+ were
obtained through mathematical analysis of the spectra.
HOX + H+ + 2I− → X− + I2 + H2O
2TEMPO+ + 2I− → 2TEMPO + I2
(2)
(3)
The DMSO containing sample includes only the reaction prod-
uct (iodine) with TEMPO+, since HOCl (or HOBr) is trapped with
DMSO. Iodide reduces TEMPO+ to TEMPO radical [1,2,30,40]. Thus
the thiosulphate consumption corresponds stoichiometrically to
the amount of TEMPO+ in the sample (Eqs. (1) and (3)).
2.3. Oxidations of the cellulose pulps and xylitol
All oxidations were carried out in a Büchi glass reactor (1.6 dm3)
at 25 ◦C. The chemical dosages and the consistency of the pulp
suspension were varied while its volume was 1.2 dm3 in all exper-
iments. Radical TEMPO was mixed with a stoichiometric excess of
NaOCl in water. The pH level of the solution was adjusted to 7.5 with
sulfuric acid. The solution was mixed in a closed vessel until TEMPO
was completely dissolved and converted to TEMPO+ by HOCl. The
pulp and the activated TEMPO solution were mixed and NaOCl was
added to the closed reactor by a pump. After the addition of NaOCl,
pH was kept constant at the target level by adding 1 M NaOH with an
automatic titrator. The oxidation rate was followed by iodometric
titration until all oxidant was consumed.
2.7. UV–vis absorption spectroscopy of TEMPO/NaOCl/buffer
solutions
NaOCl/TEMPO mixtures in a buffer solution (pH 8.3) were pre-
pared and measured with a Shimadzu UV-2550 spectrophotometer
correspondingly to the ones which were applied in the titrations.
2 mM TEMPO and 4 mM NaOCl solutions in the buffer media and
the buffer solution without any added chemicals were measured
as reference samples. All solutions were diluted with a ratio of 1/5
by the buffer solution prior to the measurements.
Xylitol was oxidized correspondingly excluding the adjustment
of oxidation solution pH with borate buffer to 8.3. Automatic titra-
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
4
2.8. Mathematical reconstruction of pure spectral components
from UV–vis data
First, the range of linear response in the UV–vis measure-
ments was identified by looking at the singularities in their loading
factors from principal component analysis (PCA) [48]; the range
46 assumed to be TEMPO, TEMPO+, HOCl/−OCl mixture (which
is referred to as Cl(+1)) and a constant background consisting
of a baseline and absorption by the buffer. A number of vary-
ing components equal to three was also confirmed by PCA: the
first three factors explain over 99% of data variance (for data at
200–400 nm). The background, derived from the spectrum of a
buffer-only solution, was subtracted from all spectra. The millimo-
lar spectra of pure TEMPO and Cl(+1) were taken from spectra of
the respective substance in buffer, after subtracting the background
and dividing for the concentration. The spectrum of TEMPO+ was
reconstructed by an algorithm, inspired by the Multivariate Curve
Resolution–Alternate Least Squares method [49], which recursively
alternates least-squares fitting of the bilinear model D = CS + E and
the application of constraints to the obtained solution. D is the n × w
matrix of UV–vis measurements (n = number of measurements,
w = number of wavechannels), C is the n × p matrix of concen-
trations in mmol/l (p = number of independent components, here
equal to 3), S is the p × w matrix of spectra of pure components
and E accounts for noise. The constraints included non-negativity
of absorbances and concentrations, stoichiometric balances and
monotonic decrease of reactants; they were applied by substitut-
ing any absorption or concentration that violated these constraints
with the closest admissible value. The initial estimate for the spec-
trum of 1 mM TEMPO+ was taken as the spectrum from the latest
point (332 min) in the reaction mixture, subtracted by the back-
ground, divided by 1.8 (i.e., 90% of the initial 1 mM TEMPO) and
again subtracted by the spectrum of 1 mM Cl(+1). The algorithm
was iterated until convergence and the final spectra were used to
obtain the concentrations of the corresponding substances by linear
least-squares fit. All calculations were performed using MATLAB®
version 8.2 R2014a (The Mathworks (USA)).
Fig. 1. The correlation between the consumed Cl(+1) and the formed TEMPO+ in a
reaction of 2 mM TEMPO and 4 mM NaOCl in a borate buffer (pH 8.3) at room temper-
ature. The concentrations of TEMPO+ formed and Cl(+1) consumed were obtained by
iodometric titration. Corresponding correlation with reaction times added is shown
in Supplementary material.
ence in the thiosulphate consumption with and without the added
DMSO by applying Eqs. (1) and (2). TEMPO+ was clearly formed in
parallel with the consumption of Cl(+1).
The correlation between the consumed Cl(+1) and the TEMPO+
formed is further illustrated in Fig. 1. The stoichiometry of the acti-
vation reaction with HOCl and TEMPO is proposed to be 1:2 [11].
The observed results correlate with the 1:2 stoichiometry except
for the longest reaction times. This deviation could be explained
with the formation of ClO3− which was detected after several hours
of reaction time. The chlorate formation correlates with the self-
decomposition of HOCl [27], which is more pronounced at long
reaction times and especially at high Cl(+1) concentrations.
3. Results and discussion
3.2. UV–vis absorption spectroscopy of TEMPO, TEMPO+ and
Cl(+1)
3.1. Iodometric titration of TEMPO+ and Cl(+1)
The conversion of TEMPO and Cl(+1) and the formation of
TEMPO+ were monitored also by UV–vis absorption spectroscopy
(Fig. 3). The spectra of the reaction solutions and the corresponding
spectra of Cl(+1) in the borate buffer are shown in the supple-
mentary material. Mathematical fitting was used instead of direct
observation of the absorption maxima shifting due to the over-
lapping of the spectra of TEMPO, TEMPO+, and the Cl(+1). The
mathematically reconstructed spectra of TEMPO, TEMPO+, and
Cl(+1), shown in Fig. 2, are based on the measurements illustrated
in Supplementary material.
The reaction of TEMPO and NaOCl was studied in a pH 8.3
buffer solution because at this pH the content of the undissoci-
ated HOCl (pKa 7.5), that is the reactive species, is still relatively
high (∼14%). The reaction was monitored as a function of time by
iodometric titrations with and without DMSO, which is used as a
masking agent for HOCl [43–46] (S1). The formation of a reaction
product was detected when DMSO was applied. This compound
disappeared when an aliphatic alcohol, n-propanol, was added
in the reaction mixture of TEMPO and NaOCl. The slow forma-
tion under the low reactant concentrations and the disappearance
with the added alcohol promotes an assumption that the com-
ponent is the oxidized form of TEMPO, namely TEMPO+ which is
reported to react stoichiometrically with iodine as described by Eq.
(3) [1,2,30,40]. TEMPO+ is converted to its reduced form, a hydrox-
ylamine (TEMPOH), during the reaction with alcohols (Scheme 2).
Primary aliphatic alcohols can be oxidized with TEMPO to the cor-
responding aldehydes in few minutes [5]. The oxidation of the
alcohols with TEMPO is clearly faster with added NaBr [11]. The co-
catalysis by bromide could be due to the faster activation of TEMPO
to its oxidized form.
The formation rate of TEMPO+ in 2 mM TEMPO solution, which is
typically applied in the TEMPO-mediated oxidation, and the forma-
tion rate in a more concentrated TEMPO solution are compared in
Fig. 4. The concentration ratio of NaOCl and TEMPO and the concen-
tration of NaOCl clearly influence on how fast TEMPO+ is formed. In
the plateau point of TEMPO+ formation the consumption of thiosul-
fate (iodometric titration in the presence of DMSO) is almost equal
to the amount of TEMPO added. The results demonstrate the bene-
fit of a separate activation step of TEMPO prior to its application as
a bromide-free oxidation catalyst e.g., for the oxidation of cellulosic
pulps. A separate activation of TEMPO by NaOCl at pH 7–8 can be
accomplished within an half of hour when 20–40 mM TEMPO and
excess of NaOCl are used (results not shown).
The thiosulphate consumption with the added DMSO was con-
verted to the concentration of TEMPO+ according to Eqs. (1) and (3)
(Fig. 3). The concentration of Cl(+1) was obtained from the differ-
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
5
Fig. 2. Mathematically reconstructed spectra of TEMPO, TEMPO+, and Cl(+1) from
monitoring the reaction of 2 mM TEMPO and 4 mM NaOCl in a borate buffer (pH 8.3)
by UV–vis spectroscopy at room temperature.
Fig. 5. The decomposition of HOBr, formed in 4 mM NaBr and 4 mM NaOCl in a
borate buffer (pH 8.3) at RT, by DMSO as monitored by iodometric titration.
3.3. Iodometric titration of TEMPO+ in the presence of added NaBr
TEMPO / UV-Vis-MRC
Cl(+1) / Iodometric titrations
Cl(+1) / UV-Vis-MRC
The use of DMSO as a masking agent for HOCl has been reported
widely [43–46]. DMSO can be applied as a masking agent for HOBr
correspondingly [47]. However, the reaction rate between HOBr
and DMSO is slower than the reaction rate between HOCl and
DMSO. The required time for removing all HOBr in the mixture of
NaBr and NaOCl by DMSO was determined by varying the incuba-
tion time prior to the addition of KI and starch (Fig. 5). The trapping
of HOBr by DMSO appeared to follow the first order exponential
decay, thus, the following equation was fitted to the titration data:
4.0
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
+
TEMPO / Iodometric titrations
TEMPO+ / UV-Vis-MRC
[HOBr] = [HOBr]0 × exp(−kt)
(4)
where the rate constant k is 0.10 s−1. Thus a delay of 2 min between
the addition of DMSO and the addition of KI was considered appro-
priate to enable the titration of TEMPO+ regardless of the amount
of Br(+1) present in the sample solution.
The iodometric titration with and without added DMSO was
applied for monitoring the NaBr/TEMPO/NaOCl reaction system
(S2) similarly as presented in supplementary data (S1) for the
TEMPO/NaOCl system. The formation rate of TEMPO+ was high
despite the 10 times lower concentration level compared to the
bromide-free system (S1 and Fig. 3). The use of bromide as an accel-
erating compound of TEMPO-mediated oxidation is well known
and reported [10,12]. Our experiments verify one of the important
effects of bromide as a co-catalyst is the faster conversion of TEMPO
to TEMPO+.
0
30
60
90 120 150 180 210 240
Time (min)
Fig. 3. Disappearance of Cl(+1) and TEMPO and formation of TEMPO+ in a reaction
mixture of 2 mM TEMPO and 4 mM NaOCl in a borate buffer (pH 8.3) at room temper-
ature. The graph compares the results obtained by iodometric titration and UV–vis
spectroscopy combined with MCR.
The correlation between the consumption of Cl(+1)/Br(+1) and
formation of TEMPO+ (Fig. 6) was similar to the one detected with
the TEMPO/NaOCl system (S1 and Fig. 3). In both cases the hypo-
halous acids were consumed and TEMPO+ formed in the theoretical
1:2 ratio.
3.4. Monitoring of the TEMPO-mediated oxidation of xylitol
The iodometric titration of Cl(+1) and TEMPO+ was applied to
study the catalytic oxidation of xylitol with preactivated TEMPO
(2 mM) at pH 8.3 in buffered conditions (Fig. 7). Interestingly,
30 mM NaOCl was consumed during reaction with 7.5 mM xylitol.
The observed 4:1 stoichiometry corresponds to the expected con-
version of xylitol into xylaric acid. The concentration of TEMPO+
remained at a constant level until all hypochlorite was consumed
after which also TEMPO+ disappeared.
Fig. 4. Effect of the concentrations of TEMPO and NaOCl on the conversion to
TEMPO+ in a borate buffer (pH 8.3) as a function of time, analyzed by iodometric
titration.
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
6
Fig. 6. Disappearance of Cl(+1)/Br(+1) and formation of TEMPO+ in a reaction mix-
ture of 0.2 mM TEMPO, 0.2 mM NaBr, and 0.2 mM NaOCl in a borate buffer (pH 8.3)
at room temperature.
Fig. 7. Disappearance of Cl(+1) and TEMPO+ during TEMPO (0.4 mM) catalyzed oxi-
dation of a 7.5 mM xylitol with 30 mM NaOCl at room temperature. pH was buffered
with borate buffer to 8.3.
3.5. Monitoring of the TEMPO-mediated oxidation of bleached
pulps by iodometric titration
The iodometric titration of Cl(+1) and TEMPO+ was applied to
study the catalytic oxidation of a bleached birch kraft pulp with
preactivated TEMPO (2 mM) at pH 9 that has been reported to
be the optimum pH for the bromide-free oxidation [11] (Fig. 8A).
During the course of the oxidation TEMPO existed mostly in the
form of TEMPO+, the concentration of which stayed at a constant
level (1.7 mM) until all Cl(+1) was consumed (Fig. 8A). Similar
observations were made with bleached eucalyptus kraft and euca-
lyptus prehydrolysis kraft pulps (Supplementary material). Then
TEMPO+ disappeared at a rate equal to the final rate of conversion of
Cl(+1). These observations indicate that the oxidation of TEMPOH to
TEMPO+ is much faster than its oxidative reaction with the primary
alcohol groups in cellulose. A similar conclusion was reported ear-
lier by Bragd et al. [11] who found that the rate of TEMPO catalyzed
oxidation of methyl ␣-d-glucopyranoside was unaffected by the
concentration of Cl(+1). However, when pH of the reaction mixture
was adjusted from 9 to 10.5 (Fig. 8B) during the reaction, TEMPO+
almost disappeared (<0.2 mM TEMPO+) after which the reaction
Fig. 8. Disappearance of Cl(+1) and TEMPO+ during TEMPO (2 mM) catalyzed oxida-
tion of a bleached birch kraft pulp (4% consistency) with NaOCl at room temperature.
[A] Oxidation with preactivated TEMPO and 57 mM NaOCl at pH 9, [B] repetition of
A oxidation except pH adjustment (9 → 10.5) with NaOH after 12 min reaction time
(dash line). [C] Disappearance of Cl(+1)/Br(+1) and TEMPO+ during TEMPO (2 mM)
and NaBr (0.4 mM) catalyzed oxidation of a bleached birch kraft pulp (3.3% consis-
tency) with NaOCl (97 mM) at pH 10 at room temperature. TEMPO was preactivated
prior to the oxidation.
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024

G Model
APCATA-15474; No. of Pages7
ARTICLE IN PRESS
T. Pääkkönen et al. / Applied Catalysis A: General xxx (2015) xxx–xxx
7
continued with a much lower rate. Thus, at the high pH reoxida-
tion of TEMPOH became the rate-limiting reaction, probably due to
the low content of HOCl at the high alkalinity.
[3] M.F. Semmelhack, C.R. Schmid, C.S. Cortes, D.A. Chou, J. Am. Chem. Soc. 106
(1984) 3374–3376.
[4] M.F. Semmelhack, D.A. Chou, C.S. Cortes, J. Am. Chem. Soc. 105 (1983)
4492–4494.
Under identical conditions more carboxylate and aldehyde
groups (0.92 and 0.21 mmol/g, respectively) were formed in the
eucalyptus prehydrolysis kraft pulp in comparison with the normal
eucalyptus kraft pulp (0.75 and 0.13 mmol/g, respectively) (Sup-
plementary material). The enhanced oxidation of the prehydrolysis
kraft pulp can be explained by its lower xylan content. Similar effect
of the xylan content on the oxidation of cellulosic pulps has been
reported earlier [29].
When a small amount of NaBr was used as a co-catalyst, the
concentration of TEMPO+ stayed close to the concentration of the
added TEMPO even though the oxidation was carried out at pH 10
(Fig. 8C). The bromide-assisted oxidation has been reported to be
fastest at pH 10 [11]. Our results verify that the NaBr addition that
leads to formation of HOBr, promotes the reoxidation of TEMPOH at
least at high pH levels. In part, this can be understood by the lower
degree of dissociation of HOBr (pKa 8.7) in comparison with HOCl
(pKa 7.5). These examples demonstrate the utility of the iodometric
titration in unravelling the rate-determining steps in the TEMPO
catalyzed oxidation of cellulosic pulps.
[5] P.L. Anelli, C. Biffi, F. Montanari, S. Quici, J. Org. Chem. 52 (1987) 2559–2562.
[6] A. Isogai, Y. Kato, Cellulose 5 (1998) 153–164.
[7] T. Saito, Y. Okita, T. Nge, J. Sugiyama, A. Isogai, Carbohydr. Polym. 65 (2006)
435–440.
[8] T. Saito, Y. Nishiyama, J. Putaux, M. Vignon, A. Isogai, Biomacromolecules 7
(2006) 1687–1691.
[9] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Tetrahedron 51 (1995)
8023–8032.
[10] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Carbohydr. Res. 269 (1995)
89–98.
[11] P.L. Bragd, A.C. Besemer, H. van Bekkum, Carbohydr. Res. 328 (2000) 355–363.
[12] P.L. Bragd, H. van Bekkum, A.C. Besemer, Top. Catal. (27 2004) 49–66.
[13] A. Saito, T. Isogai, Biomacromolecules 5 (2004) 1983–1989.
[14] B. Sun, C. Gu, J. Ma, B. Liang, Cellulose 12 (2005) 59–66.
[15] T. Saito, Y. Nishiyama, J. Putaux, M. Vignon, A. Isogai, Biomacromolecules 7
(2006) 1687–1691.
[16] T. Saito, M. Hirota, N. Tamura, S. Kimura, H. Fukuzumi, L. Heux, A. Isogai,
Biomacromolecules 10 (2009) 1992–1996.
[17] Y. Okita, T. Saito, A. Isogai, Holzforschung 63 (2009) 529–535.
[18] S. Iwamoto, W. Kai, T. Isogai, T. Saito, T. Isogai, A. Iwata, Polym. Degrad. Stab.
95 (2010) 1394–1398.
[19] A. Isogai and T. Saito, US 2010/0233481 A1 (2010).
[20] T. Isogai, T. Saito, A. Isogai, Cellulose 18 (2011) 421–431.
[21] R. Tanaka, T. Saito, A. Isogai, Int. J. Biol. Macromol. 51 (2012) 228–234.
[22] A. Isogai, T. Saito, H. Fukuzumi, Nanoscale 3 (2011) 71–85.
[23] M. Bulota, S. Tanpichai, M. Hughes, S.J. Eichhorn, ACS Appl. Mater. Interfaces 4
(2011) 331–337.
[24] M. Hirota, N. Tamura, A. Saito, T. Isogai, Carbohydr. Polym. 78 (2009) 330–335.
[25] T. Kitaoka, A. Isogai, F. Onabe, Nord. Pulp Paper Res. J. 14 (1999) 279–284.
[26] T. Saito, M. Hirota, A. Tamura, N. Isogai, J. Wood Sci. 56 (2010) 227–232.
[27] M. Lewin, M. Avrahami, J. Am. Chem. Soc. 77 (1955) 4491–4498.
[28] T. Saito, Y. Okita, T.T. Nge, J. Sugiyama, A. Isogai, Carbohydr. Polym. 65 (2006)
435–440.
[29] R. Pönni, T. Pääkkönen, M. Nuopponen, J. Pere, T. Vuorinen, Cellulose 21
(2014) 2859–2869.
[30] V.A. Golubev, E.G. Rozantsev, M.B. Neiman, Institute of chemical physics,
Acad. Sci. USSR 11 (1965) 1927–1936.
[31] I.M. Ganiev, Q.K. Timerghazin, A.F. Khalizov, V.V. Shereshovets, A.I. Grigor’ev,
G.A. Tolstikov, J. Phys. Org. Chem. 14 (2001) 38–42.
[32] I.M. Ganiev, Q.K. Timerghazin, A.F. Khalizov, N.M. Andriyashina, V.V.
Shereshovets, L.B. Volodarsky, G.A. Tolstokov, Tetrahedron Lett. 40 (1999)
4737–4740.
4. Conclusions
The quantification of TEMPO+ by iodometric titration is a useful
tool for studying the chemistry of the TEMPO-mediated oxidation
of primary alcohols. Similar and complementary information can
be obtained by UV–vis spectroscopy combined with multivariate
curve resolution, which can be applied for simultaneous quantifi-
cation of Cl(+1)/Br(+1), TEMPO+ and TEMPO. In general, conversion
of TEMPO to TEMPO+ is the slowest step in bromide-free oxida-
tion of the alcohols by Cl(+1). The use of bromide accelerates this
conversion that can also be accomplished through a separate pre-
treatment of TEMPO with an excess of Cl(+1) in neutral conditions.
The rate of the preactivation depends on the concentrations of
TEMPO and Cl(+1) and is high enough to be applied in potential
industrial processes such as catalytic oxidation of cellulosic pulps.
TEMPO+ was the dominant form of TEMPO in the catalytic oxidation
of a birch kraft pulp under optimal conditions which confirms the
earlier observation on the oxidation of the hydroxymethyl groups
being the rate-determining step in the catalytic process.
[33] A.E.J. de Nooy, A.C. Besemer, H. van Bekkum, Tetrahedron 51 (1995)
8023–8032.
[34] W.F. Bailey, J.M. Bobbitt, K.B. Wiberg, J. Org. Chem. 72 (2007) 4504–4509.
[35] A. Israeli, M. Patt, M. Oron, A. Samuni, R. Kohen, S. Goldstein, Free Radical Biol.
Med. 38 (2005) 317–324.
[36] J.T. Hill-Cousins, J. Kuleshova, R.A. Green, P.R. Birkin, D. Pletcher, T.J.
Underwood, S.G. Leach, R.C.D. Brown, ChemSusChem 5 (2012) 326–331.
[37] D. Liaigre, E.M. Breton, T. Belgsir, Electrochem. Commun. 7 (2005) 312–316.
[38] T. Isogai, T. Saito, A. Isogai, Biomacromolecules 11 (2010) 1593–1599.
[39] J. Peyre, T. Pääkkönen, M. Reza, E. Kontturi, Green Chem. 17 (2015) 808–811.
[40] E.G. Rozantsev, V.D. Sholle, Synthesis 1971 (1971) 401–414.
[41] D. da Silva Perez, A.R.P. van Heiningen, Proc. Seventh Eur. Workshop
Lignocellulosics Pulp (2002) 393–396.
Acknowledgements
[42] I. Wartiovaara, Pap. Puu 64 (1982) 534–545.
This study was supported by UPM Nanocenter and the Finnish
Funding Agency for Innovation (TEKES). We thank Mrs Mirja
Reinikainen for excellent laboratory work.
[43] N. Imaizumi, T. Kanayama, K. Oikawa, Analyst 120 (1995) 1983–1987.
[44] T. Lehtimaa, V. Tarvo, G. Mortha,[] S. Kuitunen, T. Vuorinen, Ind. Eng. Chem.
Res. 47 (2008) 5284–5290.
[45] V. Tarvo, T. Lehtimaa, S. Kuitunen, V. Alopaeus, T. Vuorinen, J. Aittamaa, Ind.
Eng. Chem. Res. 48 (2009) 6280–6286.
Appendix A. Supplementary data
[46] B.H. Yoon, L. Wang, J. Pulp. Paper Sci. 28 (2002) 274–279.
[47] J. Viroopakshappa, V. Jagannadham, J. Chem. Sci. 114 (2002) 149–154.
[48] H. Abdi, L.J. Williams, Wiley Interdiscip. Rev. Comput. Stat. 2 (2010) 433–459.
[49] J. Jaumot, R. Gargallo, A. de Juan, R. Tauler, Chemom. Intellig. Lab. Syst. 76
(2005) 101–110.
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2015.07.
024
References
[1] V.D. Sen’, V.A. Golubev, T.M. Kosheleva, Russ. Chem. B+ 26 (1977) 678–683.
[2] V. Golubev, V. Sen’, E.G. Rozantsev, Russ. Chem. B+ 28 (1979) 1927–1931.
Please cite this article in press as: T. Pääkkönen, et al., Appl. Catal. A: Gen. (2015), http://dx.doi.org/10.1016/j.apcata.2015.07.024