Journal of Colloid and Interface Science 341 (2010) 298302

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science

www.elsevier.com/locate/jcis

Adsorption of sodium polyacrylate in high solids loading calcium carbonate slurries

Joshua J. Taylor a, Wolfgang M. Sigmund a,b,*
a
b

Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, United States
Department of energy Engineering, Hanyang University, Seoul, South Korea

a r t i c l e

i n f o

Article history:
Received 3 June 2009
Accepted 24 September 2009
Available online 4 October 2009
Keywords:
Calcium
Carbonate
Slurry
Infrared
Polyacrylate
Carboxylate
Adsorption
Solids
ATR-FTIR

a b s t r a c t
The adsorption of sodium polyacrylate (NaPAA) in slurries with up to 75 wt.% calcium carbonate was
investigated with the use of attenuated total reectance-Fourier transform infrared spectroscopy (ATRFTIR) and adsorption of probe molecules. Analysis of the IR spectra demonstrated that the carboxylate
groups of NaPAA adsorbed onto ground calcium carbonate (GCC) in three different modes. These modes
were shown to be dependent on the solids loading and age of the slurry. Further investigation lead to the
determination of the chelating ability of NaPAA at high solids loading.
2009 Elsevier Inc. All rights reserved.

1. Introduction
Polyacrylic acid and its salt sodium polyacrylate (NaPAA) are
one of the most frequently applied polyelectrolytes in industry
and the household. A few examples include laundering processes,
thickening agents, and dispersion of clay and calcium carbonate.
NaPAA is used in the calcium carbonate industry as a dispersant
for the mineral because it allows for an increase in the solids loading up to 75 wt.% while maintaining the desired viscosity. However,
the complete role of stabilization and conrmation of NaPAA in dispersing calcium carbonate at high solids content is not clear.
The carboxylate group has been shown to interact with metal
cations and surfaces in four different modes [16], namely, ionic,
bridging, bidentate, and unidentate. Several papers have investigated the different carboxylate modes with IR spectroscopy because the change in bond symmetry can be detected. The ionic,
bridging, and bidentate modes have similar group symmetry. The
two oxygen atoms in the bidentate mode are interacting with
one metal cation; therefore, there will be a change in the symmetrical and asymmetrical vibration frequencies. Since the unidentate
mode has one oxygen atom coordinated with one metal cation the
symmetrical and asymmetrical vibration frequencies will be similar to a carboxylic acid group.
Geffroy et al. [2] and Dobson and McQuillan [1] have shown
that adsorption is preferred through chelation of dicarboxylates.
* Corresponding author. Address: Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611, United States. Fax: +1 352 846 3355.
E-mail address: wsigm@mse.u.edu (W.M. Sigmund).
0021-9797/$ - see front matter 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcis.2009.09.048

A ve-member chelate ring (consisting of the metal cation and

dicarboxylates) are the most stable followed by six- and sevenmember chelate rings. Lu and Miller [4] further explains that carboxylate groups interact with calcium ions in three-dimensional
seven- or eightfold coordination and it is common for calcium ions
to coordinate with both carboxylate groups and water. Katz et al.
[7] also demonstrate that both unidentate and bidentate coordination of carboxylate groups with calcium cations are possible when
the calcium ions bind in seven- and eightfold coordination.
The objective of this paper is to investigate the effect of solids
loading and aging of calcite slurries on the adsorption of NaPAA onto
GCC. The attenuated total reectance-Fourier transform infrared
(ATR-FTIR) technique was employed to collect data of ground calcium carbonate (GCC) slurries up to 75 wt.%. Previous studies have
focus on the interaction of carboxylate groups with calcium in dilute
systems but did not consider changes in the interactions within high
solids loading slurries or changes due to aging of slurries. All the differences between dilute systems and high solids loading slurries
have not been accounted for by the sciences which cause problems
for industry on a daily basis. This paper is the rst to address the
adsorption of NaPAA onto GCC in slurries up to 75 wt.%.
2. Materials and methods
2.1. Materials
Ground calcium carbonate (GCC) was provided by Imerys
(Sandersville, Georgia) with a d50 = 1 lm measured with a sedi-

299

J.J. Taylor, W.M. Sigmund / Journal of Colloid and Interface Science 341 (2010) 298302

graph. The sodium polyacrylate (NaPAA) with molecular weight

average Mw = 5967 g/mol (polydispersity 2.04) was provided by
Kemira Chemicals, Inc. (Kennesaw, Georgia). Benzoic acid, 99.5%,
gallic acid, >99%, and deuterium oxide (D2O), 100.0 at.% D, were
purchased from Fisher Scientic (Acros Organics). D2O is used in
this research because the IR spectrum of D2O allows for analysis
of certain NaPAA bands when compared to H2O bands which overlap critical NaPAA bands.

0.15

Log (1/R) (a.u.)

NaPAA in D2O

2.2. Attenuated total reectance-Fourier transform infrared (ATRFTIR)

Infrared spectroscopy detects the vibrational characteristics of
chemical groups having a dipole moment. Unlike other infrared
techniques (DRIFTS and transmittance) that require dry or transparent samples, the ATR technique allows for measurement of liquids, gels, and solids. ATR-FTIR measurements were performed
using the Thermo Electron Magna 760 FTIR Microscope with a ZnSe
crystal and a liquid nitrogen cooled MCT/A detector. FTIR spectra
were recorded from 650 cm 1 to 4000 cm 1 with 128 scans and
a resolution of 2 cm 1.

0.05

0.00
1750

1700

1650

1600

1550

1500

Wavenumber (cm-1)
Fig. 1. IR spectrum of the carboxylate stretching region for NaPAA in D2O. The IR
spectrum shows that NaPAA is in the ionic form with a peak at 1570 cm 1 for the
COO and no peak at 1700 cm 1 for the C@O.

2.3. Adsorption measurements

NaPAA in D2O
75 wt% GCC Slurry

0.0010

0.0005

d A/d (a.u.)

Slurry samples were centrifuged at 3000 rpm for 45 min. The

supernatant was then decanted into a beaker and diluted with a
measured amount of water. UltravioletVisible Spectroscopy
(UV/Vis) was utilized to measure the adsorption intensities at
wavelengths of 210 nm and 224 nm for gallic acid and benzoic
acid, respectively, with the Beckman DU 640 spectrophotometer.
Next, the measured intensities were compared to a calibration
curve and the amount of probe molecules in the supernatant was
determined. The amount of probe molecules adsorbed was calculated from subtracting the measured amount in the supernatant
from the amount in the slurry.

0.10

0.0000
bidentate
bridging

-0.0005

-0.0010
unidentate

ionic

2.4. Sample preparation

1600
NaPAA was dissolved in D2O. Amount of NaPAA in a sample was
1 wt.% of the GCC weight. Next, the GCC was slowly added to the
solution while stirring. The weight percent of GCC is a percent of
the total GCC and water weight. The aging samples used for testing
were stored in a sealed container during the aging period without
agitation.
Preparations of slurries containing benzoic acid or gallic acid
were prepared similarly. Amount of benzoic acid or gallic acid in
a sample was 1 wt.% of the GCC weight. Either benzoic acid or gallic
acid was dissolved into D2O and NaOH was added until a pH of 9.9
was reached. Next, the GCC was slowly added to the solution while
stirring.
3. Results and discussion
3.1. Ionic and adsorbed NaPAA
When NaPAA is dissolved into water the carboxylate group will
have a resonance form. This was conrmed in the IR spectra with
the absence of the C@O band which would be located around
1700 cm 1 and the presence of a COO band at 1570 cm 1 as seen
in Fig. 1. When calcium carbonate is introduced into the system
with 75 wt.% solids loading there is a change in the carboxylate
band. Fig. 2 shows that the 1570 cm 1 stretching band is split into
a band at 1581 cm 1 and 1567 cm 1 along with a band forming at
1524 cm 1 with a shoulder. From the research of Lu and Miller [4]
and Young and Miller [6] they have shown that each band repre-

1580

1560

1540

1520

1500

Wavenumber (cm-1)
Fig. 2. Second derivative of the IR spectra of the carboxylate region of NaPAA in D2O
and in a 75 wt.% GCC slurry. NaPAA adsorbs onto GCC in unidentate, bridging, and
bidentate modes.

sents a different mode of interaction. The four possible modes of

interaction are ionic, unidentate, bidentate, and bridging. Mielczarski et al. [8,9] assigned the bands at 1575 cm 1 and 1540 cm 1 to
unidentate and bidentate adsorption, respectively. Therefore, in
Fig. 2 the 1581 cm 1 band is representative of unidentate coordination, the 1567 cm 1 band represents bridging coordination,
and the bands between 1510 cm 1 and 1545 cm 1 represent
bidentate coordination with calcium. In high solid loading slurry
the adsorption of NaPAA onto GCC consists of all three coordination modes.
3.2. Increase solids and aging
As discussed previously, the adsorption of NaPAA occurs
through three coordination states but there are apparent changes
in each state with different solids loading. These differences are
demonstrated by preparing slurry samples with 10, 30, 50, and
70 wt.% GCC and analyzing the spectra of the IR carboxylate region
(Fig. 3). First, the band representative of the unidentate coordination in the 10 wt.% GCC slurry at 1586 cm 1 shifts to lower wave-

300

J.J. Taylor, W.M. Sigmund / Journal of Colloid and Interface Science 341 (2010) 298302

unidentate

0.0005

bridging

d A/d (a.u.)

0.0004

10 wt% Slurry
30 wt% Slurry
50 wt% Slurry
70 wt% Slurry

75 wt% Slurry Fresh

75 wt% Slurry Aged 24 hrs
75 wt% Slurry Aged 48 hrs

increasing age

0.0000

d A/d (a.u.)

0.0008

0.0000

-0.0004
bidentate

-0.0008
1600

1580

1560

1540

1520

Wavenumber (cm-1)

unidentate

1600

1580

bridging

1560

1540

-1

Wavenumber (cm )

Fig. 3. Second derivative of the IR spectra of the carboxylate region with increasing
solids loading. As the solids loading of the slurries increases there is a shift of the
bands toward a bridging mode.

Fig. 4. Second derivative of the IR spectra of the carboxylate region with aging. The
unidentate band shifts from 1585 cm 1 to 1580 cm 1 with increasing age indicating
a shift towards the bridging mode.

numbers as the solids loading is increased, eventually to

1578 cm 1 in the 70 wt.% GCC slurry. Similarly, the band representative of the bridging coordination in the 10 wt.% GCC slurry at
1567 cm 1 shifts to lower wavenumbers until it reaches
1560 cm 1 in a 70 wt.% GCC slurry. Also, as the solids loading increases the band between 1510 cm 1 and 1545 cm 1 becomes a
doublet with band peaks at 1539 cm 1 and 1521 cm 1, representing the bidentate coordination. As the solids loading increases
there is an increase in the bridging to unidentate ratio. These results can be explained by a decrease in the distance between calcium carbonate particles which would decrease the distance
between calcium allowing for additional bridging. These results
are important because they demonstrate that the interactions
within a low solids loading slurry are different than in a high solids
loading slurry.
During the rst couple of days after a slurry has been prepared
there are several changes in its physical properties. These changes
were suspected to originate from differences in the adsorption of
the dispersant. So ATR-FTIR was utilized to determine if any
changes in the interaction of the NaPAA with GCC could be detected during the aging process. A 75 wt.% GCC slurry was prepared
and analyzed with the ATR-FTIR as it aged. This was performed
while it was less than an hour old, 24 h old, and 48 h old. As seen
in Fig. 4 the interaction of NaPAA with GCC changes with age. Initially, the band representing the unidentate coordination is located
at 1585 cm 1. As the system ages the band shifts to 1580 cm 1.
This indicates that while the slurry ages there is a decrease in
the concentration of unidentate coordination and an increase in
the bridging and/or ionic coordination of the carboxylate groups.
These results would support the idea that as the system ages there
is an increase in the amount of dissolved calcium ions which are
then available for bridging coordination, which require two calcium ions per carboxylate group, instead of unidentate coordination, which only require one calcium ion per carboxylate group.

trons to transition from p (bonding) to p* (anti-bonding) which

would allow for determination of their concentrations in a solution
with the use of an UV/Vis spectrometer.
Adsorption experiments for benzoic acid onto GCC were performed to demonstrate the carboxylate group adsorption and/or
the hydrophobic adsorption of the benzene ring. Adsorption of
NaPAA is believed to be due to the carboxylate groups interacting
with the calcium ions [2,4,5,10]. Geffroy et al. [2] go into detail
describing the complexation of carboxylates with the surface of
calcite required for adsorption. They determine that carboxylates
adsorb through chelation. Since it has been concluded that adsorption modes of carboxylate groups are modied with solids loading,
the adsorption of benzoic acid at different solids loading was investigated. Analysis of adsorption measurements in Fig. 5 conclude
that benzoic acid does not adsorb onto GCC at any solids loading.
Since Fig. 5 does not indicate any adsorption, then the ATR-FTIR
was utilized to conrm that the carboxylate and the benzene ring
are not interacting with the surface of the GCC.
Analysis of Fig. 6 demonstrates that benzoic acid does not adsorb onto the GCC with increasing solids loading. The carboxylate
band at 1548 cm 1 for benzoic acid in D2O does not shift or split
when GCC is added to the solution which is an indication that
the carboxylate is not adsorbing onto the GCC particles. Another
possible interaction that could cause adsorption is through hydrophobic adsorption with the benzene ring. The band for a benzene

3.3. Probe molecules

Molecules which contain a carboxylate group were chosen to
probe the surface of GCC in high solids loading slurries. The interaction of the probe molecules with the surface of GCC would provide additional understanding of the interaction between NaPAA
and GCC. Benzoic acid and gallic acid were chosen because each
molecule contains a single carboxylate group along with a benzene
ring. The benzene ring absorbs ultraviolet light causing the elec-

Fig. 5. Adsorption of benzoic acid onto GCC at varying solids loading. Benzoic acid
does not adsorb onto GCC.

301

J.J. Taylor, W.M. Sigmund / Journal of Colloid and Interface Science 341 (2010) 298302

Benzoic acid in D2O

20 wt% GCC Slurry
57 wt% GCC Slurry

0.000

d A/d (a.u.)

0.004

-0.004
COO-

-0.008
C C

1640

1620

1600

1580

1560

1540

1520

-1

Wavenumber (cm )
Fig. 6. Second derivative of the IR spectra of benzoic acid in D2O, 20 wt.% GCC
slurry, and 57 wt.% GCC slurry. The bands for the benzene ring and the carboxylate
do not shift, indicating no adsorption.

ring is located at 1596 cm 1 and does not shift with increasing

solids loading so the benzoic acid is not adsorbed through the benzene ring.
The results from the adsorption measurements and IR spectra of
benzoic acid conrm each other and are in good agreement with
Geffroy et al. [2] who demonstrated that a molecule with one carboxylate group will not adsorb onto the surface of GCC. Therefore,
the next probe molecule chosen contained OH groups which were
expected to promote adsorption onto calcite.
Gallic acid was chosen as a probe molecule for the same reason
that benzoic acid was chosen but it also includes three hydroxyl
groups. Hydroxyl groups were chosen in order to determine if chelation would play an important role in adsorption. The adsorption
of gallic acid onto GCC in several different solids loading slurries is
demonstrated in Fig. 7. The amount of gallic acid adsorbed is
dependent on the solids loading of the slurry and decreases with
an increase weight percent of GCC. Calculations were performed
using bond angles and bond lengths to determine the adsorbed
monolayer concentration of gallic acid which is between 1.6 mg/
m2 and 8.5 mg/m2. The range in adsorbed monolayer concentration
is due to the different modes of adsorption (from at to vertical)
that the gallic acid could accompany while adsorbing. Comparing
the monolayer adsorption calculations to Fig. 7 would suggest that

the adsorbed amount of gallic acid at 10 wt.% would contain at

least three adsorbed layers, at 30 wt.% would contain at least two
adsorbed layers, and at 50 wt.% and 70 wt.% would contain one
monolayer with different modes of adsorption.
Further investigation of the chelating ability of gallic acid was
performed with FTIR spectroscopy. Analysis of the IR spectra conrmed that gallic acid adsorbs onto the surface of GCC. As seen in
Fig. 8, the carboxylate band at 1550 cm 1 shifts to 1547 cm 1 when
a 20 wt.% GCC slurry is prepared. The carboxylate shift is a conrmation that the gallic acid is chelating with the surface of the GCC.
Another important band includes the benzene ring band located at
1600 cm 1 for the gallic acid and D2O system at a pH of 9.9. The IR
band for the benzene ring shifts to 1594 cm 1 for a 20 wt.% GCC
slurry containing gallic acid. This is an additional indication of chelation because the benzene ring state is changed due to its OH
groups participation in the chelate ring.
The inclusion of the hydroxyl groups allows for gallic acid to
chelate with the GCC surface forming a seven-bond ring through
one hydroxyl group, calcium ion, and carboxylate group (similar
results from Geffroy et al. [2] with different molecules). A sevenmember chelate ring is not as stable as a ve-member chelate ring
but the chelating ability of gallic acid promotes adsorption unlike
benzoic acid which does not have the ability to form a chelate with
calcium. Since it has been demonstrated that the adsorption of a
carboxylate group must be performed through chelation then the
adsorption mechanism for NaPAA onto calcium must be through
an eight-member chelate ring utilizing two adjacent carboxylate
groups. This is important because the eight-member chelate ring
is not very stable and could be an explanation for the physical
changes of the slurry with aging.
Previously, it has been concluded that the adsorption modes of
carboxylate groups are modied with solids loading; therefore, the
adsorption of gallic acid at different solids loading was also investigated. Similar to the results for NaPAA, there are changes in the
adsorption modes with increase solids loading. Fig. 9 shows a shift
of the carboxylate band at 1547 cm 1 for a 20 wt.% GCC slurry to
1560 cm 1 for a 67 wt.% GCC slurry. Along with the carboxylate
shift there is a shift of the benzene band from 1594 cm 1 to
1584 cm 1 for a 2067 wt.% slurry. Comparing Fig. 9 and Fig. 3,
both of them demonstrate an increase in the ratio of bridging to
bidentate coordination with increasing solids loading. Again, these
results can be explained by a decrease in the distance between calcium carbonate particles which would decrease the distance between calcium allowing for additional bridging.
Gallic acid in D2O
20 wt% Slurry

0.0000

d A/d (a.u.)

0.0008

-0.0008
C C

-0.0016
bridging

-0.0024
1620

1600

1580

1560

1540

bidentate

1520

Wavenumber (cm-1)

Fig. 7. Adsorption of gallic acid onto GCC at varying solids loading.

Fig. 8. Second derivative of the IR spectra of gallic acid in D2O and a 20 wt.% GCC
slurry. The bands for the benzene ring and the carboxylate shift, indicating
adsorption.

302

J.J. Taylor, W.M. Sigmund / Journal of Colloid and Interface Science 341 (2010) 298302

20 wt% Slurry
67 wt% Slurry

0.0000

bidentate

d A/d (a.u.)

0.0004

C C

-0.0004

the bridging/unidentate and bridging/bidentate ratios which are

demonstrated in both the NaPAA and gallic acid systems. Second,
as a high solids loading slurry ages there is a decrease in the concentration of unidentate coordination of carboxylate groups and an
increase in the bridging and/or ionic coordination of the carboxylate groups. This can be explained by an increase in the amount of
dissolved calcium ions with age which become available for the
unidentate coordinated carboxylate groups to form bridging coordination. Third, NaPAA was demonstrated to adsorb onto calcium
carbonate through an eight-member chelate ring consisting of
two adjacent carboxylate groups and does not adsorb through single carboxylate groups alone.

bridging

Acknowledgments

1640

1620

1600

1580

1560

1540

1520

-1

Wavenumber (cm )
Fig. 9. Second derivative of the IR spectra of gallic acid in a 20 wt.% and 67 wt.% GCC
slurry. The bands for the benzene ring and the carboxylate shift, indicating change
in the coordination of the gallic acid with increasing solids loading.

4. Conclusions
Several published works have focused on the interaction of carboxylate groups with calcium and have provided valuable insight.
But these works have been performed in dilute systems and do not
account for the behavioral differences experienced in high solids
loading slurries. This paper is the rst to address the adsorption
of NaPAA onto GCC in slurries up to 75 wt.%. First, the difference
between a dilute system and a high solids loading system is demonstrated to be due to the change in adsorption coordination of the
carboxylate. As the solids loading increases there is an increase in

This research was partially supported by Kemira Chemicals, Inc.

and Imerys Clays, Inc.
References
[1] K.D. Dobson, A.J. McQuillan, Spectrochim. Acta, Part A 55 (78) (1999) 1395
1405.
[2] C. Geffroy, A. Foissy, J. Persello, B. Cabane, J. Colloid Interface Sci. 211 (1) (1999)
4553.
[3] L.J. Kirwan, P.D. Fawell, W. van Bronswijk, Langmuir 19 (14) (2003) 5802
5807.
[4] Y.Q. Lu, J.D. Miller, J. Colloid Interface Sci. 256 (1) (2002) 4152.
[5] I. Pochard, P. Couchot, C. Geffroy, A. Foissy, J. Persello, Rev. Inst. Franc. Petrole
52 (2) (1997) 251253.
[6] C.A. Young, J.D. Miller, Int. J. Miner. Process. 58 (14) (2000) 331350.
[7] A.K. Katz, J.P. Glusker, S.A. Beebe, C.W. Bock, J. Am. Chem. Soc. 118 (24) (1996)
57525763.
[8] E. Mielczarski, J.A. Mielczarski, J.M. Cases, Langmuir 14 (7) (1998) 17391747.
[9] J.A. Mielczarski, E. Mielczarski, J. Phys. Chem. 99 (10) (1995) 32063217.
[10] M. Donnet, P. Bowen, N. Jongen, J. Lemaitre, H. Hofmann, Langmuir 21 (1)
(2005) 100108.