Article in press - uncorrected proof

Holzforschung, Vol. 65, pp. 295–301, 2011 • Copyright
by Walter de Gruyter • Berlin • New York. DOI 10.1515/HF.2011.062

MgSO4 vs. Mg(OH)2 as a cellulose protector in oxygen

delignification

Jean Bouchard*, Jie Wang and Richard Berry – MgSO4 reacts with alkali to produce insoluble
FPInnovations, Pulp and Paper Division, Point-Claire, Mg(OH)2, which inactivates the transition metal by phy-
Québec, Canada sically encapsulating them either by adsorption (Erics-
son et al. 1971; Robert and Viallet 1971; Sjöström and
*Corresponding author. Valttila 1972; Gustavsson and Swan 1975) or by the
FPInnovations, Pulp and Paper Division, 570 Boulevard
St-Jean, Pointe-Claire Québec, Canada H9R 3J9 formation of coordination complexes (Gilbert et al.
Phone: q1-(514)-630-4100 1973). More recently, it has been demonstrated that
Fax:q1-(514)-630-4134 Fe2q and Mn2q can be protected against oxidation if
E-mail: [email protected] these ions are incorporated in magnesium precipitates
(Lidén and Ohman 1997; Wiklund et al. 2001).
– Magnesium ions form stable metal-carbohydrate com-
Abstract
plexes with the carbonyl groups of the partially oxidized
It has been proposed that magnesium hydroxide can replace cellulose and reduce the rate of cellulose chain cleavage
magnesium sulfate as a lower cost cellulose protector. This by the b-elimination reaction (Defaye and Gadelle
claim has been evaluated based on four commercial unblea- 1974; Lapierre et al. 2003).
ched softwood kraft pulps. The selectivity was improved in
the presence of soluble MgSO4, but little or no selectivity It has been proposed that MgSO4 can be replaced by
improvement was obtained when insoluble Mg(OH)2 was Mg(OH)2 as a cellulose protector in oxygen delignification
applied. It was further observed that in-situ iron concentra- (Gibson and Wajer 2003; Gibson et al. 2004). The major
tion in unbleached pulp affects selectivity of oxygen deli- claimed benefit is the lower cost of magnesium hydroxide at
gnification, while in-situ calcium concentration affects effi- the same level of cellulose protection. This paper focuses
ciency of magnesium protection. The results favor the on whether soluble MgSO4 can be replaced effectively by
hypothesis of formation of a Mg-carbohydrate complex as Mg(OH)2 for cellulose protection in the oxygen stage.
the major mechanism of cellulose protection by magnesium.
The application of a soluble magnesium compound is rec-
ommended to improve selectivity in oxygen delignification.
Materials and methods
Keywords: calcium compounds; delignification; ions; mag-
nesium hydroxide; magnesium sulfate; oxygen; selectivity; Pulp samples
transition metals; viscosity. Two pulps were from mills in coastal British Columbia (CBC):
CBC-1 was produced from 65% Douglas fir and 35% western hem-
lock. CBC-2 was made from a mixture of spruce, pine, and fir (SPF)
Introduction furnish. The third pulp was from an interior British Columbia (IBC)
mill based on a mixture of 70% pine, 25% spruce, and 5% of var-
One of the most important factors affecting selectivity of ious fir species. The fourth final furnish was from northern Québec
oxygen delignification is the kind and amount of transition (NQC) and made from black spruce with less than 10% Jack pine.
metal ions in pulps. During the stepwise reduction of oxygen When received, the pulps were washed three times with deionized
to water, hydrogen peroxide is formed and catalytically water and kept at 48C. The pertinent characteristics of these pulps
decomposed by transition metal ions to yield reactive are presented in Table 1. The metal ion profiles were determined
hydroxyl and oxyl anion radicals. These radicals randomly by means of a sequential inductively coupled plasma (ICP) emission
oxidize the cellulose chain by forming carbonyl groups spectrometer (Thermo Jarrell Ash, Atom Scan 25, Mississauga,
which, under alkaline conditions, undergo a b-elimination Ontario, Canada).
These pulps provided a wide range of kappa number and viscosity
reaction resulting in cleavage of the cellulose chain. The net
and, because they had similar native magnesium content (about
result is a decrease in pulp viscosity and ultimately in pulp
300 ppm), the effect of native magnesium would not interfere with
strength (McDonough 1996).
the results. Another valuable characteristic is that these pulps pro-
The discovery by Robert et al. (1964, 1966) that the addi- vided a range of transition metal concentration (Fe, Mn, and Al),
tion of a small amount of magnesium sulfate significantly varying from 61 ppm (CBC-1) to 174 ppm (NQC). It was decided
reduces the cellulose damage led to the implementation of not alter the metal ion profile of the pulps by adding specific metal
the first commercial oxygen stage in South Africa in 1970 salts as information was sought for real industrial pulps and there
(Rowlandson 1971). Two mechanisms have been proposed is no evidence that added metal ions are equivalent to intrinsic metal
to explain the protective effect of magnesium on cellulose: ions.

Brought to you by | Western University 2010/148

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
296 J. Bouchard et al.

Table 1 Kappa number, viscosity, and metal ion content (ppm) of four unbleached softwood kraft pulps
before and after acid wash.

Unbleached pulps Acid-washed unbleached pulps

Properties CBC-1 CBC-2 IBC NQC CBC-1 CBC-2 IBC NQC
Kappa no. (ml) 34 27 32 21 32.5 24.7 30.0 20.9
Viscosity (mPa.s) 39 42 38 33 34.6 39.7 35.7 32.1
Mg (ppm) 260 310 295 340 5.1 9.7 4.7 17.9
Mn (ppm) 31 65 66 120 0.7 0.9 1.3 2.6
Fe (ppm) 15 19 11 28 7.8 6.3 6.2 9.1
Al (ppm) 14 27 18 26 2.4 2.8 2.2 3.8
Ca (ppm) 1540 1190 1150 1540 29.2 19.0 15.8 38.2
CBC, coastal British Columbia; IBC, interior British Columbia; NQC, northern Québec.

Oxygen delignification with hydroxide ions to form insoluble Mg(OH)2 (Ksp:

5.66=10-12) and decrease the hydroxide concentration.
As shown in Table 2, a range of MgSO4 or Mg(OH)2 charges was
used (Fisher Scientific, Toronto, Ontario, Canada, transition metal MgSO4™Mg2qqSO42-
content below 8 ppm). The specific charge was added to the pulp
suspension and mixed for 2 min in a Hobart mixer (Model: N-50, Mg2qq2OH- l Mg(OH)2
Don Mills, Ontario, Canada). The caustic charge (2.5% based on
According to this mechanism, the addition of Mg(OH)2
oven-dried pulp) was then added and the pulp was mixed for another
2 min. The final consistency of the pulp sample was 10%. The pulp
instead of MgSO4 should not consume any hydroxide ions
was heated to 808C in a microwave oven and transferred into a pre- and, consequently, should have no impact on delignification
heated laboratory peg type reactor (designed and built at FPInno- efficiency. Kappa number as a function of the Mg concen-
vations) and mixed at 10 rpm. When the temperature reached 958C, tration is shown in Figure 1a for oxygen-delignified CBC-1
the reactor was pressurized to 690 kPa with oxygen. At the end of pulp. Similar results were obtained with the three other pulps.
the 60-min reaction, a fraction of the filtrate was collected for final With MgSO4 addition, the degree of delignification is
pH determination and the pulp was washed thoroughly with distilled reduced as the Mg concentration is increased. The increase
water before kappa number and viscosity determination (PAPTAC in kappa number for the highest Mg ion charge applied, rang-
Standard Methods G.18 and G.24P). es between 0.9 and 1.9 as compared to delignification of the
control pulp (no magnesium added). With Mg(OH)2 addi-
Acid wash tion, the degree of delignification also decreases unexpect-
The pulp was brought to 2% consistency and the pH was adjusted
edly as the Mg concentration is increased, although to a
to 1.5"0.05 with sulfuric acid. The pulp was retained for 1 h at lesser extent than in the sulfate case. The increase in kappa
508C and then washed three times with deionized water. The washed number ranges from 0.4 to 1.3 for the highest Mg ion charge
pulp was kept at 48C. For analytical data, see Table 1. applied.
These results show that the small reduction in hydroxide
concentration caused by the addition of MgSO4 (and the sub-
Results and discussion sequent formation of insoluble Mg(OH)2) does affect deli-
gnification. This effect on delignification is decreased but not
Mg addition and delignification eliminated by using Mg(OH)2. All the final pHs were above
11.7 and there is no correlation between final pH and Mg
Unpublished results from FPInnovations have shown that charge added on the pulp. The results also show that any
MgSO4 addition prior to oxygen delignification will decrease viscosity gain obtained from Mg addition must be interpreted
the efficiency of the delignification process and, under iden- taking into account the adverse effect on delignification.
tical conditions, will provide a pulp with a slightly higher The observation that Mg(OH)2 also causes a decrease in
kappa number. Under alkaline conditions, Mg ions will react delignification in all cases can be interpreted that there is a
second mechanism inhibiting the lignin oxidation reaction or
Table 2 Charges of MgSO4 or Mg(OH)2 applied on pulp to pro- the oxidized lignin solubilization. One possibility is that phe-
vide identical Mg2q concentrations. nolates, which are the critical first sites of attack for oxygen,
are made less active when the phenolate counter-ion is Mg
Charges of additives rather than sodium. Another possibility is that oxidized lignin
Conc. achieveda
MgSO4 (%) Mg(OH)2 (%) Mg2q (%) fragments are less soluble in alkaline filtrate when comp-
lexed with Mg.
0.05 0.024 0.01
0.10 0.048 0.02
Mg addition and pulp viscosity
0.20 0.096 0.04
0.40 0.192 0.08 Pulp viscosity as a function of the Mg concentration is pre-
a
Based on o.d. pulp. sented in Figure 1b and c for the IBC and CBC-1 oxygen-

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
Magnesium in oxygen delignification 297

Table 3 Viscosity gains obtained from the two additives at two

Mgq2 concentrations.

Viscosity gain of pulps

Nominal
Mg2q Form of IBC CBC-2 NQC CBC-1
conc. (%) Mg (mPa.s) (mPa.s) (mPa.s) (mPa.s)
0.01 MgSO4 1.2 2.2 0.4 0.4
Mg(OH)2 0 0 0 0.5
0.08 MgSO4 2.5 3.1 1.8 2.1
Mg(OH)2 0.6 1.6 0.7 1.7

compared at the lowest and highest charge of MgSO4 and

Mg(OH)2 applied.

Mg addition and selectivity of oxygen delignification

The selectivity index integrates the variables viscosity and

kappa number; i.e. the selectivity is defined in the present
work as the ratio of Dkappa number over Dviscosity. It is
the best way to compare the efficiency of the additives
MgSO4 and Mg(OH)2. Even though the relationship between
viscosity and the number of scissions along the cellulose
chains is not linear, this simple approach is advantageous
because of the narrow range of variation being evaluated.
The selectivity results for the four pulps studied are pre-
sented in Figure 2a and b where MgSO4 addition is repre-
sented by filled markers and Mg(OH)2 addition by open
markers. The range of selectivity shown on the y axis is
identical in both figures to allow direct comparison. For the
pulps CBC-2, IBC, and NQC, there is a significant improve-
Figure 1 (a) Kappa number obtained after oxygen delignification
ment of the selectivity in case of MgSO4 as cellulose pro-
of CBC-1 pulp as a function of the magnesium concentration. Delig- tector, even at the lowest concentration, while there is no
nification of the control pulp (%) is indicated by the arrow. (b and
c) Viscosity (mPa.s) obtained after oxygen delignification of (a) IBC
pulp and (b) CBC-1 pulps as a function of the magnesium concen-
tration added either from MgSO4 or Mg(OH)2 sources.

delignified pulps. IBC pulp shows the largest difference

between MgSO4 and Mg(OH)2 for viscosity improvement
while for CBC-1 the difference is small. The two other pulps
reveal intermediate results. In Figure 1b and c, the range of
viscosity shown on the y axis is identical (4 mPa.s) to allow
easy comparison between the pulps.
Of the four pulps studied, three showed significant impro-
vements of viscosity when the Mg is added as MgSO4 rather
than Mg(OH)2. The IBC pulp (Figure 1b) exhibited the larg-
est difference, and a large viscosity increase was obtained
with the lowest charge of MgSO4; larger charges caused only
small further increases. With Mg(OH)2 almost no viscosity
gain was observed even with the highest charge.
The CBC-1 pulp (Figure 1c) responded differently; there
was no substantial viscosity increase at low MgSO4 charge
but rather a steadily increasing viscosity as MgSO4 concen-
tration was increased. Mg(OH)2 addition was as effective as
MgSO4 for the CBC-1 pulp. With the two other pulps (CBC- Figure 2 Selectivity of oxygen delignification of softwood kraft
2 and NQC) the viscosity improvement was always superior pulps as a function of the magnesium concentration added either
when adding MgSO4 instead of Mg(OH)2. These results are from MgSO4 or Mg(OH)2 source. (a) IBC and CBC-1 pulps.
summarized in Table 3 where viscosity gains (mPa.s) are (b) CBC-2 and NQC pulps.

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
298 J. Bouchard et al.

improvement at all with the addition of Mg(OH)2. As it was

the case for the viscosity data, the CBC-1 pulp responds
differently than the three other pulps to Mg addition. For
CBC-1 pulp, there is no significant improvement in selectiv-
ity with both MgSO4 and Mg(OH)2 additives. This obser-
vation is a reminder that pulps respond differently even
under identical experimental conditions, and it is against
good scientific praxis to draw general conclusions from a
single furnish.

Furnish and selectivity of oxygen delignification

Figure 4 Maximum selectivity gain obtained from magnesium
Figure 2 also confirms that the initial selectivity of oxygen addition. Numbers on top of the bars indicate the calcium concen-
tration (ppm) found in each pulp.
delignification without addition of Mg ions depends on the
furnish, irrespective of the kappa number. Without additives,
the wide range of selectivity between 0.8 (NQC) and 1.1 stabilization of the glycosidic linkage of oxidized cellulose
(IBC) is remarkable for four kraft pulps having similar in- by formation of stable Mg-carbohydrate complexes. The
situ magnesium contents (Table 1) which were oxygen-deli- effect of a higher Ca concentration indicates that there is an
gnified in the same reactor under the same conditions. interference caused by either Ca competing with Mg in form-
The data of transition metals (Table 1) allow an explana- ing the carbohydrate complex or by Ca removing active Mg
tion for the differences: there is a relationship between the from the system through a Ca-Mg compound.
initial iron concentration in pulp and the selectivity of oxy-
gen delignification as illustrated in Figure 3. Order of addition of Mg additives and selectivity
It has been demonstrated (Sjöström 1981; Cardona-Barrau The superior performance of MgSO4 over Mg(OH)2 as a
and Lachenal 2001) that the addition of either FeII or CuII to cellulose protector can be reasonably explained as the dif-
pulp will induce a strong degradation of cellulose during ference between the addition of soluble Mg2q ion as com-
oxygen delignification. These transition metals catalyze the pared to insoluble Mg(OH)2. In all the results presented
decomposition of hydrogen peroxide formed during the above, the MgSO4 solution and the Mg(OH)2 suspension
reduction of oxygen and lead to generation of hydroxyl and were always added on pulp and mixed prior to NaOH addi-
oxyl anion radicals (Cardona-Barrau and Lachenal 2001; tion. The positive effect on selectivity of soluble Mg as com-
Bouchard et al. 2010). In the quoted literature, there was pared to insoluble Mg can be explained by either of the
always a concern about the equivalence of added metal ions mechanisms proposed in the literature:
versus in-situ transition metal ions. The present work dem-
onstrates that the selectivity of oxygen delignification of a • The soluble magnesium ion leads to complexation with
series of pulps is determined, at least partially, by their in- carboxylic acids or carbonyl groups present on the cel-
situ iron concentration. lulose in the case if some contact time with the pulp is
Another observation is that the magnitude of the selectiv- available. This complexation mechanism has much less
ity gain obtained by MgSO4 addition is also a function of chance to occur if the Mg is added in an insoluble state.
the furnish, but it is not related to the iron content. The • Presence of soluble Mg is a prerequisite to explain the
alkaline metal data from Table 1 in the plots of Figure 4 physical encapsulation of transition metal hydroxide with-
permit the interpretation: the lower the Ca content of the in the Mg(OH)2 precipitate formed when sodium hydrox-
pulp, the higher is the selectivity gain from Mg addition. This ide is added.
observation is in agreement with the proposed mechanism of In the following series of experiments, the order of addi-
tion of chemicals on the CBC-2 pulp was varied in four
ways. In all cases, the Mg concentration was varied between
0.0% and 0.08% (% Mg2q on o.d. pulp):
• MgSO4qNaOH: Mg2q addition, 2 min of mixing, NaOH
addition, 2 min of mixing. Soluble Mg is contacting a
neutral well-washed pulp.
• wNaOH and MgSO4x: Both solutions are added simulta-
neously on pulp followed by 4 min of mixing. The pulp
is becoming alkaline at the same time that the soluble Mg
is contacting it.
• NaOHqMgSO4: NaOH addition, 2 min of mixing, Mg
addition, 2 min of mixing. Soluble Mg is contacting an
alkaline pulp.
Figure 3 Selectivity of oxygen delignification of four softwood • Mg(OH)2qNaOH: Mg(OH)2 suspension addition, 2 min
kraft pulps as a function of their initial iron content. of mixing, NaOH addition, 2 min of mixing. Alkaline

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
Magnesium in oxygen delignification 299

insoluble Mg(OH)2 is contacting a neutral well-washed kappa number and viscosity reduction obtained through the
pulp. complete AO sequence.
Moreover, a good efficiency of metal removal can be con-
The calculated selectivities of the oxygen stages as a func-
firmed as the metal profiles of each of the acid-washed pulps
tion of the Mg2q charge applied are shown in Figure 5.
are quite similar. However, the degree of iron removal is low,
When adding MgSO4 on a well-washed neutral pulp (),
varying between 30% and 50%, which is in agreement with
the lowest Mg charge (0.01% Mg2q) provides the highest
previous observations that pulps contain some recalcitrant
selectivity. The presence of higher Mg charges decreases the
iron (Sjöström 1981; Lapierre et al. 1997).
selectivity slightly. This peak can be explained through the
These acid-washed pulps were oxygen delignified without
difference in the delignification and carbohydrate protection
any Mg additives or after addition of 0.04% Mg2q as either
responses to soluble Mg ions; adding more Mg than needed,
MgSO4 or Mg(OH)2. The results are very similar for all pulps
marginally increases the pulp viscosity while it linearly
and the results for CBC-2 pulp are presented in Table 4.
decreases the delignification, as demonstrated previously,
A simple metal ion removal step by acid wash signifi-
providing a net effect of reduced selectivity.
cantly increases the selectivity of oxygen delignification,
This peak positive effect on selectivity at the lowest
because it will decrease the quantity of hydroxyl radicals
charge applied is still observed when MgSO4 and NaOH are
generated from transition metal catalysis of hydrogen per-
added together (m), although the effect is less dramatic then
oxide decomposition. Consequently, it can be concluded that
when MgSO4 is added on a neutral pulp. Adding NaOH
Mg ion addition on untreated pulp cannot completely deac-
before MgSO4 (j) eliminates the peak completely. Effec-
tivate transition metals and/or achieve the full protective
tively, at a charge of 0.04% Mg2q and above, all the modes
complexation with carbohydrates.
of MgSO4 addition provide equivalent selectivity although it
Once the transition metal ions have mostly been removed
is still lower than that when a low charge of MgSO4 is added
by acid wash, any significant effect of adding Mg ions can-
on a well-washed neutral pulp. In contrast, the direct addition
not be expected, if the role of Mg is only to co-precipitate
of Mg(OH)2 (d) did not increase the selectivity which dem-
the transition metals in the form of Mg(OH)2-based com-
onstrates that the very low concentration of free Mg2q present
plexes. The metal ions remaining in the pulp after an acid
in a Mg(OH)2 suspension is too low to have any significant
wash (particularly iron) are strongly bound to the pulp and
effect.
there is little chance that they will detach from the pulp by
means of a complex with insoluble Mg(OH)2. To verify this
Mg addition and selectivity of acid-washed pulps
hypothesis, the four acid-washed pulps were also oxygen
It is well-known that a removal step of transition metals, delignified in the presence of MgSO4 or Mg(OH)2 (0.04%
such as a chelation or an acid wash stage, prior to oxygen Mgq2 on o.d. pulp).
delignification improves the selectivity of the process. An As shown in Table 4 for CBC-2 pulp, the selectivity of
acid wash stage was applied to the four pulps and the data oxygen delignification of the acid-washed pulp still signifi-
concerning kappa numbers, viscosities, and metal ion pro- cantly increases when MgSO4 is added prior to the oxygen
files of these acid-washed pulps are listed in Table 1 (right- stage. The results indicate that either the low concentration
hand side). of residual iron in the acid-washed pulps is deactivated
By comparing the acid-washed pulps in Table 1 with the (encapsulation mechanism) or, once the metals like Ca have
initial pulps, it can be observed that there is a slight decrease been stripped from the pulp by the acid wash, a greater level
in both the kappa numbers and the viscosities after the acid of Mg complex formation is achieved (carbohydrate-Mg-
treatment. For comparison with the previous data, the selec- complex mechanism).
tivity calculation for the acid-washed pulps will be based on When adding Mg(OH)2 on acid-washed pulps, a similar
selectivity improvement is the result (Table 4). This positive
effect, which was not observable when applied on untreated
brownstock, provides support for the following concept:
there is a localized neutralization of the carboxylic acid by
the hydroxyl groups and dissolution of free Mg2q, when all
the carboxylic acid groups in pulp are in a protonated form
and when the colloidal suspension of Mg(OH)2 is added on

Table 4 Effect of metal removal by acid wash on the selectivity

of oxygen delignification of CBC-2 pulp with (0.04% Mgq2) and
without (w/o) magnesium compounds addition. Selectivity: ratio of
Dkappa number over Dviscosity.
Figure 5 Selectivity of oxygen delignification of a softwood kraft
Initial pulp Acid-washed pulp
pulp (CBC-2) as a function of the magnesium concentration added
using different order of addition of chemicals. MgSO4qNaOH: ; w/o Mg 2q
MgSO4 Mg(OH)2 w/o Mg2q MgSO4 Mg(OH)2
wNaOH and MgSO4x: m; NaOHqMgSO4: j; Mg(OH)2qNaOH:
d. 0.84 0.93 0.84 1.09 1.22 1.23

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
300 J. Bouchard et al.

the pulp. This Mg ion can then bind to the pulp as soluble • Only a small charge of MgSO4 is needed for maximum
Mg. selectivity improvement.
• Removing most of the transition metals by acid washing
Possible protective mechanisms of magnesium ions improves the selectivity to a greater extent than adding
Mg to an untreated pulp. Adding Mg to an acid-washed
During oxygen delignification, transition metals catalyze the pulp provides a further benefit and even insoluble Mg in
formation of unselective hydroxyl and oxyl anion radicals (Mg(OH)2) is effective because of its localized solubili-
from hydrogen peroxide decomposition. A decrease in the zation after contacting a pulp in its acidic form.
rate of formation of radicals can be expected, if the mecha- • All the observations support the theory that Mg protects
nism accounting for the selectivity improvement by Mg addi- the carbohydrate during oxygen delignification by form-
tion involves an inactivation of the transition metal by ing carbohydrate complexes.
co-precipitation with Mg(OH)2. However, it has been report-
ed by Reitberger and Gierer (1988) that the rate of radical
formation is not decreased by the presence of Mg which
strongly suggests that Mg plays another role.
The observations summarized below all support the con- Acknowledgements
cept of Mg-carbohydrate-complexes which, for example, can
stabilize the oxidized glucose unit within cellulose and inhib- Special thanks to R. Osseni (stagiaire from Pagora-INP, Grenoble)
it the chain scission by alkaline b-elimination. The co-pre- for the preliminary work. We also acknowledge the Member Com-
cipitation of transition metals can still play a role but as a panies that provided pulp samples and Dr. John Schmidt for his
way of removing multivalent ions that otherwise could com- critical review of the manuscript.
pete with Mg at carbohydrate binding sites. These key obser-
vations are:
References
• Soluble Mg2q ions are the active species responsible for
selectivity improvement in oxygen delignification. Insol-
Bouchard, J., Wang, J., Berry, R. (2010) The role of hydroxyl and
uble Mg(OH)2 is not effective. oxyl anion radicals in selectivity of oxygen delignification.
• Most of the selectivity improvement is achieved with a Holzforschung 64:153–159.
low Mg charge and then levels off. This behaviour is Cardona-Barrau, D., Lachenal, D. (2001) Action of oxygen on the
consistent with binding to a finite number of available carbohydrates of a kraft pulp. The influence of a sodium boro-
sites. hydride pre-treatment. J. Wood Chem. Technol. 21:169–179.
• Metal removal by acid wash improves the selectivity to a Defaye, J., Gadelle, A. (1974) Magnesium salts in the alkali-oxygen
much greater degree than only adding Mg ions. This degradation of cellulose. Pulp Paper Can. 75:50–53.
observation is consistent with a significant proportion of Ericsson, B., Lindgren, B.O., Theander, O. (1971) Factors influenc-
the binding sites in an untreated pulp being unavailable ing carbohydrate degradation during oxygen-alkali bleaching.
Svensk Papperstid. 74:757–765.
for maximal Mg-complex formation.
Gibson, A., Wajer, M. (2003) The use of magnesium hydroxide as
• Mg addition improves the selectivity of an acid-washed
an alkali and cellulose protector in chemical pulp bleaching.
pulp. This result is consistent with the view that all sites Pulp Paper Can. 104:28–32.
for Mg-complex formation are being made available to Gibson, A., Wajer, M., Schmidtchen, P. (2004) The use of Cell-
allow maximum possible selectivity increase. guardTMOP magnesium hydroxide slurry as a cellulose protector
in oxygen delignification. In: Proceedings of PAPTAC 90th
Annual Meeting, Montreal, Canada. pp. B1409–B1413.
Gilbert, W., Pavlova, E., Rapson, W.H. (1973) Mechanism of mag-
Conclusions nesium retardation of cellulose degradation during oxygen
bleaching. Tappi J. 56:95–99.
Gustavsson, R., Swan, B. (1975) Evaluation of the degradation of
• For all the pulps investigated in this work, soluble Mg cellulose and delignification during oxygen bleaching. Tappi J.
provided a selectivity gain when added before the oxygen 58:120–123.
delignification stage. Insoluble Mg(OH)2 does not Lapierre, L., Bouchard, J., Paleologou, M., Berry, R. (1997) The
improve selectivity and is not a good replacement for limits of metal removal from kraft pulp by acid treatment. J.
MgSO4. Pulp Paper Sci. 23:539–542.
• As both MgSO4 and Mg(OH)2 negatively affect deligni- Lapierre, L., Bouchard, J., Rancourt, I., Berry, R.M. (2003) The
fication, the process selectivity rather than the viscosity effect of magnesium ions and chelants on peroxide bleaching.
Holzforschung 57:627–633.
gain is a better indicator of any benefit. The negative
Lidén, J., Ohman, L.-O. (1997) Redox stabilization of iron and man-
effect on delignification may partly be due to deactivation
ganese in the qII oxidation state by magnesium precipitates and
of ionized free phenolic hydroxyl groups or a decrease in some anionic polymers; Implications for the use of oxygen-
lignin fragment solubility. based bleaching chemicals. J. Pulp Paper Sci. 23:193–199.
• In-situ iron concentration in unbleached pulp affects oxy- McDonough, T.J. (1996) Oxygen delignification. In: Pulp Bleach-
gen delignification selectivity while in-situ calcium con- ing: Principles and Practice. Eds. Dence, C.W., Reeve, D.W.
centration affects the efficiency of Mg protection. TAPPI Press, Atlanta. pp. 213–239.

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM
 Article in press - uncorrected proof
Magnesium in oxygen delignification 301

Reitberger, T., Gierer, J. (1988) Chemiluminescence as a means to Rowlandson, G. (1971) Continuous oxygen bleaching in commer-
study the role of hydroxyl radicals in oxidative processes. Holz- cial production. Tappi J. 54:962–967.
forschung 42:351–356. Sjöström, E. (1981) The chemistry of oxygen delignification. Paperi
Robert, A., Viallet, A. (1971) Degradation of polysaccharides by Puu 63:438–442.
oxygen in the presence of alkali: protective action of magnesium Sjöström, E., Valttila, O. (1972) Inhibition of carbohydrate degra-
salts and of silicates. ATIP Rev. 25:237–241. dation during oxygen bleaching. 1. Comparison of various addi-
Robert, A., Rerolle, P., Viallet, A., Martin-Borret, C. (1964) Possible tives. Paperi Puu 54:695–700.
use of oxygen in bleaching paper pulps. 1. Improvements con- Wiklund, L., Ohman, L.-O., Lidén, J. (2001) Solid solution for-
tributed by using a catalyst: optimal conditions for the treatment mation between Mn(II) and Mg(II) hydroxides in alkaline aque-
with oxygen. ATIP Bull. 18:151–165. ous solution. Nord. Pulp Pap. Res. J. 16:240–245.
Robert, A., Viallet, A., Rerolle, P., Andreolety, J.-P. (1966) Improve-
ments in the bleaching of cellulosic pulps with oxygen and sodi- Received August 20, 2010. Accepted February 3, 2011.
um hydroxide. ATIP Rev. 20:207–217. Previously published online March 14, 2011.

Brought to you by | Western University

Authenticated
Download Date | 3/6/15 10:44 PM