Removal of Natural Organic Matter from Water Using Ion-Exchange
Chromatography and Cyclodextrin Polyurethanes
T.I Nkambule1, R.W Krause1*, B.B Mamba1, and J Haarhoff2
1
University of Johannesburg, Nanotechnology Innovation Centre, Department of Chemical Technology, P.O. Box
17011 Doornfontein, Johannesburg, 2028, South Africa
*Email: rkrause@uj.ac.za, Tel no: +2711 559 6152, Fax no: +2711 559 6425
2
University of Johannesburg, Department of Civil Engineering Science, P.O. Box 524 Auckland Park, Johannesburg,
2006, South Africa
Abstract
Natural organic matter (NOM) consists of a complex mixture of naturally occurring organic compounds. NOM present
during water disinfection may result in the formation of disinfection by-products (DBPs), many of which are
carcinogenic or mutagenic. Although it is difficult to completely characterize NOM due to its complex and large
structure, this understanding is necessary for determining the mechanism of NOM removal from water. In this study,
water from the Vaalkop water treatment plant was characterized for its NOM composition by fractionation on ionic
resins. Fractionation at different pH and with different ion exchange resins resulted in the isolation of a neutral, basic
and acidic fraction of both the hydrophobic and hydrophilic NOM components. The hydrophilic basic fraction was
found to be the most abundant fraction in the source water. When each isolated NOM fraction was percolated through
cyclodextrin (CD) polyurethanes, an adsorption efficiency of 6%-33% was obtained. The acidic fractions were the most
adsorbed fractions by the CD polyurethanes, with the neutral fractions being the least adsorbed. Trihalomethanes
(THMs) were formed by all six NOM fractions, but in varying proportions. The hydrophilic basic fraction was found to
be the most reactive precursor for THM formation. Treatment of the water sample with the CD polyurethanes after
ozonation resulted in a removal efficiency of up to 59%, which was mainly the hydrophilic basic fraction. The effect of
the combination treatment in removing NOM resulted in the polymer exhibiting a relatively good capability to remove
NOM from water as evidenced by an up to 88% reduction of the hydrophilic acidic fraction, and similar result for the
other NOM components.
Keywords: Cyclodextrin polyurethanes; Disinfection by-products; Natural organic matter; Trihalomethanes; Ozonation
Introduction
Natural organic matter (NOM) is a conglomeration of organic compounds, having diverse chemical
properties and it occurs in all natural water sources when animal and plant material breaks down
(Xie, 2003). Due to the differing sources of NOM, it is expected that the composition of NOM in
different water bodies may also not be uniform.
During the production of potable water, NOM may react with chlorine or other disinfectants to
produce disinfection by-products (DBPs) many of which are carcinogenic or mutagenic (Xie, 2003).
For example, haloacetic acids (HAAs), a component of DBPs, are considered harmful to human
health and also have diverse toxicological effects on laboratory animals, such as impaired
reproductive and developmental effects (Kanokkantapong et al. 2006). Trihalomethanes (THMs)
are another example of DBPs and have been classified as possible carcinogens to humans. NOM
itself also causes membrane fouling, aesthetic and malodour problems in water, the organic acids
which emanate from NOM oxidation have been implicated in the corrosion of turbines and
engineering systems (McDonald et al. 2004). Thus, understanding the impact of NOM in water
treatment process is crucial not only to human health, but also in industrial processes where pure
water is a requirement.
NOM can be broadly categorized into humic substances, microbial byproducts and colloidal natural
organic matter. Humic substances constitute the more hydrophobic fraction of NOM and exhibit
relatively high specific ultra-violet absorbance (SUVA) values as they usually contain a relatively
large proportion of aromatic moieties (Chen et al. 2002). Microbial byproducts, the second category
of NOM, consist of acids, with relatively high charge density, as well as polysaccharides, amino
sugars and proteins (Vanboon et al. 2005), making this group hydrophilic. Finally, colloidal natural
organic matter includes relatively polar amino sugars and may cause high membrane fouling
potential due to its neutrality.
Granular activated carbon (GAC) has been widely used to remove organic matter from water, but
often fails to remove it to sufficiently low levels so as to be inconsequential to treatment process
(McDonald et al. 2004). Ozonation, though used for disinfection, leads to the oxidation of NOM
and formation of low molecular weight organic compounds, bromated organics and bromates.
However, these fragments, when not properly adsorbed on GAC tend to be more difficult to remove
due to their mobility and general increased polarity (Wuilloud et al. 2003). It is anticipated that
these ozonation fragments will be removed more efficiently by the use of water-insoluble
cyclodextrin (CD) polyurethanes, whose synthesis and application have been previously reported
(Mhlanga et al. 2007). These polymers demonstrated huge capacity to remove a wide range of
organic contaminants present in water at concentration levels of ng/L (Mhlanga et al. 2007).
Furthermore, NOM once adsorbed onto the active sites of GAC, renders the GAC ineffective at
removing other micro-organic pollutants that may be present in the water but the CD polyurethanes
appear to retain their ability to remove small organics in the presence of humic acid (Mamba et al.
2008).
Cyclodextrins (CDs) are cyclic glucose oligomers made from enzymatic degradation of starch
through the action of Bacillus macerans. These glucose units are joined via α-(1,4) glycosidic
linkages (Bender and Komiyama, 1978). The CDs, once polymerized into water insoluble polymers
are capable of forming inclusion complexes with guest molecules and hence can be utilized for the
removal of organic contaminants from water.
In order to better understand the mechanism of NOM removal from water, it is important to
understand the composition of the NOM in the source water, especially as applied to the local NOM
conditions. This comprehension would then propel the need to characterize the water source for its
NOM composition before attempting to remove it.
NOM, which is usually found in drinking water at concentration levels between 2 and 15 mg/L
(Hepplewhite et al. 2004), can be fractionated i.e. isolated and separated, into three hydrophobic
and three hydrophilic fractions using suitable ion exchange resins. The hydrophobic fractions
contain mostly humic and fulvic acids whereas the hydrophilic fractions comprise low molecular
weight carbohydrates, proteins and amino acids (Karnik et al. 2005). The fractionation technique
allows for the independent evaluation of each organic fraction present in the water sample and how
effectively each fraction can be removed from water using adsorbents such as GAC and CD
polyurethanes. The fractions can also be individually tested against DBP formation to ascertain
which group of its constituent molecules could be targeted for removal.
Herein, we report on the characterization of NOM present in water samples from aforementioned
water treatment plant using ion exchange resins for fractionation in combination with the use of
water insoluble cyclodextrin polyurethanes in the removal of these NOM fractions from water. We
also report on the individual isolated NOM fraction’s potential to form THMs. Furthermore, the
effect of ozonation and subsequent compositions of these fractions followed by their removal by the
CD polyurethanes will be evaluated.
Experimental
Sample collection and preservation
Water samples were collected from the Vaalkop water treatment plant (situated in North-West of
Johannesburg), which uses ozonation as part of its standard water treatment protocol. Samples were
taken before and after the ozonation process and the sampling was repeated three times at intervals
of one month. All water samples were filtered through a 0.45 μm cellulose filter paper and stored in
the refrigerator at 4˚C for not more than 48 hrs. Two litres of each water sample was then
fractionated to observe the variation of the organic species.
Conditioning of resin columns
Three types of ion exchange resins, namely XAD-7HP, Dowex ® 88 and Diaion-WA-10, were
packed into three separate glass columns (inner diameter: 2.5 cm) up to a resin height of 20 cm.
Each resin was cleaned thoroughly before the isolation of the different NOM fractions to
significantly reduce the level of dissolved organic carbon (DOC) bleeding from the resins.
The XAD-7HP resin was sequentially extracted for 24 hrs with acetone and hexane using Soxhlet
technique to clean and prepare the resin prior to use. Methanol was pumped through the resin until
the effluent was free of hexane. Following methanol, distilled water was pumped through the resin
until the DOC of the effluent was less than 1 mg/L. Finally, the resin was rinsed with 0.1 M NaOH
and 0.1 M HCl to remove any remaining impurities.
The Dowex ® 88 resin was subjected to Soxhlet extraction with methanol for 24 hrs, after which
600 ml of 3 M NH4OH was pumped through the resin. The resin was then protonated to saturation
with 300 ml of 2 M HCl. Distilled water was finally passed through the resin to remove any
remaining impurities.
The Diaion-WA-10 resin underwent Soxhlet extraction with acetone for 24 hrs. Following the
extraction, 1 M HCl was pumped through the resin until the DOC of the effluent was reduced to
less than 1 mg/L. An excess of 3 M NH4OH was pumped through the resin until the resin changed
from its off-white colour to a yellowish colour (indicating that resin was in its free base form).
Finally, distilled water was passed through the resin until the DOC of the effluent was less than 1
mg/L.
Fractionation procedure
Fractionation of the NOM was performed using three types of ion exchange resins namely; XAD7HP, Diaion-WA-10 and Dowex ® 88, as described in the procedure by Marhaba et al. (2003).
According to this technique, the organic matter could be divided into six fractions namely;
hydrophobic acid (HpoA), hydrophobic base (HpoB), hydrophobic neutral (HpoN), hydrophilic
acid (HpiA), hydrophilic base (HpiB) and hydrophilic neutral (HpiN). The fractionation procedure
is shown in Figure 1.
Hydropobic neutral:
Eluted by methanol
XAD-7HP
pH 7
pH 2
Hydrophobic acid:
Eluted by NaOH
XAD-7HP
pH 10
Hydrophobic base:
Eluted by Hcl
XAD-7HP
Hydrophilic base:
Eluted by NaOH
Dowex-88
Hydrophilic acid:
Eluted by NaOH
DiaionWA10
Hydrophilic neutral:
Obtained by freeze concentration
Figure 1: Fractionation method
Sample
Organic Carbon Analysis
Dissolved organic carbon (DOC) was used as an indicator of the organic content in the water
samples. DOC is the organic constituent that can pass through the 0.45μm filter paper. The DOC of
each NOM fraction was measured using a total organic carbon (TOC) analyzer (Tekmar Dormann
Apollo 9000). Standards of 1, 2, 5, and 10 mg/L C were prepared with potassium hydrogen
phthalate (KHP), and ultra pure water was used in all dilutions. A minimum of three replicates of
each measurement were carried out and values averaged, discarding any significant outliers.
Ultra Violet-Visible (UV-Vis) Spectrophotometric analysis
The water samples were also analyzed on a UV-Vis spectrophotometer at a wavelength of 254 nm.
Absorption at this wavelength has been reported to represent the “aromatic character” of the organic
species. The Varian Cary-50 UV spectrophotometer with a 1-cm quartz cell was utilized for all UVVis Spectrophotometric measurements.
Absorption of each NOM fraction using the CD-polyurethane
The CD polyurethane polymer was synthesized as per procedure described by Li and Ma
(2000).The water insoluble polymer synthesized was obtained through the polymerization of β
cyclodextrin with an excess of hexamethylene diisocynate (HMDI). (This polymer will be
abbreviated β-CD-HMDI). The polymer was washed first to remove un-reacted cyclodextrin and
other pollutants that may interfere with TOC measurements. This was done by first heating the
polymer for an hour at 156˚C to displace any trapped moisture and solvents. It was then allowed to
cool to room temperature. The polymer (300 mg) was then loaded into empty SPE cartridges and
washed with de-ionized water until TOC readings, taken at intervals of 30 minutes, were less than
0.5 mg/L. The isolated NOM fractions were then treated with the polymer by passing 30 cm 3 of
each organic fraction through the polymer at a filtration rate of 10 cm3/min. The polymer-treated
water was then analyzed for its carbon content using the TOC analyzer to determine how much of
the percent carbon had been adsorbed onto the polymer.
Chlorination of NOM fractions
For the determination of the THM formation potential, a chlorine solution of 18 mg/L was prepared
from granular chlorine. The samples were first adjusted to a pH of 9±0.2 by means of a borate
buffer. After which, chlorine water was added to the samples in accordance with the samples
chlorine demand, as determined by the TOC and amount of ammonium salts present in the samples.
Finally the samples were incubated at 25˚C in amber bottles for seven days to allow for the process
of THM formation. All sample bottles were head-space free to prevent any air bubbles trapped with
the incubated NOM fractions. After seven days of chlorination, 30% ascorbic acid was added to the
samples to react with any remaining free active chlorine.
Gas Chromatography/ Mass Spectrometry analysis
The Gas Chromatography-Mass Spectrometry (GC-MS) analyses were carried out using a Varian
CP-3800 capillary Gas Chromatograph coupled with a Saturn 2000 Mass Spectrometer. THMs were
extracted using a 100 μm polydimethylisoxane (PDMS) solid phase micro extraction (SPME) fibre.
To ensure that the fibre was clean before performing an analysis it was first conditioned. The
conditioning was performed by exposing the fibre to injector analysis temperature of 200˚C for 20
minutes. This was followed by a blank analysis carried out by desorbing the fibre for a further three
minutes at similar analysis temperature. Standards of 10, 50 and 100 μg/L were prepared from a
2000 μg/mL THM calibration mix from Supelco. The THMs that were analyzed were chloroform,
bromodichloromethane, dibromochloromethane and bromoform. The fibre was immersed in 10 mL
aliquots of each standard solution for 20 minutes at 70˚C. The solution was constantly stirred
throughout the extraction time to enhance extraction. After the extraction, the fibre was exposed
into the GC-MS injector port for three minutes. THMs in the samples were extracted the same way
as for the standards. GC-MS conditions used for analysis are shown in Table 1.
Table 1: GC-MS conditions
Parameter
Condition
Column type
VF,5ms, 30 x 0.25 mm, 0.25 µm
Injector
Splitless, 3min sample exposure
Injector temperature
200˚C
Oven temperature
35˚C (hold for 10 min) @ 9˚C /min to 120˚C @15˚C /min (hold for 5 min)
Ionization
Electron impact
Carrier gas
Helium
Flow rate
1 mL/min
Detector
Mass spectrometer (ion trap)
Mode
Full Scan (m/z = 50-255)
Results and Discussion
The Vaalkop water treatment plant
The Vaalkop water treatment works consists of three water treatment plants which use different
water treatment protocols. Samples for this study were sourced from the plant (Plant 1) which
utilizes both ozonation and chlorination as part of the standard water treatment procedure. This
treatment plant is located in the North-West province and sources its water from the Vaalkop dam, a
large open-water source. Figure 2 is a flow chart of the water treatment processes at plant one,
where A and B depict the sampling points.
Raw water
Granular activated
carbon
Flocculation
Coagulation
A
B
Chlorination
Sedimentation
Ozonation
Sand filtration
Figure 2: A flow chart showing the water treatment process at plant 1, together with the two sampling points A and B.
Characterization of samples before fractionation
The samples exhibit a relatively high TOC content as shown by the TOC values in Table 2.The
Specific Ultra Violet Absorbance (SUVA) was used as an indicator for the aromaticity of the NOM
in the water samples. SUVA gives an indication of the amount of humic substances relative to nonhumic substances in the water samples (Kiwa, 2006: Chen et al. 2002). Since the SUVA for all
samples (Table 2) was low, this suggested that the NOM in the samples had a lower aromatic
content in nature.
Table 2: Characterization of samples obtained after 1st, 2nd and 3rd sampling
TOC
SUVA
Sample 1
pH
(March 2008)
(mg/L)
[L/(mg.M)]
Before
12.0
1.1
7.0
ozonation
After
10.5
0.9
6.6
ozonation
Sample 2
Conductivity
(μS)
Turbidity
(NTU)
602
1.3
611
0.3
(April 2008)
Before
ozonation
After
ozonation
Sample 3
14.6
1.5
7.6
599
0.4
12.8
0.7
7.0
620
0.4
6.6
3.4
9.0
636
1.3
6.0
2.7
9.1
626
1.0
(May 2008)
Before
ozonation
After
ozonation
The TOC of both sample 1 and 2 decreased by approximately 12% after ozonation (from 12.0 to
10.5 mg/L in sample 1 and from 14.6 to 12.8 mg/L in sample 2), while the TOC of sample 3
decreased by approximately 9% after ozonation (from 6.6 mg/L to 6.0 mg/L). The ozone dosage at
this particular plant was set as 3.5 mg/L. The slight decrease in TOC was expected, since ozonation
as reported by Bose et al. (2007) tends to convert biodegradable organic matter to biodegradable but
smaller organic molecules. In other words higher molecular weight compounds are converted to
lower molecular weight compounds.
The effect of ozonation on the NOM was also confirmed by the slight decrease in pH (see Table 2).
There is a slight decrease in the pH values for all samples after ozonation, confirming the oxidation
effect of ozone on NOM and the generation of lower molecular weight substances which were most
probably acidic in nature. Zhang et al. (2008) also reported that oxidation of NOM results to an
increase of the more acidic functional groups hence a slight decrease in pH values after ozonation.
Figure 3 shows an infra-red (IR) spectrum displaying the different functional groups that were
present in the water samples for samples 1, 2 and 3
100
Sample 1 before o3
Sample 2 before o3
Sample 3 before o3
Transmission (% T)
80
60
40
C=N
C-N
20
C=O
O-H
0
4000
3500
3000
2500
2000
1500
1000
500
-1
Wavenumber (cm )
Figure 3: An IR spectrum displaying the functional groups mainly associated with the NOM in raw water
Fractionation of organic matter
The mass balance of the different organic fractions is summarized in Table 3. The negative
percentage surplus (i.e. mass deficit) meant that some of the organic fractions are irreversibly
trapped in the resins and therefore lost during elution. Notably the losses were greater after
ozonation - an effect which is still being investigated, but we believe the increased polarity of the
ozonated fractions would cause their stronger binding to the ionic resins. In fact, Kiwa (2006)
reports that all resins remove NOM fractions but with significant differences.
Table 3: Mass balance of organic species for samples 1, sample 2 and sample 3.
TOC
HpoB
HpoA
HpoN
HpiB
Sample 1
(March 2008)
(mg/L)
(mg/L)
(mg/L)
(mg/L)
(mg/L)
Before
12.0
1.5
2.5
1.7
3.1
ozonation
After
10.5
0.8
2.6
2.2
1.4
Ozonation
HpiA
(mg/L)
HpiN
(mg/L)
%
Surplus
1.0
1.5
-5.8
0.9
1.5
-10.5
Sample 2
(April 2008)
Before
ozonation
After
ozonation
Sample 3
14.6
1.4
3.0
2.2
4.1
0.7
1.7
-10.3
12.8
1.1
3.5
2.5
1.8
0.3
1.6
-16.4
6.6
0.8
1.8
0.7
1.5
0.9
0.4
-7.6
6.0
0.5
2.1
0.8
0.5
0.8
0.5
-8.3
( May 2008)
Before
ozonation
After
ozonation
As can be seen from Table 3, the water samples were mainly composed of the hydrophobic NOM
fractions (51% for sample 1 and 2, and 54% for sample 3) thus implying that the NOM consisted
mainly of humic substances. This result supports the general assertion that hydrophobic substances
constitute a larger fraction than the hydrophilic fraction in natural waters (Yavich et al. 2004).
Ozonation led to a decrease in TOC, especially in the hydrophilic basic fractions. The hydrophilic
basic fractions (HpiB) TOC was reduced by 55%, 56% and 65% for samples 1, 2 and 3 respectively
by ozonation, resulting in an overall relative increase in the hydrophobic fractions to 60%, 66% and
65% for samples 1, 2 and 3 respectively. The high reduction of this fraction was due to the fact that
the hydrophilic base fraction is not aromatic and the ozone reacts readily with these functional
groups, hence the significant decrease in TOC. The hydrophobic basic (HpoB) TOC was also
slightly reduced by ozonation but not as significantly as the HpiB’s. This was due to the aromatic
character of the hydrophobic fraction. The neutral fraction’s TOC (both hydrophobic and
hydrophilic) was generally not affected by ozonation. However, the TOC of the hydrophobic acid
fraction for all samples increased after ozonation.
Treatment of the organic fractions with β-CD HMDI polymer
The isolated NOM fractions were then passed through water-insoluble β-CD HMDI polyurethane to
determine the extent to which each organic fraction can be removed by the polymer. Table 4 shows
the percentage removal of each NOM fraction by the polymer. On the whole, the polymer exhibited
low capability at removing the NOM fractions as shown by the values in Table 4. The acidic
fractions (HpoA and HpiA) were the best adsorbed by the polymer at 27% and 33% removal
efficiency for HpoA and HpiA respectively. The acidic fractions are generally more ionic and have
a high charge density, thus their greater affinity both for the cavity of the CD polyurethanes as well
as the polar surface. The neutral fractions were the least adsorbed fractions by the polymer, with a
minimum removal of 9% and 0% for HpoN and HpiN respectively (Table 4). The low adsorbance
of the neutral fractions was due to the fact that these fractions are less ionic with a low charge
density; hence they cannot be easily adsorbed by the CD polyurethane. While this poor removal of
the neutral fractions was expected, this understanding is necessary as it helps in the understanding
of how to best modify the polymers to better remove the NOM fractions.
Table 4: The percentage absorbance of each NOM fraction by the CD polymer
TOC before polymer
TOC after polymer
NOM fraction
(mg/L)
(mg/L)
0.8-1.4
0.7-1.2
HpoB
1.8-3.0
1.5-2.2
HpoA
0.7-2.2
0.6-2.0
HpoN
1.5-4.1
1.3-3.5
HpiB
0.7-0.9
0.5-0.6
HpiA
0.4-1.7
0.4-1.6
HpiN
% Absorbed
13-14
17-27
9-14
13-15
29-33
0-6
If the combined effect of ozonation and subsequent treatment of the ozonated water samples with
the CD polyurethanes are evaluated, the results obtained suggest that this strategy meets our
intended objective. Results from Table 5 indicate an up to 88% removal of the hydrophilic acid
(HpiA) fraction from the water source, which was not attained by either process independently. For
example, the removal rate of the hydrophilic basic fraction was 73%, showing a great achievement
of this combination approach as this study reports this fraction to have the highest THM formation
potential.
Table 5: Demonstrates the success of the combination of ozone treatment and CD polymers in removing the NOM
fractions from water.
Process
HpoB
HpoA
HpoN
HpiB
HpiA
HpiN
-35%
+13%
+19%
-59%
-57%
-10%
Ozonation
-14%
-22%
-12%
-14%
-31%
-3%
CD polymer
Overall
-49%
-9%
+7%
-73%
-88%
-13%
reduction
Trihalomethanes (THMs) analysis of the NOM fractions
The isolated NOM fractions were then individually tested against their DBP formation potential.
THMs were used as a representative of the disinfection by products in the NOM fractions. The
THMs that were analyzed are; chloroform, bromodichloromethane, dibromochloromethane and
bromoform. Standards of 10, 50 and 100 μg/L of the THMs were prepared from a 2000 μg/L THM
calibration mix from Supelco. THMs concentrations in the NOM fractions were then determined
from the peak areas of the samples and standards as given by equation 1.
THM conc. (μg/L) = Peak area (sample) x peak area (CH2Cl2/std)
x THM conc. (std)
Peak area (std)
peak area (CH2Cl2/sample)
(1)
Figure 4: A GC/MS chromatogram of the 100 μg/L THM standard.
Figure 4 shows a GC/MS chromatogram of a 100 μg/L standard which was used for the calculation
of the concentrations of the determined THMs.
Only chloroform and bromodichloromethane were detected in the NOM fractions.
Dibromochloromethane and bromoform were not detected by the GC/MS in the NOM fractions.
Table 6 shows the THMs concentrations in the different NOM fractions.
Table 6: NOM fractions THMs concentrations (in μg/L)
Sample
Chloroform
Bromodichloromethane
HpoA
0.032
0.015
HpoB
0.017
0.008
HpoN
0.023
0.011
HpiA
0.024
0.012
HpiB
0.034
0.012
HpiN
0.022
0.008
The THM formation potential experiment of each fraction provides insight into the chlorinated
reactivity of each contributing precursor (Marhaba et al. 2000). THMs were formed by all six NOM
fractions, but with varying and different proportions. The hydrophilic base fraction was found to
have the highest concentration of chloroform and the hydrophilic acid fraction had the highest
concentration of bromodichloromethane. In total the hydrophilic base fraction was the most reactive
precursor fraction for the total formation of THMs. This finding differs slightly from that reported
by Marhaba et al. 2000 where the hydrophilic acid fraction was found to be the most reactive
precursor in THMs formation. This could be expected since the water sources for both studies were
different, and hence so too was their NOM composition. Marhaba et al. attributed the hydrophilic
acid fractions potential to form most THMs to the fact that since the chlorine species are
electrophiles, they tend to react with electron-rich sites in organic structures.
The hydrophobic acid fraction had the least concentration of chloroform and the hydrophobic
neutral fraction had the least concentration of bromodichloromethane. In total the hydrophilic
neutral fraction had the lowest concentration of total THMs (chloroform and
bromodichloromethane).
Conclusion
Characterization results (SUVA and mass balance of fractions) indicated that the water samples
mainly consist of humic substances in the form of hydrophobic NOM. Ozonation, at dosage of 3.5
mg/L, was less effective at removing NOM since it only resulted in a 12% overall decrease of the
TOC of the water samples. However, ozonation reduced the TOC of the hydrophilic basic fractions
by over 50% and that of the hydrophobic basic by approximately 35%. THMs were formed by all
six NOM fractions, but with varying proportions. The hydrophilic base fraction was found to
contain the highest concentration of THMs. The combination (i.e. ozonation and treatment with CD
polyurethanes) method employed by this study results to an overall reduction of the hydrophilic
base fraction by 73%. The acidic fractions (HpoA and HpiA) were the best adsorbed fractions by
the polymer (27% and 33% respectively). However, the effect of the combination treatment at
removing NOM resulted in the polyurethane exhibiting a relatively good capability to remove NOM
from water as evidenced by an up to 88% reduction of the HpiA fraction.
Acknowledgements
Funding obtained from the National Research Foundation (NRF) of South Africa, ESKOM’s
Tertiary Support Program (TESP), the University of Johannesburg and the DST/Mintek NIC water
platform is gratefully acknowledged.
References
1. Bender M.L. and Komiyama M., 1978. Cyclodextrin chemistry, Springler-Verlag Berlin
Heidelberg, NewYork. 1-3.
2. Bose P. and Reckhow D.A., 2007. The effect of ozonation on natural organic matter removal
by alum coagulation. Water Research 41, 1516-1524.
3. Chen J., Gu B., LeBoeuf E., Pan H. and Dai S., 2002. Spectroscopic characterization of the
structural and functional properties of natural organic matter fractions. Chemosphere 48, 5968.
4. De Wuilloud J.C., Sadi B.B. and Caruso J.A., 2003. Trace humic and fulvic acid determination
in natural water by cloud point extraction/pre-concentration using non-ionic and cationic
surfactants with FI-UV detection. The Analyst 128, 453-458.
5. Hepplewhite C., Newcombe G. and Knappe D.R.U., 2004. NOM and MIB, who wins in the
competition for activated carbon adsorption sites? Water Science and Technology 49,257–267.
6. Kanokkantapong V., Marhaba T.F., Pavasant P, and Panyapinyophol B., 2006.
Characterization of haloacetic acid precursors from source waters. The Journal of
Environmental Management 80, 214-221.
7. Karnik B.S., Davies S.H., Baumann M.J. and Masten S.J., 2005. The effects of combined
ozonation and filtration on disinfection by-product formation. Water Research 39, 2839-2850.
8. Kiwa N.V., 2006. Selection of Anionic Exchange Resins for NOM Removal. BTO 042, 9-16.
9. Li D. and Ma M., 2000. Nanosponges for water purification. Clean Products and Processes 2,
112-116.
10. Mamba B.B., Krause R.W., Malefetse T.J., Sithole S.P. and Nkambule T.I., 2008. Humic acid
as a model for natural organic matter (NOM) in the removal of odorants from water by
cyclodextrin polyurethanes. Submitted for publication. Journal of physics and chemistry of the
earth.
11. Marhaba T.F., Pu Y. and Bengraine K., 2003. Modified dissolved organic matter fractionation
technique for natural water. Journal of Hazardous Materials B101, 43-53.
12. Marhaba T.F., Doanh V., 2000. The variation of mass and disinfection by-product formation
potential of dissolved organic matter fractions along a conventional surface water treatment
plant. Journal of Hazardous Materials A74, 133-147.
13. Mcdonald S., Bishop A.G., Prenzler P.D. and Robards K., 2004. Analytical chemistry of
freshwater humic substances. Analytica Chimica Acta 527, 105-124.
14. Mhlanga S.D., Mamba B.B., Krause R.W. and Malefetse T.J., 2007. Removal of organic
contaminants from water using nanosponges cyclodextrin polyurethanes. Journal of Chemical
Technology and Biotechnology 82, 382-388.
15. Vanboon G.W., Duffy S.J., 2005. Environmental Chemistry: a Global Perspective. Oxford
university press, 2nd edition, 257-320.
16. Xie Y.F. (2003). Disinfection By-products in drinking water. Boca Raton, Lewis publishers. 815.
17. Yavich A., Lee K., Chen K. and Pape L., 2004. Evaluation of biodegradability of NOM after
ozonation. Water Research 38, 2839-2846.
18. Zhang T., Lu J., Ma J. and Qiang Z., 2008. Comparative study of ozonation and synthetic
goethite-catalyzed ozonation of individual NOM fractions isolated and fractionated from
filtered river water. Water Research 42, 1563-1570.