ARTICLE IN PRESS

Water Research 38 (2004) 33733381

The role of zeolites in wastewater treatment of printing inks

A. Metes a,*, D. Kovacevicb, D. Vujevica, S. Papica
a

Faculty of Chemical Engineering and Technology, University of Zagreb, Zagreb, Savska 16, Croatia
b
Faculty of Science, University of Zagreb, Zagreb, Marulicev trg 19, Croatia
Received 11 June 2003; received in revised form 13 February 2004; accepted 19 April 2004

Abstract
The adsorption of residual organic pollutants from occulated printing ink wastewater onto several synthetic zeolites
was investigated as a nishing method for additional reduction of TOC. The nonselective removal of total organic
content was studied.
The amount of adsorbed organics was largest for ZSM-5 and NH4-Beta while the other zeolites studied showed lower
efciency, suggesting that adsorption is independent of pore structure. The adsorption rates of organic pollutants were
fast. Although the TOC removal increases with increasing amount of zeolite, because of the necessity of additional
ltration to lower turbidity to required levels, 5.0 g/L of zeolite was found to be optimum.
The 88% reduction of TOC obtained with a single occulation treatment was improved with the combination of
occulation and adsorption with ZSM-5 which resulted in the overall TOC efciency of 95%.
The addition of zeolites in decantated supernatant water, obtained after occulation, was also studied in order to
assess the effect of oc on zeolite capacity. A decrease in adsorption capacity occurred only if a coagulant concentration
less than optimal was applied. Removal efciency then decreased by around 10%. It was concluded that occulation
followed by adsorption with zeolites is an effective treatment method for this kind of wastewater.
r 2004 Elsevier Ltd. All rights reserved.
Keywords: Water-based printing inks; Wastewater; Zeolites

1. Introduction
Synthetic and natural zeolites have become increasingly important in last 30 years due to the wide range of
their chemical and physical properties [1].
Zeolites have been used as adsorbents, molecular
sieves, membranes, ion exchanger and catalysts in
municipal and industrial pollution control, as well as
in horticulture, agriculture, and environmental soil
remediation [2]but their primary use has been in
water and wastewater treatment.
The chemical and structural features of zeolites make
them very effective for the removal of toxic metal ions
*Corresponding author. Tel.: +385-1-4597-128; fax: +3851-4597-142.
E-mail address: ametes@marie.fkit.hr (A. Metes ).

(e.g. lead, nickel, zinc, cadmium, copper, chromium and/

or cobalt), radionuclides as well as ammoniacal nitrogen
(ammonia and ammonium) from municipal and industrial wastewaters, such as from electroplating efuents
[3], agricultural wastewaters [4,5] or coke-plant efuents
[6].
The investigation of several natural zeolites for the
selective removal, and further recovery, of organicsubstituted ammonium ions showed that they are
promising for clean up and recycling of photo-nishing
and photo-developing washwaters [7].
The possibility of the successful encapsulation of
organic dyes within the channels of various zeolites by in
situ diazotation and coupling and the basic adsorption
properties of zeolites were reported [8,9]. However, it
was also reported that the zeolites were not suitable to
adsorb reactive dyes [10].

0043-1354/$ - see front matter r 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2004.04.012

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

3374

Nomenclature
G
IC
k
1=n
N
NTU
S

mean velocity gradient, s1

inorganic carbon content, mg C/L
Freundlich adsorption coefcient
Freundlich exponent
impeller speed, rpm
nephelometric turbidity unit
specic area, m2/g

The application of zeolites for printing ink wastewater

treatment has not been previously reported. Globally,
the printing ink industry is facing stringent requirements
to produce new and better performing and environmentally friendly printing inks. The reduction of volatile
organic content, the toxicity of products, and the
quantity of waste generated during production is
necessary for advancement of this industry.
The development and application of water-based
printing inks has lead to the reduction of volatile
organic compounds (VOC) emissions, as one of the
main driving forces of product innovation. However, wastewater obtained after cleaning/washing of
the laboratory and industrial equipment is characteristically high in both colour and organic content and
many challenges remain to nd inexpensive and effective
treatment processes for this water.
Printing ink wastewater is usually difcult to treat
biologically but the possibility of utilizing bacteria to
treat and degrade printing ink in wastewater was
proposed [11].
Since occulation was shown to be a simple and
efcient method for removing a substantial portion of
the organic content and to be very good for the removal
of colour and turbidity [12,13], the present study is
primarily concerned with additional TOC reduction
from the pretreated efuents by occulation. Although
zeolites are well known to be highly selective adsorbents,
in this research, the nonselective removal of total
organic content was investigated.
The adsorption efciencies of pretreated ltered
wastewater by various selected zeolites are compared
with that of powdered activated carbon (PAC), as a
standard adsorption medium. Equilibrium and kinetic
studies were also performed.
Since the turbidity of water treated by occulation is
strongly dependent on occulation concentration, adsorption of organics from water that was treated with
different occulant concentrations was studied. Furthermore, the non-ltered occulated water, after the
oc settled, was treated with zeolites in order to
determine the inuence of residual ocs on zeolite
capacity and to assess the possibility of avoiding
ltration after occulation.

T
TC
TOC
TOCe
VOC
x50
ZP
slog

temperature, K
total carbon content, mg C/L
total organic content, mg C/L
equilibrium total organic content, mgC/L
volatile organic compounds
particle size distribution median, mm
zeta potential, mV
dispersity parameter

2. Experimental methods and materials

2.1. Wastewater
Simulated wastewater was obtained by combining six
prepared single wastewaters of a specic printing ink,
each at a concentration of 5.0 g/L printing ink in tap
water. Each exographic, water-based, printing ink
represents a complex mixture, mainly based on methylmethacrylate copolymer as a binder (ca. 60%), also
including the corresponding pigment (ca. 10%), polyethylene wax (o4%), defoaming agent (0.5%) and
water. Pigments differ according to their chemical
composition. Yellow pigment has an anisidine structure
while red pigment is of the quinacridone type. Both red
and yellow pigments contain azo chromophores, while
blue and green pigments are of phthalocyanine group.
The other two pigments titanium dioxide and carbon
black, are inorganic. The commercial printing ink
samples were provided from Chromos-Printing Ink
Factory Samobor, Croatia.
The simulated water was characterized with respect to
turbidity, pH, TOC content (Table 1) and particle size.
2.2. Flocculation
On the basis of previous research with similar printing
inks [12], AlCl3
6H2O and FeCl3
6H2O were chosen as
the most efcient coagulants for this type of wastewater.
Laboratory jar tests were carried out in order to
determine the optimal concentration ranges to destabilize the printing ink suspension without adjustment
of pH. Inorganic salt was added to the dispersion
and mixed for 2 min under rapid mixing conditions
Table 1
Characteristics of simulated printing ink wastewater
pH

TC
(mg/L)

IC
(mg/L)

TOC
(mg/L)

NTU

ZP
(mV)

pH0=7.57.8
6.0
5.0

1094.0
629.7
147.5

41.1
29.7
13.1

1052.9
600.0
134.4

3980
1450
33.0

24.88
20.21
14.16

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

(250 rpm) after which slow mixing was applied

(50 rpm) for 10 min. The resulting ocs were allowed
to settle.
The electrokinetic (zeta) potential (ZP, mV), NTU
and TOC (mg/L) values were used for determination of
process efciency.
The electrokinetic data were calculated from the
measured electrophoretic mobilities (ue ; m2/s V) on
the basis of the Smoluchowski equation. The measurements were carried out with Otsuka ELS-800 electrophoretic light scattering instrument.
A HACH Model 2100P turbidimeter was used for
turbidity measurements. TOC was analysed by the
standard method with a Shimadzu Total Organic
Carbon Analyzer Model TOC-VCPN. The TOC values
were obtained by subtracting the IC concentration
from the TC concentration. The pH measurements
were made using pH/conductivity meter, Oakton model
WD-35630-60.
Based upon the results of the jar tests, occulation
was performed in the laboratory occulator (V =3 L)
with the addition of the optimal occulant dose
(g(FeCl3)=0.70 g/L; g(AlCl3)=0.56 g/L), with initial
rapid mixing (N 200 rpm, G 293:1 s1) for t
150 s, followed by slow mixing (N 50 rpm). In this
study rapid mixing, slow mixing, and sedimentation
were conducted in the same vessel. Mixing was
performed by an IKA Viscoklick VK 250 Control,
Eurostar Power control-viscosity agitator (Germany).
At the end of the settling period, the NTU of the
supernatant was measured, and the sizes of the ocs
remaining in the supernatant and those in the sludge
were determined by microscopic measurements. The
ratio of sludge volume to supernatant volume was
determined.
The batch experiments were repeated with lower
occulant concentration than optimal in order to
compare the sludge volume and to investigate the
inuence of residual ocs in the supernatant on the
subsequent adsorption step.
At the end of the batch experiments, the quality of the
treated water was assessed on the basis of measured
TOC and NTU, as well as on the volume of sludge.

3375

2.3. Adsorption
The water pretreated by occulation was ltered and
experiments were performed in order to determine the
best type of zeolite, the proper quantity of zeolite and
the sufcient time for mixing. A variety of synthetic
zeolites were evaluated according to their dimensional
structure, pore size, ratio of Al/Si and cations. The
zeolites used, obtained from Zeolyst International, (are
presented in Table 2. The adsorption efciencies of the
zeolites were compared with that of PAC (Kemika,
Croatia).
The adsorption of residual organic compounds on
zeolites and activated carbon, was performed at 298 K
by adding a specic amount of zeolite (120 g/L) into
pretreated water, mixing for a given period (t 5; 15,
30, 60, 120, 180, 240, 300 min), followed by sedimentation of the adsorbent. After additional ltration through
a Whatman qualitative cellulose lter paper (grade 5; 2.5
particle retention), TOC was measured in the ltrates, as
an indicator for adsorption efciency. NTU and pH
were also recorded.
An additional set of experiments was performed with
occulated pretreated water that was not ltered but was
decantated after the ocs settled. Samples of this
wastewater were either treated with optimal dosage of
FeCl3
6H2O (0.70 g/L) or AlCl3
6H2O (0.56 g/L) and
with concentrations less than optimum (0.50 and 0.40 g/
L, respectively). These adsorption experiments were
repeated by adding the optimal dosage of NH+
4 -ZSM-5
(5.0 g/L) into water. The ltrates were then measured for
TOC, NTU and pH.

3. Results and discussion

3.1. Characteristics of printing ink wastewater
The simulated wastewater contains high concentrations of organics (TOC=1053 mg/L) showing high
turbidity (NTU=3980) and pH around 7.5. It has
been previously reported that the stability of such
dispersions is highly dependent on pH, e.g. they show

Table 2
Characteristics of adsorbents used
Adsorbent

Producer and
number

Si/Al ratio

S (m2/g)

Adsorbent

Producer and
number

Si/Al ratio

S (m2/g)

H+-ZSM-5

CBV8020
CBV5020
CBV8014
CP814B-25
Kemika

40
25
80
12.5

430
420
425
730
1150

NH+
4 -mordenite
H+-Y
Na+-Y
13X

CBV30A
CBV720
CBV100
Aldrich

17.5
15
2.5
2.6

500
780

NH+
4 -ZSM-5
NH+
4 -beta
PAC

900

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

3376

high stability in basic pH while at low pH occulation

occurs.
The degree of organic pollutant removal, as measured
by TOC, signicantly decreases to 600 and 134 mg/L
when pH is decreased to 6.0 and 5.0, respectively
(Table 1). In both cases, the samples were ltered before
measuring TOC.
3.2. Flocculation
Common inorganic salts (aluminium and ferric
chlorides) were used as agents for maximal removal
(destabilization) of the particles, with size approximately
7 mm (x50 7:2 mm, slog 0:438), suspended in the
wastewater. According to ZP values between 5 and
0 mV (Fig. 1), the optimal occulant concentration

10.0

150
AlCl36H2O
FeCl36H2O
ZP(mV)
TOC(mg/L)

140

0.0

130

-5.0

120

-10.0

TOC (mgC/L)

ZP,mV

5.0

110
0.40

0.50

0.60

0.70

Concentration of coagulant, g/L

Fig. 1. The effect of occulant addition on zeta potential (ZP,

mV) and organic content (TOC, mg/L).

range between 0.450.60 and 0.530.72 g/L for

AlCl3
6H2O and FeCl3
6H2O, respectively, was determined for wastewater at initial pH. These results
correspond to the turbidity values shown in Table 3.
Larger concentrations of Al or Fe ions lead to larger
NTU values and ZP to the positive values, with
negligible change of TOC.
According to results from above jar tests, the
simulated wastewater was treated with optimal 0.70 g/
L FeCl3
6H2O or 0.56 g/L AlCl3
6H2O in the laboratory occulator under constant mixing conditions.
The mixing intensity is important because it has
been determined that for initial rapid mixing in the
range of 150200 rotations per minute, the sedimentation rate is the largest while the turbidity of the
supernatant water does not exceed NTU of 10. At
higher impeller speed ocs were smaller and remained
suspended in the supernatant, giving NTU between 10
and 30.
The characteristics of water treated with the optimal
occulant concentration are given in Table 3. These
results combined with those from Table 1 indicate that
the reduction of 88.56% and 88.92% of organic
pollution was obtained, with a ratio of sludge to water
of 1:5.1 and 1:4.1 for the case with Fe and Al salts,
respectively.
The relatively narrow optimal coagulant concentration range implies that even small changes in coagulant
concentrations result in signicantly different characteristics of treated water (TOC, NTU and pH) and
(slighter) different volumes of sludge. Table 3 also
shows the results of occulation when the coagulant
concentration is outside of the optimal range. For these
conditions it is still possible to obtain clearly separated
zones. At lower coagulant dose (0.5 g/L FeCl3
6H2O
and 0.35 g/L AlCl3
6H2O), the organic content is 15.5%
and 22.7% higher than when the optimal dose is applied.
In addition, the turbidity of ltrates (6.76 and 4.82) or of

Table 3
Characteristics of water after occulation at initial pH=7.5
Flocculant,
g (g/L)

NTU superna
tant

NTU
ltrate

PH

TC
(mg/L)

IC
(mg/L)

TOC
(mg/L)

TOC removal
(%)

Vsludge :V

FeCl3
6H2O
0.75
0.70
0.60
0.50

10
5.2
8.2
88

0.78
0
0.29
6.76

4.5
5.5
6.0
6.6

126.1
127.0
142.5
164.4

6.6
6.6
9.8
22.0

119.5
120.4
132.7
142.4

88.75
88.56
87.40
86.48

1:5.2
1:5.1
1:4.8
1:4.5

AlCl3
6H2O
0.60
0.56
0.50
0.45
0.35

5.4
5.5
5.5
5.5
73

0.27
0
0.02
0.39
4.82

6.4
6.5
6.5
7.4
7.5

116.1
120.0
127.8
137.1
152.9

3.5
3.3
2.6
2.6
2.8

112.6
116.7
125.2
134.5
150.1

89.31
88.92
88.11
87.22
85.74

1:4.2
1:4.0
1:4.0
1:3.5
1:2.9

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

supernatant water (88 and 73) at the lower dose is much

higher than that at the optimal dose.
A comparison of the quality of the water treated with
aluminium and ferric chlorides, shows that there is no
signicant difference between these two agents. The
percentages of organic removal are very close but the pH
of treated water is more acidic and the sludge volume is
slightly increased for the case with Fe salt.

3.3. Adsorption
Among several removal technologies, adsorption with
zeolites has great importance due to its simplicity and
comparable low cost of application. Because of the
complexity of wastewater studied here, non-selective
adsorption by zeolites is considered. In order to
investigate the inuence of residual ocs on zeolite
adsorption, it was rstly necessary to investigate the
effect of zeolite types, the optimal concentration of
zeolite, the appropriate contact times, and to determine
zeolite capacity. Therefore, zeolites with different pore
size and structure were applied in order to investigate
their sorption properties and effect on further pollutant
removal from ltered pretreated water (with optimal
concentration of 0.70 g/L FeCl3).
The distribution of organics between zeolites or
activated carbon and water, at equilibrium, is important
to obtain the capacity of adsorbents. Because of the
complexity of water studied in this investigation and of
the wide range of pore sizes and structure of zeolites

3377

considered, an analysis of the adsorption behaviour is

quite complex.
Generally, when discussing the adsorption process,
the verication of the adsorption mechanism can be
proposed on the basis of the correlation of the
experimental data with different, mostly empirical,
adsorption isotherms. The most commonly used isotherms for description of adsorption are the Langmuir
and Freundlich isotherms. The experimental data did
not t the Langmuir isotherm, giving negative slope and
intercept (meaning the equilibrium reaction constant
and the area per adsorbed molecule) which leads to the
conclusion that the adsorption behaviour of the tested
systems does not follow the assumption on which the
Langmuir approach was based. When the Freundlich
isotherm was used, a better t was obtained, although
not for all zeolites used. The Freundlich isotherm is used
in the form
q kce1=n ;
where q (mg/g) represents the mass concentration of
total organics adsorbed on the zeolites per unit of mass
of zeolites, while ce (mg/L) represents the equilibrium
concentration of total organics, expressed as TOCe
(mg/L).
The equilibrium concentration of organics (TOCe)
was determined by measuring TOC as a function of
time. Fig. 2 indicates the characteristic shape of the
kinetic curves that were obtained for ZSM-5 and PAC.
The other zeolites showed similar kinetics. Over the rst

2.9

2.6

2.3

TOC0/TOCt

1.7

1.4
H-ZSM-5, 10 g/L

1.1

H-ZSM-5, 5 g/L
NH4-ZSM-5, 5 g/L
PAC, 10 g/L

0.8

PAC, 5g/L
NH4-ZSM-5, 1 g/L
NH4-ZSM-5, 10 g/L

0.5
0

50

100

150

200

250

300

350

t, min

Fig. 2. Kinetic curves for ZSM-5 zeolites and PAC at different amount of adsorbents.

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

30 min there was a rapid increase in adsorption for all

cases, however following this initial period there were
minor variations with characteristically small oscillations of TOC. Such behaviour indicated that adsorption
mainly occurred on the external surface of the zeolites.
The TOC removal increases with increasing amounts
of zeolite, however larger amounts of zeolite necessitate
additional ltration to lower turbidity to required levels.
The addition of H+ or NH+
4 zeolites leads to either a
decrease in pH or an increase in pH, respectively.
Two different types of protonated ZSM-5 zeolites
with ratios of 25 and 40 Si/Al were studied in order to
determine the effect of the Si/Al on TOC removal and
pH change. The Si/Al reects the physical properties of
the zeolites including the afnity for polar adsorbates
and cation exchange properties.
H+-ZSM-5 was shown to be very effective for TOC
removal. There was no signicant effect of the Si/Al
ratio on the rates and quantity of contaminants
removed. The lower of Si/Al ratio increases H+
exchange resulting in a pH decrease (pH=34). Zeolite
(ZSM-5) concentrations of 5.0 and 10 g/L reduce TOC
to approximately 60 and 50 mg C/L, respectively.
When the same amounts of NH+
4 -ZSM-5 zeolite were
applied, approximately the same pollutant removal
(TOC=53.57 and 48.43 mg C/L) was obtained but the
exchange of NH+
4 ions caused the pH to change to
neutral (pH=77.6)
In order to investigate the role of adsorption in the
internal pores, a series of experiments was conducted
with larger pore size zeolites. Therefore, in addition to
ZSM-5, which is a pentasil zeolite characterized
by parallel straight channels of 5.35.6  1010 m
intersected by zig-zig channels of 5.15.5  1010 m,
several other zeolites were investigated [1416]. Mordenites contain unconnected parallel channels of
6.57.0  1010 m with smaller side pockets. Tetragonal
crystal structure with straight 12-membered ring channels (7.6  6.4  1010 m) with crossed 10-membered
ring channels (5.56.5  1010 m) is characteristic for
zeolite Beta. Zeolite Y and X exhibit the faujasite
structure and have a three-dimensional pore structure,
with pores running perpendicular to each other in the x;
y; and z planes. The pore diameter is large at
7.4  1010 m.
The roles of adsorption, the adsorption capacity, and
the shape of the isotherm for these larger pore zeolites
were similar to those for the smaller pore zeolites, thus
implying that the molecular size of the majority of TOC
is too large to penetrate the internal pores or pathways
of the zeolites. Furthermore, comparing NH+
4 -Mordenite with ZSM-5, it can be seen that TOC removal was
signicantly higher when ZSM-5 was applied although
it has smaller specic area and pore size. Zeolites
with higher adsorption capacity, S>700 m2/g, and
larger pore size than ZSM-5 and Mordenite, such as

NH+
4 -Beta, Y and 13X, behaved differently. While
X and Y showed poor adsorption for the given organic
pollutants, resulting in insignicant TOC removal,
NH+
4 -Beta was a very efcient adsorbent and behaved
in a similar manner to ZSM-5. Concentrations of NH+
4 Beta of 5 and 10 g/L reduce TOC to 57.3 and 39.3 mg/L,
respectively, giving neutral treated water (pH=7.1)
Since PAC is extensively used in water treatment, the
results obtained with ZSM-5 were compared with those
of PAC. The zeolites were found to have a higher
organic capacity in comparison to that obtained with
PAC.
For the same amount of PAC (5 g/L), and same
contact time, it was possible to reduce TOC only to
87.65 mg/L. To obtain the same TOC reduction as with
the zeolites, it was necessary to treat the water with 50 g/
L of PAC. In both cases pH of treated water was in the
range of 5.25.5.
Fig. 3 shows a comparison of the results obtained for
systems containing ZSM-5 and activated carbon. The
corresponding adsorption parameters together with
those obtained for all other experiments are summarized
in Table 4 except for 13X and Na-Y when adsorption
behaviour could not be described applying either the
Freundlich or Langmuir isotherms. The values of
the constants 1=n are relatively high in all cases due to
the large pore volume but no signicant difference in 1=n
values was observed, indicating the similar adsorption
behaviour of investigated adsorbents.
Fig. 4 shows the percentage of additional TOC
removal that occurs with different amounts of zeolites
in wastewater occulated with an optimal amount of
ferric chloride. Although the TOC removal increases
with increasing amount of zeolite, because of the
necessity of additional ltration to lower turbidity to
1.8
1.6
1.4
1.2
logq

3378

1
0.8
0.6

PAC
NH4-ZSM-5

0.4

H-ZSM-5, Si/Al=25

0.2

H-ZSM-5, Si/Al=40

0
1.6

1.7

1.8
logTOCe

1.9

Fig. 3. Adsorption isotherms of organic pollutants on ZSM-5

zeolites and PAC.

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381

3379

3.4. Influences of flocs on zeolite capacity

required levels, the amount of 5 g/L has been found as

the optimum.
From the results obtained, ZSM-5 and Beta zeolites in
the NH+
4 form have been selected as most effective
zeolites for satisfactory TOC removal. The treated water
had pH around 7.
The combined process, whereby occulation was
followed by adsorption with the zeolites, gave much
higher efciency. The initial wastewater TOC (1052.9 mg
C/L) could be decreased to 120.4 mg C/L by a single
occulation process while by the two-step process with
either ZSM-5 or NH+
4 -Beta, TOC could be further
decreased to around 50 mg/L.
In the case of NH+
4 -ZSM-5, the overall efciency
increased to around 95% when its optimal amount was
used (Fig. 5).

After occulation at the optimal occulant concentration (treatment 1), and after the sludge zone was
formed, the supernatant water had turbidity between 5
and 10, indicating that small ocs remained in the entire
settling zone. However, the turbidity was completely
removed when the supernatant water was ltered. The
microscope measurements showed that ocs remaining
in the supernatant had irregular shape and a wide range

2.20

2.00

TOCo/TOC

1.80

Table 4
Freundlich parameters in the equilibrium isotherms for the
organicszeolites and organicsactivated carbon systems
Freundlich adsorption isotherm
constants
k (mg/g)/(mg/L)
H-ZSM-5 (Si/Al=25)
H-ZSM-5 (Si/Al=40)
NH4-ZSM-5
PAC
NH4-beta
H-Y
NH4-mordenite

7

8.0  10
7.1  103
2.7  108
2.1  109
6.8  104
1.4  101
6.5  108

1=n

R2

4.16
1.79
5.32
4.93
2.49
5.26
4.34

0.9313
0.9168
0.9900
0.9817
0.9704
0.9908
0.7553

1.60

1.40
supernatant in treatment 1
filtrate in treatment 2
supernatant in treatment 2

1.20

1.00
0

30

60

90

120 150 180 210 240 270 300

t, min

Fig. 5. Kinetic curves for NH+

4 -ZSM-5 (5 g/L) applied in
differently pretreated water.

80

70

60

59.75
53.39
52.25

TOC removal, %

10 g/L

67.36

5 g/L

61.68

59.27

2.5 g/L

57.65

54.67

51.6

52.4
48.14

50

48.25

46.39

40.85

40

36.32
33.13

35.87

29.29

27.19

30

25.45

18.28

20

14.94

13.57

10

12.12

3.82
0.75

0
H-ZSM-5
(Si/Al=25)

H-ZSM-5
(Si/Al=40)

NH4-ZS M-5

PAC

NH4 -Beta

NH4 -Mordenite

Na-Y

H-Y

Adsorbent

Fig. 4. Additional TOC removal, from occulated water, in dependence of adsorbent type and its amount.

13X

ARTICLE IN PRESS
3380

A. Metes et al. / Water Research 38 (2004) 33733381

of sizes but they were not larger than 35 mm. On the

contrary, the ocs in sludge, obtained at applied mixing
conditions, are crosslinked and therefore their size could
not be dened.
Mixing intensity affects the ocs size and shape as well
as the settling velocity and, consequently, the turbidity
of the supernatant. At higher impeller speed
(n > 300 rpm), in supernatants there were ocs not larger
than 20 mm while in the sludges there were more
individual ocs with length varying between 0.2 and
1.5 mm and with widths 23 times smaller.
Decantated supernatant was treated with an optimal
amount of NH+
4 -ZSM-5. The kinetic curve had the same
shape, including characteristic time oscillations of TOC,
as the ltrated water. Equilibrium was established within
the time interval of 23 h. The presence of the ocs did
not affect the adsorption rate or the zeolite capacity. The
TOC removal (57%) was also unaffected by the ocs.
When the same experiments were repeated with
occulant concentration smaller (0.5 g/L FeCl3
6H2O)
than optimal (treatment 2), the supernatant water showed
much higher turbidity (NTU=88) than the ltered water
(6.76), while the resulting TOC was 142.4 mg/L.
Comparing the adsorption results for ltrated water
samples from treatment 1 and 2, the effect of higher
turbidity on TOC removal was observed particularly in
the rst 60 min of contact with adsorbent. Within the rst
15 min, the difference between TOC values was around
10% while the difference decreased to 4% near 1 hour of
mixing. After 1 h of adsorption, the TOC values were
approximately the same, with the same tendency of
oscillation around 60 mg C/L in the studied time interval.
The series of experiments with supernatant water
showed the inuence of ocs on the adsorption process.
Comparing the results for ltered and supernatant water
in treatment 2, it was seen that the ocs had a signicant
inuence on TOC removal in the rst 60 min. Much
greater variations of TOC removal for appropriate time
were observed indicating that ocs affected process
efciency due to covering the zeolite surfaces, thus
showing that adsorption of organics was mainly
occurring on the external surfaces of the zeolite.
The adsorption rate was lower and the results did not
t the Freundlich isotherm. Decreased TOC removal of
more than 15% in the rst 30 min occurred.
However, the nal TOC values of treated water were
between 61 and 64 mg/L. Removal efciency decreased
to around 10% for the same amount of zeolite and in the
same time interval but the overall removal efciency was
still above 94%.

4. Conclusions
Although a large percentage of organic compound
removal (88%) was achieved by a single occulation

step, the combined occulation/zeolite adsorption process for printing ink removal from simulated wastewater
showed some advantages for potential application such
as higher efciency (95% of organic removal) and water
neutralization.
ZSM-5 and Beta zeolites, both in NH4-form, were
found to be the most efcient in enhancing TOC
removal. At concentrations of 5 g/L, they decreased
TOC more than 50% from pretreated water by
occulation. In comparison with PAC of the same
concentration, the zeolites are over 20% more
effective.
It was shown that ltration is not required before the
adsorption process even when a coagulant concentration
lower than optimal is applied because ocs do not
signicantly inuence zeolite capacity. Furthermore, the
volume of sludge in that case is smaller than when the
optimal coagulant concentration is applied. Therefore,
while optimizing the process, the consumption of
coagulant should be balanced with the generation of
sludge.
It would be of practical signicance to operate such a
combined process because it allows use of a wider
concentration range of coagulant, which will not result
in signicantly different characteristics of nal water
(TOC, pH, NTU).

Acknowledgements
The authors thank D. Holjevac, from Chromos
Printing Ink Factory, Samobor, Croatia, for providing
printing ink samples and Professor Keith Smith,
University of Wales Swansea, UK, for providing zeolites
samples.

References
[1] Tomlinson AAG. Modern zeolites, Structure and function
in detergents and petrochemicals, vol. 3. Switzerland:
Trans Tech. Publications, Ltd.; 1998.
[2] Ming DW, Allen ER. Use of natural zeolites in agronomy,
horticulture, and environmental soil remediation (Review).
In: Bish DL, Ming DW, editors, Natural zeolites:
occurrence, properties, applications, 45PG. Washington,
USA: Mineralogical Society of America; 2001. p. 619654.
[3] Leinonen H, Lehto J. Purication of metal nishing waste
waters with zeolites and activated carbons. Waste Manage
Res 2001;19(1):4557.
[4] Chung YC, Son DH, Ahn DH. Nitrogen and organics
removal from industrial watewater using natural zeolite
media. Water Sci Technol 2000;42(56):12734.
[5] Hedstrom A. Ion exchange of ammonium in zeolites: a
literature review. J Environ EngASCE 2001;127(8):
67381.

ARTICLE IN PRESS
A. Metes et al. / Water Research 38 (2004) 33733381
[6] Singh G, Prasad B. The removal of ammonia from cokeplant wastewaters by using synthetic zeolite. Water
Environ Res 1997;69(2):15761.
[7] Vlessidis AG, Triantallidis CS, Evmiridis NP. Removal
and recovery of p-phenylendiamines developing compounds
from photonishing and lab-washwater using clinoptilolite
tuffs from Greece. Water Res 2001;35(6):16038.
[8] Richards MD, Pope CG. Adsorption of methylene blue
from aqueous solutions by amorphous aluminosilicate gels
and zeolite. J Chem Soc Faraday Trans 1996;92:31723.
[9] Huddersman K, Patruno V, Blacke GJ, Dahm RH. Azo
dyes encapsulated within aluminosilicate microporus
materials. J Soc Dyers Colourists 1998;114:1559.
[10] Karcher S, Kornmuller A, Jekel M. Screening of commercial sorbens for the removal of reactive dyes. Dyes
Pigments 2001;51(23):11125.

3381

[11] Zhang Y, Shi H, Qian Y. Biological treatment of

printing ink wastewater. Water Sci Technol 2003;47(1):
716.
[12] Metes A, Koprivanac N, Glasnovic A. Flocculation as a
treatment method for printing ink wastewater. Water
Environ Res 2000;72(6):6808.
[13] Nandy T, Vyas RD, Shastry S, Kaul SN. Optimization of
coagulants for pretreatment of printing ink wastewater.
Environ Eng Sci 2002;19(1):17.
[14] IZAstructure Commission website (2003), http://www.
iza-structure.org/databases
[15] Zeolyst international website (2003), http://www.zeolyst.
com
[16] Meirer WM, Olson DH. Atlas of zeolite structure types,
2nd ed. London: Butterworths; 1987.