Natural Surfactants for Flotation Deinking in Paper Recycling

R. A. Venditti, O. J. Rojas, H. Morris, J. Tucker, K. Spence, C. Austin, L. G. Castillo*

Forest Biomaterials Science and Engineering, North Carolina State University, USA and *Department of Wood
Pulp and Paper, University of Guadalajara, Mexico


ABSTRACT

The objective of this research was to evaluate new types of surfactants based on renewable materials (sugar and
protein-based surfactants) for use in ink removal from recycled paper via flotation deinking. By applying green
chemistry approaches we aim to minimize the environmental impact of deinking agents and also to open an avenue
for a number of products that are to be generated from the utilization of biomass. Foamability and foam stability by
the respective surfactants were considered and detergency experiments via piezoelectric sensing were used to unveil
their fundamental differences in terms of surface activity. Lab scale flotation deinking efficiency was measured
primarily by image analysis and flotation yield determined gravimetrically. We demonstrated that sugar-based
surfactants are viable replacements to petroleum-based surfactants in flotation. Differences in flotation efficiency
could be explained based on foaming, detergency and adsorption characteristics of the surfactants.


INTRODUCTION

In paper recycling, deinking operations are used to remove the ink from the recovered paper by washing and
flotation processes. In flotation, the ink is separated from the fibers by the injection of air in the presence of a
foaming agent (McCool, 1993). Rising bubbles (foam) carry away the ink particles which are separated from the
top of the flotation vat (Figure 1). Some fibers are lost in the reject stream or froth (foam and ink) and therefore the
fiber yield is less than 100%. Likewise, some ink particles remain in the fiber accepts and therefore the final paper
quality depends on the selectivity of the separation process. The key operational steps in flotation involve the use of
surfactants to ensure the detachment of the ink from the fibers (detergency) and the formation of a stable foam that
can be separated from the pulp during the flotation stage. In a typical process the amount of surfactant used is about
0.025-0.25 % based on oven-dry fiber mass.

The effect of the surfactant on the attachment of ink and other hydrophobic particles to the air bubbles is complex.
The surfactant can exist at the ink-water interface, the ink-air interface and the air-water interface, changing the
surface characteristics. There has been some recent interest in flotation processes in which the surfactant is sprayed
on the top of the flotation cell rather than mixed with the pulp prior to flotation (DeLozier et al, 2005). Laboratory
and pilot plant experiments have demonstrated improvements in the flotation process. It is suggested that mixing the
surfactant into the pulp before flotation may cause surfactant to adsorb to the pulp fibers making them more
hydrophobic and adsorb to the ink particles making them more hydrophilic. This decreases the affinity of the air
bubbles to the ink relative to the fibers and lowers flotation deinking efficiency and fiber yield. In the spray
surfactant process this indiscriminant adsorption of surfactants on fiber and ink is reduced.

Rejects:
foam + ink
Accepts: Pulp
suspension
(deinked Pulp)
air
Pulp + Surfactant
Rejects:
foam + ink
Accepts: Pulp
suspension
(deinked Pulp)
air
Pulp + Surfactant
Figure 1: A simple flotation
cell for recycled paper
Foaming is another key performance attribute of flotation surfactants. It has been found that surfactants have the
ability to promote foaming above and below their critical micelle concentration (a concentration in which the
surfactants aggregate into micelles in the bulk of the solution) (Borchardt, 1992). It has been found for linear alkyl
benzene sulfonates that the foaming activity of the surfactant increases with length of the alkyl chain. It has also
been shown that anionic surfactants tend to foam more readily than non-ionic ones with the same hydrophobe.

Two of the most common types of surfactants used for deinking are fatty acid soaps and polyethyleneoxide alkyl
ethers, which are classified as anionic and non-ionic surfactants, respectively. The case of nonylphenol ethoxylates
(NP) is particularly significant. Nonylphenol is a mixture of isomeric monoalkyl phenols, predominately para-
substituted, found in the environment primarily as a biodegradation product of nonylphenol ethoxylates.
Nonylphenols used as a nonionic surfactant result in its release to the environment through various waste streams in
paper recycling operations. The National Library of Health (NIH) reports these surfactants as severely irritating to
skin and eyes if present in high concentrations (International Labour Office, 1998). NIOSH (NOES Survey 1981-
1983) has statistically estimated that 306,211 workers are potentially exposed to nonylphenol in the US (NIOSH,
1983). Monitoring data indicate that the general population may be exposed to nonylphenol via dermal contact or
ingestion of water containing nonylphenol (Lewis, 97). Nonylphenols are suspected to be endocrine disruptors,
meaning they have adverse effects on the workings of the endocrine system in humans and animals (Ren, 1997).
Many European countries have banned the use of NP at given concentrations. No recycling mill in Europe is using
NP and most mills are asking for readily biodegradable deinking agents. Therefore, there is an important need to
understand the behavior environmentally-friendly surfactants in the paper deinking process.

EXPERIMENTAL

Surfactant Materials
The main deinking agents discussed in this study involve alkyl phenol ethoxylates, sugar-based, protein-based and a
proprietary surfactant mixture. The ethoxylated surfactant consisted of octylphenol ethoxylated with an average
number of ethylene oxide units per molecule of 9.5 (CASR No: 9002-93-1) with 33 mN/m surface tension and 108
mm Ross Miles foam at 5 min (0.1%, 25 C in DI water). The sugar-based surfactant was an alkyl (C10-C16) mono
and oligomeric D- glucopyranose (CASR No. 110615-47-9) with 28.3 mN/m surface tension and 110 mm Ross
Miles foam at 5 min (0.1%, 25 C in DI water). The protein-based surfactant is derived from soybean and is a
polymer composed of around 25 different types of amino acids linked with amide bonds with a weight averaged
molecular weight of around 100k-300k. The amino acids contain a basic amino group (NH2) and an acidic
carboxyl group (COOH). The polymer is amphoteric in solution and has a net anionic charge. Only at pH values
above 7 is the polymer fully soluble, with the acid groups ionized under alkaline conditions (this surfactant has
negligible foaming ability under the conditions of the Ross Miles standard). A proprietary commercial flotation
deinking aid intended for mixed office recovered paper based on a non-ionic surfactant blend was used as a
comparison (42.5 mm Ross Miles foam at 5 min (0.1%, 25 C in DI water).
All of these surfactant mixtures were received as donations from different chemical suppliers. The experimental
approach involved the measurement of the deinking ability for a standard printed (recycled) paper. A fundamental
study was conducted on surfactant activity that involved detergency, foamability and foam stability tests. The
application of the surfactants in bench-scale operations by flotation deinking was performed thereafter.

Procedures

Detergency
We studied the detergency ability of the different surfactants by using a model ink. Ink in printed paper is the main
component of concern in the recycling operation, i.e., a number of energy-intensive steps is aimed at removing ink
from the substrate (printed paper). The use of model inks deposited on flat, smooth surfaces allowed us to
understand, from first principles, the fundamental differences between the surfactants tested. Tripalmitin, a fatty
acid, was chosen to create model thin films of ink on gold surfaces. Gold was chosen due to the fact that
complementary techniques such as piezoelectric sensing (via the Quartz Crystal Microbalance or QCM) are based
on sensors coated with this metal. Vacuum sublimation was performed in order to deposit a uniform layer of the
model ink on the gold surface using the setup in Figure 2. A constant temperature in the sublimation chamber of
115C (above the sublimation temperature for tripalmitin) was ensured by reflux condensation (acetic acid). The
chamber was subject to negative pressure and the effective temperature on the gold surface was 75C.



We allowed sublimation times ranging from 30 to 220 minutes in order to optimize the coating procedure and to
ensure good reproducibility in the preparation of the model inks. After sublimation, the coated (printed) surfaces
were subjected to detergency tests by using different types of surfactants in the setup shown in Figure 3. Evaluation
of detergency effectiveness was performed by measurement of contact angles with water on the coated samples
(Figure 4) for different surfactant (detergency) treatment times. Surfactant treatment was carried out until the
contact angle with a water drop probe reached a constant value. Following, a washing (rinsing) sequence was
performed by exposing the surface to surfactant-free solution (water) and the contact angle was then measured again.




Figure 4: Contact angle
measurement apparatus.
Contact angle between a
drop of water and the
coated (printed) surface
was measured to
determine the quality of
the ink film after various
treatments.
Milli-Q
Water
Source
Gold
Wafer
Light
Source
Measurement
Observation
Condenser
Vacuum
Chamber
Ink Source
Gold
Wafer
Figure 2: Experimental setup to produce model
ink films for subsequent detergency
experiments. The picture shows the vacuum
sublimation chamber (left) and a view of the
gold surfaces close to the tripalmitin (ink)
source (right).
Heat
Source
Surfactant
solution
printed
gold wafer
Figure 3: Experimental setup for
surfactant treatment (detergency
tests). The beaker contains a gold
substrate coated (printed) with the
model ink. The substrate was
exposed to a surfactant solution
under shear (stirring) and then was
rinsed with water.
A Quartz Crystal Microbalance with Dissipation (QCM-D, Q-Sense, Sweden) was used to measure the changes in
film (ink) mass due to the detergency effects by the various surfactants. Substrates (QCM resonators) were
subjected to vacuum sublimation as explained before and placed in the QCM flow module to measure the change in
resonant frequency before and after the addition of surfactant solution as well as after rinsing with water. The
monitored change in frequency was related to the mass uptake or release.

Foamability and Stability
To test the foamability of different surfactants we used dynamic and static methods. In the static method of Ross-
Miles we used a pipet with 0.1% surfactant solution and a receiver with same solution, ASTM standard D1173-
53(ASTM, 2001) . The pipet was positioned at the top of the receiver and stopcock opened. When all of the solution
has run out of the pipet, we took a reading of the foam height generated at the end of 5 min. This height is
proportional to the volume of air remaining in the foam and therefore is an indication of the foamability. The values
measured are reported in the Surfactant Materials section above. In the dynamic method a 400 ml sample of the
surfactant solution (0.025g/L) was placed in a 2L graduated cylinder, and air was bubbled into the sample through a
dispersing stone at 185 ml/min (see Figure 5). This produced foam in the graduated cylinder which rose at a
decreasing rate until it reached a maximum height and began to collapse. The height of the foam was recorded every
20 seconds. The maximum height of the foam was used as a measure of the foamability of the surfactant solution.


Flotation Deinking
Recycled Husky Xerocopy paper of 92 brightness and 20 lb/75 gsm (manufactured by Weyerhaueser) which
contained 30% recycled fiber was copied with a Dialta Di 3510 copy machine with toner from Konica Minolta, MT
Toner 303A. Text was printed on both sides of the paper, 12 point font, single spaced, 1 inch margins. Pulping of
the printed paper was performed in a Tappi British Disintegrator for ten minutes at 3% consistency. Flotation
deinking was performed in a Wemco Laboratory Flotation Machine equipped with the Wemco 1+1 rotor disperser
with 2000 gram cell tank (Figure 6). Handsheets were made in a standard Tappi handsheet mold and pressed using
a standard Tappi press method. The ink content and brightness were measured after conditioning at 50% relative
humidity and 23
o
C.


Figure 6. Wemco Laboratory Flotation Machine and simplified schematic (not to scale).

Figure 5. Foam tests
conducted in a
graduated cylinder with
dispersing stone.
Foam removed
only above
this level


Surfactant was added to the printed paper at one of two addition points, either to the raw fibers and water prior to
pulping in the British disintegrator (step 2 below) or with the pulp slurry in the flotation cell before loading the
Wemco device (step 4 below).

Six surfactants were evaluated for de-inking efficiency and yield. Four different surfactant addition levels of 0.1%,
0.25%, 0.5%, or 0.75% based on oven dry fiber were evaluated. Lower levels of surfactants were trialed but in
many cases did not generate enough foam.

Details of the procedure follow.
1. Shred 11 sheets of printed paper (50 OD grams) in a paper shredder and place in British Disintegrator.
2. Add 1.61 L of tap water at 52
o
C and the surfactant if required (about 3% consistency) and pulp for 10
minutes at 55% on the variable speed motor (approximately 1570 RPM), checking the slurry after two
minutes to ensure all fibers are being pulped.
3. After pulping, wash disintegrator with 3.34 L water at 52
o
C into a five gallon bucket to dilute to 1%
consistency.
4. Measure 2 L of the well-mixed sample and place into a tared flotation cell vat. If required, add the
surfactant to the sample. Place vat in the flotation device.
5. With the air flow valve closed, set Wemco rotor at 900 rpm and allow mixing for 30 seconds.
6. At 30 seconds, open the air valve to introducing air to the cell.
7. Allow foam to rise to the top of the flotation cell. Scrape off foam and ink that is only above the overflow
lip of the flotation cell. Leave foam in the cell below the lip. The flotation experiment was stopped once
foam ceased to exist above the flotation cell lip (flotation times ranged from 60 to 210 seconds).
8. Stop the Wemco machine. Clean the fibers and ink from the mixer shaft with a spray bottle to knock them
back into the cell and weigh the flotation cell. Test the flotation cell for consistency (by means of a filter
pad) and calculate the yield from the deinking process. Make three 3 OD gram handsheets from the
deinked slurry, press and air dry in a conditioning room, 50% RH and 23 C.
9. Test the handsheets for dirt count and ISO brightness. The Apogee SpecScan 2000 image analysis system
was used with operating parameters of 2 handsheets with both sides scanned, 10 cm
2
areas scanned, 256
grey scale values (GSV), threshold of 80% of the average GSV, 0.007 mm
2
smallest particle size detected.
ISO Brightness was performed on the Technidyne Color Touch Spectrophotometer.

Quartz Crystal Microbalance
A Quartz Crystal Microbalance with Dissipation monitoring, QCM-D (Q-sense D-300, Sweden) was used to study
surfactant adsorption and activity on model thin films of ink deposited on quartz/gold electrodes.

QCM-D consists of a thin plate of a piezoelectric quartz crystal, sandwiched between a pair of electrodes. It
measures simultaneously changes in resonance frequency, f, and dissipation, D (the frictional and viscoelastic
energy losses in the system), due to adsorption on a crystal surface. Mechanical stress causes electric polarization in
a piezoelectric material. The converse effect refers to the deformation of the same material by applying an electric
field. Therefore, when an AC voltage is applied over the electrodes the crystal can be made to oscillate. Resonance
is obtained when the thickness of the plate is an odd integer, n, of half wavelengths of the induced wave, n being an
integer since the applied potential over the electrodes is always in anti-phase. If something is adsorbed onto the
crystal, it can be treated as an equivalent mass change of the crystal itself. The increase in mass, m, induces a
proportional shift in frequency, f. This linear relationship between m and f was for the first time demonstrated by
Sauerbrey (Rodahl, 1995).

n
f C
n f
f
n f
f t
m
q q q q
=

=

=
2
0
0
2


where
q
and v
q
are the specific density and the shear wave velocity in quartz respectively; t
q
is the thickness of the
quartz crystal, and f
0
the fundamental resonance frequency (when n =1). For the crystal used in these measurements
the constant C has a value of 17.7 ng cm
2
Hz
1
. The relation is valid when the following conditions are fulfilled: (i)
the adsorbed mass is distributed evenly over the crystal. (ii) m is much smaller than the mass of the crystal itself
(<1%), and (iii) the adsorbed mass is rigidly attached, with no slip or inelastic deformation in the added mass due to
the oscillatory motion. The last condition is valid when the frequency decreases in proportion to the true mass of the
adsorbate with no change in energy dissipation. Variations in the energy dissipation upon adsorption thus reflect the
energy dissipation in the adlayer or at its interface. We note that the mass detected with the QCM-D device includes
any change in the amount of solvent that oscillates with the surface.

RESULTS AND DISCUSSION

Detergency- Sublimation curves were generated to monitor the amount of ink on the surfaces and how the amount
differs with distance from the ink source. Figure 7 shows that the maximum contact angle was achieved in
approximately 1.5 hours.

The distance of the sample from the source had no impact on the quality of the model ink (data not shown). Another
experiment using the bulk addition of the model ink by manually coating resulted in a surface with a contact angle
of water on tripalmitin coated gold of 84. The typical contact angle obtained from vacuum sublimation was 72.
However, sublimation was performed in order to better mimic a thin layer of printed ink on the substrate and to
provide a homogeneous, reproducible and ultra-thin film, more suitable for the QCM experiments.

The results from experiments with surfactant treatment and rinsing sequence are shown in Figure 8 in terms of
changes in the contact angle of the printed substrate. The protein-based surfactant had a significant decrease in
contact angle. This result is explained by the hydrophobic end groups of the protein-based surfactant strongly
attaching to the ink surface and the hydrophilic groups making the surface hydrophilic, resulting in a contact angle
smaller than that of the ink surface of 84
o
. The commercial surfactant mixture had the smallest change in contact
angle because the rinsing step removed more of the surfactant. An experiment with water without a surfactant was
also performed; it was found that the contact angle of water on the ink film did not change due to the treatment (data
not shown). This indicated that fluid flow and transport effects associated with procedures were negligible.

In flotation, it is desirable for the contaminant to be hydrophobic (i.e., have a very high contact angle with water). If
the adsorption of surfactant to the ink is an important factor to determine flotation efficiency of an ink, then
surfactants that promote a high contact angle in Figure 8 would be expected to have higher efficiency.

Figure 7: Vacuum sublimation curve. The
broken line at ca. 85 degrees shows the
maximum attainable angle with a sample
simply coated with tripalmitin.
0
10
20
30
40
50
60
70
80
90
0 20 40 60 80 100 120 140 160 180 200 220
Time (min)
C
o
n
t
a
c
t

A
n
g
l
e

(
d
e
g
r
e
e
s
)
0
10
20
30
40
50
60
70
80
90
0 20 40 60 80 100 120 140 160 180 200 220
Time (min)
C
o
n
t
a
c
t

A
n
g
l
e

(
d
e
g
r
e
e
s
)


The contact angle change (% change) after surfactant treatment and after washing with water (rinsing) sequence is
shown in Table 1. The surfactant percent change was calculated using the initial contact angle and the
equilibrium contact angle after surfactant treatment. The rinsing percent change was calculated using the initial
contact angle and the equilibrium contact angle after the rinsing sequence. The protein-based surfactant produced
the highest percent change after rinsing, mainly because the surfactant was not removed during rinsing, resulting in a
lower contact angle (due to the net hydrophilic character of the surface after surfactant adsorption). This suggests
that the protein based surfactant would have low flotation efficiency (see later).


Surfactant adsorption onto the bare gold surface was also studied to provide control experiments of the sub-surface
(gold) (see Figure 9). The commercial surfactant rapidly adsorbed onto the gold and reached equilibrium with an
increase in contact angle of 13
o
. The protein-based surfactant reached equilibrium slower than the sugar-based and
commercial surfactants. The sugar-based surfactant experienced a contact angle change of 15
o
whereas the synthetic
surfactant produced a change in contact angle of 21
o
. These differences reveal the effect of surfactant structure and
nature on the adsorption, wetting and detergency behaviors.

Table 1: Surfactant Treatment Results
Surfactant type
Surfactant
% change
Washing
% change
Commercial 30.43 15.94
Synthetic 50.00 30.00
Protein-based 40.28 41.67
Sugar-based 34.72 16.67

Figure 8: Changes in contact angle (CA)
after treatment of printed surfaces with
surfactants, before (solid symbol) and after
rinsing with water (washing, open symbols) .
20
30
40
50
60
70
80
0 50 100
Time (min)
C
o
n
t
a
c
t

a
n
g
l
e

(
d
e
g
r
e
e
s
)
Sugar-based
Synthetic
Protein
Commercial
20
30
40
50
60
70
80
0 50 100
Time (min)
C
o
n
t
a
c
t

a
n
g
l
e

(
d
e
g
r
e
e
s
)
Sugar-based
Synthetic
Protein
Commercial

Figure 10 illustrates the mechanisms of detergency: a change in wetting (contact angle) occurs first by the addition
of the surfactant followed by ink separation under high shear forces. The first step, i.e., favorable adsorption of
the surfactant on the hydrophobic substrate (ink) is critical in order to guarantee a change in contact angle and to
facilitate the separation process that follows. In this study we quantified the interaction between surfactant and ink
substrate by using a piezoelectric sensor technique, namely, the quartz crystal microbalance (QCM). Conventional
QCM sensors (quartz crystal coated with a thin film of gold) were coated via sublimation with the model ink (see
previous sections). Following, QCM experiments were performed by monitoring the vibration frequency of the
respective sensor as a function of time. These measurements allowed the elucidation of (a) the extent of adsorption
and adsorption kinetics for the respective surfactant in contact with the model ink and, (b) the affinity between
surfactant with the ink substrate. This last parameter was accessed by monitoring the mass of surfactant released
upon rinsing, as will be explained in the context of Figure 11.



Figure 11 shows the change of sensors frequency after injection of surfactant solution (note the scale is inverted).
A reduction in f indicates mass uptake by the sensor, i.e., adsorption of surfactant, while a reduction in the signal
value is related to the release of mass (surfactant desorption). As an illustration we discuss here the case of all the
surfactants tested and summarize later the overall QCM results in Table 2. The base line (data up to about 500 s)
shows the frequency for each sensor after equilibration in water. The substrates consisted of (model) ink-coated
sensors. At about 500 minutes the surfactant solution is injected followed by rinsing with water at ca. 2500 min time.

In all cases adsorption to the ink surface follows a fast adsorption kinetics and relatively high degree of binding. In
the case of the commercial surfactant mixture it is seen that after rinsing (with water) the signal returns to the
original level, within the experimental error. In the case of the other surfactants it is interesting to note a larger
degree of binding and, most importantly, the fact that upon rinsing (with surfactant-free solution) the signal doesnt
return to the original value. This phenomenon may indicate a stronger binding/affinity of the respective surfactant
with the ink substrate as compared to the case of the commercial surfactant. The mass released after rinsing is
Figure 10: Schematic illustration of detergency
mechanism by surfactant addition. The oil is the
base component of ink (e.g., tripalmitin)

ini tial state: no
surfactant
change i n
wetting
Separati on by
shear -->
(roll i ng)
substrate
oil
water

change i n
wetting
Separati on by
shear -->
(roll i ng)
substrate
ink
water
+ surfactant:

ini tial state: no
surfactant
change i n
wetting
Separati on by
shear -->
(roll i ng)
substrate
oil
water

change i n
wetting
Separati on by
shear -->
(roll i ng)
substrate
ink
water
+ surfactant:
Figure 9: Changes in contact angle after
surfactant adsorption on gold
0
5
10
15
20
25
30
0 20 40 60 80 100
Time (min)
C
o
n
t
a
c
t

a
n
g
l
e

(
d
e
g
r
e
e
s
)
Sugar
Synthetic
Protein
Commercial
0
5
10
15
20
25
30
0 20 40 60 80 100
Time (min)
C
o
n
t
a
c
t

a
n
g
l
e

(
d
e
g
r
e
e
s
)
Sugar
Synthetic
Protein
Commercial
related to the amount of surfactant that desorbs from the interface or, if ink is removed from the interface, to the
amount of ink that is released. A higher affinity (lower surfactant release) is a favorable situation in terms of the
detachment of ink from fiber process because in this case a better wetting of the substrate (see Figure 10), under
conditions of shear, is ensured. However, for the attachment of ink to air bubbles, this may not be the case.

The Sauerbrey equation can be used to relate the change in frequency to the change in mass and the calculated
values are presented in Table 2, for all surfactants considered in this study. It is interesting to note the case of
protein-based surfactant: it is well known that proteins adsorb strongly and irreversibly to a variety of surfaces. Our
results indicate agreement with this observation since it is the protein-based surfactant that shows the largest degree
of binding. The signal for the commercial surfactant mixture, after rinsing, is lower than the base line, which
indicates that a portion of the ink substrate may have been removed from the interface.


Figure 11: QCM frequency change after exposing a coated (printed) sensor (model ink) to a solution of the
respective surfactant (at ca. 500 min s) and after rinsing with water at about 2500 s time. The change in frequency,
plotted as f (note reversed scale), is proportional to the change in film mass due to surfactant uptake and/or release.


As seen in Table 2, all surfactants, adsorb to a similar extent on the model ink surface. However, the degree of
binding was quite different for the commercial mixture in that is showed the lowest degree of binding.

Table 2: Mass Change with Surfactant Treatment
Type Amount adsorbed (ng) Surfactant released (ng)
Commercial 10.117 11.3
Synthetic 12.836 3.0
Protein-based 13.675 3.1
Sugar-based 12.496 3.7

In general it is expected that surfactants that adsorb to the surface of hydrophobic particles like ink will have their
hydrophilic portion facing the water phase, stabilizing the particle in the water phase and decreasing the flotation
efficiency. However, this may not be the entire explanation for the flotation efficiencies (see later) as the data in
Table 2 does not explain why the sugar based surfactant performed as well as the commercial mixture in flotation.

Foamability- A plot of the foam height of the surfactants versus time are shown in Figure 12. It can be observed
that the protein-based surfactants made no foam under these conditions and the commercial surfactant generated
about as much foam as the sugar-based surfactant. There is a strong trend between foamability and removal
efficiency (see later), also shown in Figure 12. Since there is a strong correlation between foamability and removal
efficiency as seen in Figure 12, foamability and performance of a surfactant are closely related, and finding natural
surfactants that create the correct quantity and quality of foam should also remove more ink from paper in recycling.

F
o
a
m

H
e
i
g
h
t

(
c
m
)
0
5
10
15
20
0 500 1000
Synthetic
Sugar based
Protein based
Commercial mixture
Time
0
5
10
15
20
Synthetic
Sugar based
Protein based
Commercial mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture
F
o
a
m

H
e
i
g
h
t

(
c
m
)
0
5
10
15
20
0 500 1000
Synthetic
Sugar based
Protein based
Commercial mixture
Time
0
5
10
15
20
Synthetic
Sugar based
Protein based
Commercial mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture
0
5
10
15
20
25
0 20 40 60 80
Removal Efficiency as % of Control
M
a
x

F
o
a
m

H
e
i
g
h
t

(
c
m
)
Synthetic
Sugar based
Protein Based
Commercial
mixture

Figure 12: Left: Foam height vs time. Right: Foam height vs flotation removal efficiency.

Deinking by Flotation - Recycled paper deinking was performed as explained in the experimental section.
Detergency and foam phenomena are expected to be key factors in the overall process. Detected toner particle areas
as measured with image analysis (reported as parts per million or ppm) were used to calculate the ink removal
efficiency for the given sample (with application of the respective surfactant) compared to the control (a sample
subjected to deinking operations in the absence of surfactant). The removal efficiency is defined as:
( )
% *100
Control Sample
Control
PPM PPM
RE
PPM

=

An illustration of typical handsheet areas made from the pulped paper if no deinking is performed (control) and for
the resulting paper after flotation deinking with commercial surfactant and with sugar-based surfactant as well as the
rejects (stream rejected in the flotation cell using the commercial surfactant, rich in ink and surfactant) are presented
in Figure 13.









Figure 13: Handsheets samples produced after recycling printed paper. Left to right: Control of feed (no
surfactant treatment); sample after commercial surfactant treatment; sample after treatment with the
sugar-based surfactant and sample produced from the rejects from flotation. The residual ink on paper is
seen as black particles (dirt) quantified with an optical scanner.

The removal efficiency was plotted against the yield to determine each surfactant overall performance when added
to the flotation cell (Figure 14). Similar trends were obtained for the case of addition of the surfactant in the pulper
(Figure 15). The synthetic and sugar-based surfactants exhibited optimal removal efficiencies versus yield at both
addition points. The commercial surfactant performed markedly better when added in the flotation cell rather than
the pulper, in agreement with the intended addition point of the supplier. The protein based surfactant had the least
desirable removal efficiency versus yield performance when added to the pulper or the flotation cell. The flotation
results are plotted in terms of removal efficiency versus surfactant charge in Figure 16. The protein based
surfactant has a significantly lower removal efficiency versus surfactant charge than all of the others. The
decreased performance of the protein-based surfactant can be explained by the following two observations. The
protein based surfactant produced the lowest amount of foam (Figure 12). Further, the protein based surfactant had
the highest affinity/adsorption for the model ink (Figure 11) and produced the lowest contact angle with water after
adsorption to the model ink surface (Figure 8). This indicates that the protein based surfactant is acting as a
resistance to effective air bubble-toner contact that is required in flotation. A similar finding has been reported for
cationic starch and toner particle agglomeration by Venditti and coworkers (Zheng et al, 1999 and 2001) and by
Berg and coworkers (Snyder and Berg, 1994). In that case, starch adsorbed onto the surface of toner particles acts as
a hindrance for toner-toner contact which is required for agglomeration, similar to the air-toner contact required in
flotation. Similar results were found for starch in water interfering with acrylic micro sphere-acrylic microsphere
contact (Huo et al, 2001) and acrylate particle-polyester fiber contact (Huo et al, 1999).

Surfactant Added in Flotation Cell
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
50 55 60 65 70 75 80 85 90 95 100
Yi eld (%)
R
e
m
o
v
a
l

E
f
f
.

(
%
)
Sugar based
Commercial mixture
Protein based
Synthetic

Figure 14. Removal efficiency versus yield for surfactants added in the flotation cell.

Surfactant Added in Pulper
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
50 55 60 65 70 75 80 85 90 95 100
Yiel d (%)
R
e
m
o
v
a
l

E
f
f
.

(
%
)
Sugar based
Protein based
Synthetic
Commercial mixture

Figure 15. Removal efficiency versus yield for surfactants added in the pulper.


Surfactant Added to Flotation Cell
0
10
20
30
40
50
60
70
80
90
100
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Surfactant Charge (% on OD fiber)
R
e
m
o
v
a
l

E
f
f
.

(
%
)
Sugar based
Commercial mixture
Protein based
Synthetic

Figure 16. Removal efficiency versus surfactant charge for surfactants added in the flotation cell.

CONCLUSIONS

The alteration of a model ink surface can be investigated by measuring contact angles of water on the model ink
after exposure to different surfactant solutions. Adsorption of surfactant onto model ink film can be monitored
using a QCM technique. The ability of the surfactants to produce foam was positively correlated to the flotation
efficiency. The efficiency of a flotation cell (with respect to ink removal and process yield) was very sensitive to the
surfactant chemistry utilized. It was demonstrated that a sugar-based surfactant had flotation ink removal efficiency
versus overall yield that was similar to conventional surfactants. A protein surfactant that had low foamability and
adsorbed to the ink surface rendering the surface much more hydrophilic has a very low flotation ink removal
efficiency versus overall yield, confirming that surfactant adsorption and foaming phenomena are important in the
flotation deinking process.

ACKNOWLEDGMENTS

This research was funded by the NCSU Undergraduate Research Program and EPA P3 funds. Support by University
of Guadalajara for visiting research experience of Luis Castillo at NCSU is gratefully acknowledged. We would
also like to acknowledge the assistance of Dr. Xavier Turon with the QCM experiments.


REFERENCES:

ASTM standard D1173-53 (Reapproved 2001).
Borchardt, J. K. Progress in Paper Recycling, 1992, 2(1), 55-63.
DeLozier, G.; Zhao, Y.; Deng, Y.; White, D.; Zhu, J.; Prein, M. Tappi 2005, 4(10), 25-30.
Huo, X.; Venditti, R. A.; Chang, H. M. TAPPI Proceedings of the 1999 Recycling Symposium, 1999, 681-692.
Huo, X.; Venditti, R. A.; Chang, H. M. Journal of Pulp and Paper Science, 2001, 27(6), 207-212.
International Labour Office. Encyclopaedia of Occupational Health and Safety. 4th edition, Volumes 1-4 1998.
Lewis, RJ Sr, ed; Hawley's Condensed Chemical Dictionary. 13th ed. NY, NY: John Wiley and Sons, Inc, p. 805
(1997).
McCool, M. A. Flotation Deinking. In Secondary Fiber Recycling; Spangenberg, R. J. Ed.; Tappi Press: Atlanta,
1993; pp 141-161.
NIOSH; National Occupational Exposure Survey (NOES) (1983).
Rodahl, M.; Hook, F.; Krozer, A.; Brzezinski, P.,; Kasemo, B. Review of Scientific Instruments 1995, 66(7): 3924-
3930.
Ren, L., Marquardt, M.A., Lech, J.J. Chem Biol Interact , 1997, 104 (1), 55-64.
Snyder, B. A.; Berg, J.C. Tappi Journal, 1994, 77(5), 79-84.
Zheng, J.; Venditti, R. A.; Olf, H. G. Progress in Paper Recycling, 1999, 9(1), 30-37.
Zheng, J.; Venditti, R. A.; Olf, H. G. Journal of Pulp and Paper Science, 2001, 27(3), 98-102.