You are on page 1of 9

Journal of Environmental Management 73 (2004) 155–163

www.elsevier.com/locate/jenvman

Combined chemical and biological oxidation of penicillin


formulation effluent
I. Arslan Alaton*, S. Dogruel, E. Baykal, G. Gerone
Department of Environmental Engineering, Faculty of Civil Engineering, Istanbul Technical University, Room Nr. K320, 34469 Maslak, Istanbul, Turkey
Received 18 July 2003; revised 19 February 2004; accepted 29 June 2004

Abstract
Antibiotic formulation effluent is well known for its important contribution to environmental pollution due to its fluctuating and
recalcitrant nature. In the present study, the chemical treatability of penicillin formulation effluent (average filtered CODoZ830 mg/l;
average soluble CODoZ615 mg/l; pHoZ6.9) bearing the active substances penicillin Amoxicillin Trihydrate (C16H19N3O5S$3H2O) and the
b-lactamase inhibitor Potassium Clavulanate (C8H8KNO5) has been investigated. For this purpose, the penicillin formulation effluent was
subjected to ozonation (applied ozone doseZ2500 mg/(l!h)) at varying pH (2.5–12.0) and O3CH2O2 (perozonation) at different initial
H2O2 concentrations (Z2–40 mM) and pH 10.5. According to the experimental results, the overall Chemical Oxygen Demand (COD)
removal efficiency varied between 10 and 56% for ozonation and 30% (no H2O2) and 83% (20 mM H2O2) for the O3CH2O2 process. The
addition of H2O2 improved the COD removal rates considerably even at the lowest studied H2O2 concentration. An optimum H2O2
concentration of 20 mM existed at which the highest COD removal efficiency and abatement kinetics were obtained. The ozone absorption
rate ranged between 53% (ozonation) and 68% (perozonation). An ozone input of 800 mg/l in 20 min was sufficient to achieve the highest
BOD5/COD (biodegradability) ratio (Z0.45) and BOD5 value (109 mg/l) for the pre-treated penicillin formulation effluent. After the
establishment of optimum ozonation and perozonation conditions, mixtures of synthetic domestic wastewaterCraw, ozonated and
perozonated penicillin formulation effluent were subjected to biological activated sludge treatment at a food-to-microorganisms (F/M) ratio
of 0.23 mg COD/(mg MLSS!d), using a consortium of acclimated microorganisms. COD removal efficiencies of the activated sludge
process were 71, 81 and 72% for pharmaceutical wastewater containing synthetic domestic wastewater mixed with either raw, ozonated or
perozonated formulation effluent, respectively. The ultimate COD value obtained after 24-h biotreatment of the synthetic domestic
wastewaterCpre-ozonated formulation effluent mixture was around 100 mg/l instead of 180 mg/l which was the final COD obtained for the
wastewater mixture containing raw formulation effluent, indicating that pre-ozonation at least partially removed the non-biodegradable COD
fraction of the formulation effluent.
q 2004 Elsevier Ltd. All rights reserved.

Keywords: Advanced oxidation processes; Biodegradability improvement; Combined chemicalCbiological treatment; Pharmaceutical wastewater;
Ozonation; O3CH2O2 process; Penicillin formulation effluent

1. Introduction exert toxic effects to living organisms even in the ng/l–mg/l


concentration range (Halling-Sorensen, 2000). Among all
Recently, the presence of pharmaceuticals in sewage other pharmaceutical drugs and substances, the literature on
effluent—unmodified or in the form of metabolites—has antibiotics is much more developed due to its serious
been reported and documented as an emerging risk to the irreversible effects on the aquatic and terrestrial environment
biotic environment (Editorial, 2000; Ternes et al., 2003). (Kümmerer et al., 2000; Reinthaler et al., 2003).
Some available data on antibiotics indicate their capability to The toxic effects of antibiotics on several organisms, e.g.
bacteria, algae, and water flea (Daphnia magna), have been
found not only in high concentrations, but also in low
* Corresponding author. Tel.: C90-212-285-65-76; fax: C90-212-285-
65-87. concentrations and in chronic tests. For instance, Richard-
E-mail address: arslanid@itu.edu.tr (I.A. Alaton). son and Brown (1985) examined a number of antibiotics
0301-4797/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jenvman.2004.06.007
156 I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163

such as ampicillin, erythromycin, sulphamethoxazole,


tetracycline and penicilloyl groups for their biodegradabil-
ity during sewage treatment. Their study indicated that these
antibiotics were practically non-biodegradable and had the
potential to survive sewage treatment. These properties lead
to persistence of these compounds in the environment and
potential for bio-accumulation (Wollenberger et al., 2000).
Studies concerning antibiotics in the aquatic environment
have also clearly shown that elimination in municipal
sewage treatment plant is often incomplete (Kümmerer
et al., 1997, 2000). More recently, we have reported that
pharmaceutical wastewater from the antibiotic formulation
process has BOD5 (Biochemical Oxygen Demand) values of
less than 20 mg/l but COD (Chemical Oxygen Demand)
values in the range of 800–1500 mg/l. Hence, antibiotic
formulation effluent can be regarded as practically non- Fig. 1. Molecular structures of Amoxicillin Trihydrate (a) and Potassium
biodegradable. Clavulanate (b).
Although ozone is a strong oxidizing agent and
disinfectant, it may not effectively work for more complex a pharmaceutical company located in Istanbul, Turkey.
and highly concentrated ingredients found in industrial According to the technical staff of the treatment plant, the
wastewater. Acceleration of ozone decomposition to more selected formulation effluent corresponded to approxi-
active oxidants, i.e. free radicals such as OH% and HO%2, mately 30% of the total daily effluent (Z150 m3/d), of
upon H2O2 addition has received particular interest for its which the remaining portion was wastewater of domestic
potential applicability to a variety of medium-strength nature. The collected effluent was stored at 2–4 8C before
industrial effluents (Vel Leitner et al., 1994; Gulyas et al., use for up to 4 weeks. Prior to all experiments, effluent
1995; Beltran et al., 1998; Zwiener and Frimmel, 2000; suspensions were filtered through glass fiber filters with an
Mitani et al., 2002). In the present experimental work, the average pore size of 1.2 mm to remove particulate matter
advanced oxidation of a penicillin formulation effluent that may interfere with the ozonation reactions. The
bearing Amoxicillin Trihydrate and Potassium Clavulanate penicillin formulation composition and resulting effluent
as the active substances has been investigated by means of are summarized in Tables 1 and 2, respectively.
ozonation at varying pH as well as the ‘perozonation’
process (H2O2-enhanced ozonation) at different H2O2
2.2. Ozonation experiments
concentrations and a fixed pH. Treatment performances
have been comparatively examined in terms of COD
One liter-penicillin formulation effluent was ozonated for
abatement kinetics. After optimization of the above-
1 h in a borosilicate glass bubble column wherein the
indicated ozonation processes, raw and chemically pre-
ozoneCoxygen gas mixture was continuously sparged at a
treated penicillin formulation effluent were exposed to
rate of 1.3 l/min through a fritted glass dispersion disc (5 cm
biological activated sludge treatment in order to examine
in diameter). Ozone was produced by a corona discharge
the effect of chemical pre-treatment on changes in the
PCI GL-1 model pilot scale ozone generator with a
biological degradation efficiency. This information is of
maximum ozone production capacity of 20 g/h. Teflon
particular importance to highlight whether advanced
tubing was used for all connections from the ozone
chemical oxidation is capable of improving the biocompat-
generator to the reaction vessel. All excess (unreacted)
ibility of the studied industrial wastewater or may serve as a
gaseous ozone leaving from the reactor top was collected in
stand-alone treatment option.
two gas washing bottles connected in series and filled with
Table 1
2. Materials and methods Ingredients of the penicillin formulation effluent

Formulation componenta Function


2.1. Penicillin formulation effluent and analytical reagents
Amoxicillin Trihydrate Penicillin active substance
Potassium Clavulanate Assistant active substance
The penicillin effluent used in this study originated from Croscarmellose Sodium (dried) Flavor
the BIOMED BIDw formulation (active ingredients: HDK-N20 Binding agent
Amoxicillin Trihydrate; C16H19N3O5S$3H2O, molecular Sodium Stearyl Fumarate Additive
weightZ419.4 g/mol, and the b-lactamase inhibitor Avicel pH 11.2 Buffering agent
Potassium Clavulanate; C8H8KNO5, molecular weightZ a
The concentration of active agents and formulation additives is strictly
237.3 g/mol, shown in Fig. 1a and b respectively), at confidential.
I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163 157

Table 2 COD removal rates followed first order kinetic behavior


Environmental characterization of the penicillin formulation effluent during the ozonation and perozonation processes. First order
Parameter Value COD removal rate constants were calculated in the
COD (mg/l) a
830 following manner:
Soluble COD (mg/l)b 615
DOC (mg/l) 450 COD=CODo Z expðKkCOD !tÞ (3)
Suspended solids (mg/l) 150
Volatile suspended solids (mg/l) 65
where kCOD stands for the first order COD removal rate
Alkalinity (mg/l CaCO3) 80
Chloride (mg/l ClK) 85 constant (timeK1) and COD/CODo represents normalized
TKN (mg/l) 36 COD as a function of ozonation time t (in min).
NH3–N (mg/l) 2.0
pH (–) 6.9
Electrical conductivity (ms/cm) 600
2.4. Biodegradation experiments
Total p (mg/l) !0.1
Color (1/cm)c 0.017
a
BOD5 was measured in raw as well as chemically pre-
Filtered through 1.20 mm-cutoff filter paper (filtered COD). treated effluent using sewage sludge acclimated to raw,
b
Filtered through 0.45 mm-cutoff filter paper (soluble COD).
c
Measured as absorbance at lZ436 nm.
ozonated and perozonated wastewater samples in order to be
able to estimate the most appropriate ozonation time (ozone
10% KI solution in order to determine the amount of off-gas dose) to achieve the highest possible biotreatability level in
ozone production. Two other gas washing bottles filled with the present study. However, BOD5 measurements do not
2% KI solution were placed in another ozonation line in mirror the actual situation in biological treatment units and
order to determine the ozone input rate. Ozone was fed to hence were only used as an approximate estimation of
one of the lines (input line or ozonation line) by means of a biodegradability improvement.
three-way valve. The ozone dose was set as 2500 mg/(l!h) In order to better assess the biotreatability of the pre-
for all ozonation and perozonation experiments. The input ozonated effluents, modified biodegradation experiments
and off-gas ozone concentrations were determined iodo- were carried out in addition to the above mentioned BOD5
metrically (IOA, 1987) to calculate ozone absorption rates measurements. For that purpose, raw (untreated), ozonated
(O3A values) for each ozonation experiment as follows: and perozonated penicillin formulation effluent that was
previously mixed with synthetic domestic wastewater at a
O3A ðmgÞ Z input O3 ðmgÞ K off  gas O3 ðmgÞ (1) ratio typical of the selected pharmaceutical factory (i.e. at a
3:7 volumetric ratio) making up a total of 1 l mixed
O3A ð%Þ Z fabsorbed O3 ðmgÞ=input O3 ðmgÞg !100 (2) pharmaceutical wastewater, was subjected to 24-h biode-
gradation experiments. For that purpose, a microbial
consortium being previously fed with synthetic domestic
2.3. Analytical procedures wastewater was first prepared in accordance with a modified
version of ISO Standard Ref. Nr. 8192 (1986) and
All experimental analyses were conducted in accordance acclimated to a mixture of 70% v/v synthetic domestic
with standard methods (APHA/AWWA/WPCF, 1998) wastewater (carbon sourceZmeat extract and peptone) plus
except dissolved organic carbon (DOC) analysis which 30% v/v raw, ozonated and perozonated penicillin formu-
was accomplished using a Shimadzu TC-1000 model total lation effluent. Acclimation of sewage sludge to mixed
organic carbon analyzer and spectrophotometric measure- pharmaceutical wastewater samples was carried out in three
ments which were performed on a Shimadzu UV-120 model different 4-l capacity fill-and-draw reactors that were
single-beam spectrophotometer. Prior to DOC, color initially operated at a sludge age of 7 days. After 9 weeks,
(measured at lZ436 nm) and UV absorbance (measured the sludge age was fixed at 10 days. It took 3 months to
at lZ344 and 274 nm) analyses, sample aliquots were acclimate the mixture containing raw penicillin formulation
filtered through 0.45 mm syringe membranes and their pH effluent. The biodegradation experiments were conducted
was adjusted to pH 7G0.2 in order to obtain clear, neutral by simultaneously operating three activated sludge reactors
pH solutions. All measurements were done at least in (reactor volumesZ1 l; food-to-microrganisms ratioZ
duplicate. 0.23 mg COD/mg MLVSS/d; initial COD of the domestic
A comprehensive environmental characterization of the wastewater fraction in the mixtureZ480 mg/l; final COD of
penicillin formulation effluent was only performed for the wastewater containing 100% synthetic domestic wastewater
raw effluent to decide on the most appropriate chemical after 24 h biodegradation %25 mg/l) where the three
oxidation conditions. Only (1.2 mm) filtered COD, soluble pharmaceutical wastewater mixtures were treated for 24 h.
COD (DCOD) and pH were measured throughout all Effluent COD was followed at regular time intervals (i.e.
experiments to compare the efficiencies of different after 0.5, 1, 2, 4, 6, 12 and 24 h of activated sludge
advanced oxidation processes. treatment) throughout the biodegradation experiments.
158 I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163

3. Results and discussion Table 3


First order COD removal rate constants (kCOD values, presented in hK1) and
percent overall COD removal efficiencies at varying pH
3.1. Pre-treatment with O3: effects of initial pH Initial pH Final pH kCOD COD removal
(–) (–) (hK1) efficiency (%)
Ozonation experiments were performed for 1 h at four
2.5 2.1 0.114 10
different initial pH values, i.e. at pH 2.5, 6.9 (i.e. the actual 6.9 6.4 0.168 15
pH of the formulation effluent), 10.5 and pH 12 in un- 10.5 8.6 0.336 30
buffered samples. During all experiments, the pH decreased 12.0 8.2 0.804 56
considerably as a consequence of acid product formation. Experimental conditions: ozonation timeZ1 h; applied ozone doseZ
Fig. 2 presents normalized COD removal rates as a function 2500 mg/l h. The correlation coefficient R2 for kCOD determination was
of ozone dose. As was expected, the COD removal always R0.90.
increased with increasing pH due to accelerated ozone
decomposition to free radicals which is particularly 3.2. Pre-treatment with O3/H2O2: effects of initial H2O2
concentration
pronounced at pH R11.0 (Staehelin and Hoigné, 1982).
For instance at acidic pH, when direct oxidation by ozone is
Staehelin and Hoigné (1982) have demonstrated that the
the dominant reaction pathway, only 10% COD removal
conjugate base of H2O2 (HOK 2 ) can enhance ozone
was found revealing that the degradation of penicillin
decomposition resulting in OH% formation at a faster rate
formulation mainly proceeds via the indirect, free radical
than in the case of OHK initiation. In addition, as can be
oxidation pathway. It can be concluded that the COD
concluded from the net reaction (Eq. (4), the OH%
abatement mechanism of ozonation seemed to change and production yield is 1.0 mol OH%/mol O3 instead of only
was more efficient at elevated pH values, especially when 0.67 mol OH%/mol O3 for the OHK-enhanced initiation
the pH was above 10.5. reaction.
Table 3 lists first order COD removal rate constants The overall reaction for H2O2-mediated O3 decompo-
(kCOD values) and percent total COD removal efficiencies sition is (Glaze et al., 1987):
obtained at varying initial ozonation pH’s together with the
pH values measured before and after 1 h ozonation. From 2O3 C H2 O2 / 2OH$ C 3O2 (4)
the table, it can be inferred that the increase in kCOD values Accordingly, the theoretical H2O2:O3 molar ratio for
paralleled the decrease in pH and COD removal efficiencies perozonation is 1:2. H2O2-initiated ozone decomposition
with the same tendency observed with increasing initial and initiation of free radical chain reactions are given below
reaction pH. (Staehelin and Hoigné, 1982):
H2 O2 % HOK
2 CH
C
pKa Z 11:8 (5)

k2
2 C O3 
HOK / HO$2 C O$K
3 k2 Z 2:8 !106 l=ðmole !sÞ
(6)

HO$2 % O$K
2 CH
C
pKa Z 4:8 (7)

k3
2 C O3 
O$K / O$K
3 C O2 k3 Z 1:6 !109 l=ðmole !sÞ
(8)

k4
HC C O$K
3 % HO3
$
k4 Z 5:0 !1010 l=ðmole !sÞ (9)

k5
HO$3 
/ OH$ C O2 k5 Z 1:4 !105 l=ðmole !sÞ (10)

k6
2 
HO$2 C O$K / HOK
2 C O2 k6 Z 8:7 !107 l=ðmole !sÞ
(11)
The reaction rate dependence on H2O2 concentration
Fig. 2. COD abatement rates at varying initial pH values. Experimental suggests that H2O2 is acting as a free radical scavenger
conditions: ozonation timeZ1 h; soluble CODoZ637 mg/l, filtered at concentrations above the theoretical, optimum require-
CODoZ838 mg/l; applied ozone doseZ2500 mg/l h. ment, i.e. at excess H2O2 concentrations. In this case,
I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163 159

the following radical scavenging reactions may become 30 mM, COD removal rates began to decrease revealing
important (Buxton et al., 1988): that in accordance with reactions (12) and (13), inhibition of
k7
COD removal was observed when H2O2 was applied in
H2 O2 C OH$ 
/ HO$2 C H2 O C O2 excess and started to consume OH%. The presence of an
(12)
k7 Z 2:7 !107 l=ðmole !sÞ optimum, critical H2O2 concentration or concentration
range for the O3/H2O2 process that usually coincided with
k8
the theoretical, stoichiometric requirement given in Eq. (4),
2 C OH 
HOK  OHK C HO$2
$
/ has already been evidenced in previous work (Beltran et al.,
(13)
k8 Z 7:5 !109 l=ðmole !sÞ 1998; Mitani et al., 2002). In Fig. 3, it is evident that the
inhibition of COD removal at [H2O2]oZ40 mM was only
Instead of OH%-initiated oxidative chain reactions with observed in the later stages of perozonation, i.e. at
organic pollutants symbolized as R: tR30 min. The ‘delayed’ inhibition of COD removal is
k7 speculatively attributable to the formation and accumulation
OH$ C R 
/ ROH$  $2 
/RO 
// of more selective HO%2 oxidation products in accordance
(14)
k9 Z 108 –1011 l=ðmole !sÞ with reaction (12).
Table 4 summarizes kCOD values and percent overall
H2O2/O3 experiments were conducted at six different COD removal efficiencies obtained at different initial H2O2
initial H2O2 concentrations in the range of 2–40 mM and concentrations. For perozonation in the presence of 40 mM
pHoZ10.5. This range was selected considering typical H2O2, only the first 30 min were considered for kinetic
concentrations that have been investigated in previous evaluation of COD abatement rates.
experimental studies (Gulyas et al., 1995; Arslan-Alaton From Table 4, it can be inferred that although the best
and Akmehmet Balcioglu, 2002) and the fact that the ozone results were obtained in the presence of 20 mM H2O2, it
feed rate selected for the present work made up a total ozone would be more feasible to select 2 mM for pre-treatment
dose of 52 mM (Z2500 mg/l) after 1 h ozonation. Again, process economy. Perozonation treatment with 2 mM H2O2
the reaction pH was not controlled throughout the process resulted in comparably good COD abatement kinetics and
because it decreased to pH z8 at the end of the reaction removal efficiencies. On the other hand, it should be kept in
period which is a pH range required for the succeeding mind that because of the fact that a second, biological
biotreatment step. treatment step would follow the perozonation process, COD
Fig. 3 depicts changes in normalized COD values during ‘over-oxidation’ should be minimized. Consequently, 2 mM
perozonation at different initial H2O2 concentrations. Final H2O2 was selected as the optimum for the perozonation
COD/CODo values were obtained as 0.70, 0.24, 0.29, 0.35, process at pH 10.5.
0.17,0.30 and to 0.63 for initial concentrations of 0, 2, 5, 10, Table 5 shows O3A values and specific COD removal
20, 30 and 40 mM H2O2, respectively. From Fig. 3, it can be rates for ozonation at pH 10.5 and varying H 2O 2
stated that even at the lowest studied H2O2 concentration, concentrations. From Table 5, it is obvious that the specific
COD removal kinetics were significantly enhanced and the COD removal was highest for perozonation at 2 mM H2O2,
OH%-initiated chain reaction mechanism was favored for the speaking for the most efficient use of ozone at this H2O2
chemical, oxidative pre-treatment of the formulation concentration. Ozone absorption rates on the other hand
effluent. However, at H2O2 concentrations exceeding increased gradually with increasing H2O2, indicating that an
increase in ozone absorption rate does not necessarily result
in a higher COD removal efficiency. 51% of the applied
ozone (Z1272 mg) was absorbed during perozonation in
Table 4
First order COD removal rate constants and percent COD removal
efficiencies at varying initial H2O2 concentrations

[H2O2]o kCOD COD removal


(mM) (hK1) efficiency (%)
0 0.336 30
2 1.194 76
5 1.218 71
10 1.290 64
20 1.668 82
30 1.164 70
40 1.344 37
Fig. 3. COD abatement rates at varying initial H2O2 concentrations.
Experimental conditions: ozonation timeZ1 h; soluble CODoZ615 mg/l; Experimental conditions: ozonation timeZ1 h; applied ozone doseZ
filtered CODoZ835 mg/l; applied ozone doseZ2500 mg/l h; pHo 10.5 2500 mg/l h; initial pH 10.5; final pH 8.4–8.6. The correlation coefficient R2
(no pH buffer was used). for kCOD determination was always R0.90.
160 I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163

Table 5
Ozone absorption (O3A) and specific COD removal rates at varying initial
H2O2 concentrations

[H2O2]o O3A Specific COD removal rate


(mM) (mg) (mg DCOD/mg O3A)
0 1339 0.14
2 1272 0.52
5 1368 0.37
10 1344 0.30
20 1344 0.48
30 1632 0.39
40 1512 0.16

Experimental conditions: ozonation timeZ1 h; applied ozone doseZ


2500 mg/l h; initial pH 10.5; final pH 8.4–8.6.

the presence of 2 mM H2O2 revealing that 1.92 mg O3 was


absorbed to remove 1 mg of COD.

3.3. Comparison of optimized O3/OHK and O3/H2O2


processes

From the experiments described in the previous sections,


it could be demonstrated that an optimum pH (12.0) and
H2O2 concentration (Z2 mM at pH 10.5) existed for the
ozonation and perozonation processes, respectively. How-
ever, for the selection of the most suitable chemical pre-
treatment conditions, determination of a pollutant-specific
optimum ozone input (i.e. ozonation time) to achieve the
highest biodegradability becomes very critical in a sequen-
tial chemicalCbiological treatment approach. In order to
serve this purpose, changes in UV (ultra violet) optical
density and BOD5 rates or, alternatively, BOD5/COD ratios,
observable during chemical oxidation can be followed
providing approximate, preliminary tools to evaluate
structural modifications and biodegradability changes,
respectively.
Fig. 4 depicts changes in UV absorbance (a), BOD5 (b)
and BOD5/COD (c) values for O3/pH 12.0 and O3/2 mM
H2O2/pH 10.5 processes as a function of ozonation time. The
behavior of UV absorbance at the two characteristic Fig. 4. Changes in characteristic UV absorbance (a), BOD5 (b) and
wavelengths during ozonation and perozonation is displayed BOD5/COD (c) values during ozonation (pHo 12.0) and perozonation (pHo
in Fig. 4a. Two distinct UV absorption band maximum values 10.5; H2O2oZ2 mM) of penicillin formulation effluent. Experimental
were detected for the penicillin formulation effluent at the conditions: ozonation timeZ1 h; soluble CODoZ620 mg/l for ozonated
and 615 mg/l for perozonated formulation effluent; filtered CODoZ
wavelengths lZ344 and 274 nm. These absorption bands 880 mg/l for ozonated and 873 mg/l for perozonated formulation effluent;
represent the aromatic and olefinic groupings of the applied ozone doseZ2500 mg/l h.
formulation based effluent samples. As is evident form the
figure, an abrupt increase (significant for the ozonated for O3 and O3/H2O2, respectively. However, upon closer
sample) could be observed after 5 and 10 min of chemical inspection of the figure it can be seen that the increase in
pre-treatment indicating the formation of intermediates that BOD5 values practically stopped after 20 min correspond-
exert higher UV absorbance than the parent compounds ing to an ozone input of 800 mg/l. Unlike the respective
present in the original effluent. Fragmentation of these early BOD5 profiles, BOD5/COD ratios obtained for both
oxidation products occurred in the later stages of both the oxidation processes continued to increase throughout the
ozonation and perozonation processes. However, overall UV entire reaction period from an initial value of 0.06–0.38
removal was rather incomplete. after single ozonation and from 0.02 to 0.45 for perozona-
Fig. 4b indicates that after 1 h ozonation, the BOD5 tion (Fig. 4c) as a consequence of the parallel decreasing
values increased from 35 to 101 mg/l and from 19 to 96 mg/l trend of COD values. After 20 min chemical pre-treatment,
I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163 161

COD values dropped to around 400–550 mg/l for the


perozonation and ozonation processes, respectively.
The specific COD removal rate was appreciably higher for
the perozonation process (Z0.52 mg COD/mg O3A) than
for single ozonation (Z0.27 mg COD/mg O3A), although
O3A values increased in parallel to the applied, initial H2O2
concentrations.
From Fig. 4c, it is also evident that acceptable
improvements in the BOD5/COD ratios were achieved
after 20 min ozonation and perozonation. Ozonation and
perozonation in the presence of 2 mM H2O2 did not proceed
further and an ozonation time of 20 min was selected for the
combined chemicalCbiological treatment of the effluent.
Fig. 5. Biological treatment of simulated pharmaceutical wastewater
containing synthetic domestic wastewaterCraw, ozonated and perozonated
3.4. Biological treatment of raw, ozonated and perozonated penicillin formulation effluent at a volumetric ratio of 700:300 (Z1000 ml).
penicillin formulation effluent Chemical oxidative pre-treatment conditions are as in Fig. 4 except that
ozonation and perozonation time were selected as 20 min (applied ozoneZ
Like many other industrial effluents, pharmaceutical 800 mg/l). Biological treatment conditions: treatment timeZ24 h;
F/MZ0.23 mg COD/(mg MLSS!d); CODoZ627 mg/l (mixed raw);
wastewater usually contains high amounts of toxic and/or CODoZ522 mg/l (mixed ozonated); CODoZ468 mg/l (mixed perozo-
recalcitrant compounds and hence chemical pre-treatment is nated); pH 7.0–7.2.
often required. However, integrated chemical–biological
treatment of industrial effluents is generally found to derive penicillin formulation effluent was only partially eliminated
both economic and technical advantages over chemical after chemical pre-treatment.
oxidation alone (Scott and Ollis, 1995). Studies have
already been conducted successfully coupling biologicalC
chemical oxidation for the treatment of originally refractory 4. Summary and conclusions
industrial pollutants in wastewater (Gulyas et al., 1995;
Stockinger et al., 1995; Jochimsen et al., 1997; Benitez In the present study, combined chemical (O3/OHK and
et al., 2001; Akmehmet Balcioglu and Arslan, 2001; O3/H2O2) and biological oxidation of penicillin formulation
Arslan-Alaton and Akmehmet Balcioglu, 2002; Hess and effluent was investigated. After optimization of the ozona-
Schrader, 2002). To our knowledge, the integrated chemi- tion and perozonation processes, raw and chemically
calCbiological treatment of real antibiotic formulation pre-treated effluent samples were exposed to biological
wastewater has not been reported so far. treatment using acclimated sewage sludge. The following
Simulated pharmaceutical wastewater containing raw, conclusions could be drawn from the experimental work:
ozonated (pH 12; 800 mg/l O3) and perozonated (pH 10.5;
2 mM H2O2; 800 mg/l O3) effluent from the penicillin † COD removal rate and efficiency increased with
formulation process was subjected to biological activated increasing pH as a consequence of enhanced O3
sludge treatment. Changes in COD abatement rates for the
Table 6a
effluent samples during 24 h-biological treatment are
COD abatement rates during 24 h biodegradation experiments for
depicted in Fig. 5. From Fig. 5, it can be seen that biological simulated pharmaceutical wastewater containing raw, ozonated (applied
COD removal was more complete for the chemically pre- ozoneZ800 mg/l in 20 min; pH 12.0) and perozonated (initial H2O2
treated effluent samples, although the difference is not very concentrationZ2 mM; applied ozoneZ800 mg/l in 20 min; pH 10.5)
significant. The almost 2-fold increase in COD removal rates Biological treatment Raw sample Ozonated Perozonated
obtained for the perozonated effluent samples as compared to time (h) sample sample
the biodegradation rates of untreated (raw) pharmaceutical 0.0 627 522 468
wastewater is worth mentioning. This observation speaks to a 0.5 603 508 424
considerable reduction in microbial inhibition and recalci- 1.0 557 481 274
trance. The effluent sample which was previously ozonated 2.0 502 342 128
still had a COD value of 99 mg/l and a final COD of 130 mg/l 4.0 407 215 104
6.0 219 82 133
remained after biotreatment of the wastewater containing 12.0 208 112 132
perozonated effluent (Table 6a). Percent COD removal 24.0 180 99 130
efficiencies obtained after 24 h biological treatment were 71,
Experimental conditions: biotreatment timeZ24 h with sewage sludge
81 and 72% for samples containing synthetic domestic
acclimated to raw, ozonated and perozonated effluent; F/M ratioZ0.23 g
wastewater mixed with raw, ozonated and perozonated COD/(g MLVSS!d); initial volumetric ratio of the synthetic domestic
penicillin formulation effluent, respectively (Table 6b). effluent fraction in the pharmaceutical wastewater mixtureZ70% v/v; total
These results revealed that the biologically inert fraction of simulated pharmaceutical wastewater volumeZ1 l.
162 I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163

Table 6b
Treatment performance of stand-alone biological treatment of simulated pharmaceutical effluent mixture, ozonation of penicillin formulation effluentC
activated sludge treatment of whole pharmaceutical effluent as well as perozonation of penicillin formulation effluentCactivated sludge treatment of whole
pharmaceutical effluent

Raw Ozonated Perozonated


COD removed after 24 h biotreatment (mg/l) 447 423 338
Biological COD removal efficiencies (%) 71 81 72
Overall absolute chemical and biological COD removal (mg/l) 447 743 802
COD removal efficiencies of combined chemicalCbiological 71a 84b 79b
oxidation (%)
COD of penicillin formulation effluent before mixing; PFE (mg/l) 813 560 416
COD of synthetic domestic wastewater before mixing; SDW (mg/l) 547 505 490
COD of the simulated pharmaceutical wastewater mixture; PFE!0. 627 522 468
3CSDW!0.7ZPW (mg/l)

Experimental conditions as in Table 6a.


a
Subjected to biological treatment without any pre-treatment.
b
Initial COD is the COD of the pharmaceutical wastewater mixture containing untreated (raw) synthetic domestic wastewater and untreated (raw) penicillin
formulation effluent; final COD is the COD of the pharmaceutical wastewater mixture after biological treatment containing untreated synthetic domestic
wastewater and ozonated or perozonated penicillin formulation effluent.

decomposition to free radicals at elevated pH values. The in the final, treated effluent. Overall COD removal
COD abatement mechanism of ozonation seemed to efficiencies of 84 (O3) and 79% (O3/H2O2) were obtained
change and was more efficient at high pH values, for chemical oxidation of segregated penicillin formu-
especially when exceeding pH 10.5. The highest COD lation effluent combined with biological oxidation of the
removal efficiency (Z56%) was achieved at the highest whole pharmaceutical effluent.
pH tried (12.0).
† Even the lowest studied H2O2 concentration resulted in From the results obtained via combined chemical and
an appreciable enhancement in COD removal rates. COD biological oxidation of penicillin formulation effluent, it can
removal efficiency increased from 30% without H2O2 be stated that advanced oxidation has proven to be
addition to 76% in the presence of 2 mM H2O2 at pH reasonably suited and quite feasible for application as a
10.5 and an ozone dose of 2500 mg/(l!h). The highest pre-treatment method. However, a non-degradable fraction
COD removal efficiency obtained was 83% for perozo- remained in the chemically pre-treated effluent that may
nation at a H2O2 concentration of 20 mM. The COD inhibit or at least retard a succeeding biological treatment
removal rate obtained by perozonation with 20 mM system. This led us to the conclusion that factories may
H2O2 was about 50 times faster than ozonation alone at consider applying advanced oxidation as a stand-alone
pH 10.5. When the H2O2 concentration exceeded option whenever stream segregation into biodegradable and
20 mM, COD removal was inhibited seriously, revealing non-biodegradable fractions is possible so that effluent
the presence of an optimum, critical H2O2 concentration. volume and hence treatment costs can be reduced
† Preliminary experiments that were conducted to select an considerably.
optimum ozone dose to obtain the highest possible
biological treatment efficiency were simply based on Acknowledgements
BOD5 measurements. The BOD5/COD (biodegradabil-
ity) ratio increased by a factor of 6 and 23 for ozonation The authors wish to thank TUBITAK (Turkish National
at pH 12.0 and perozonation with 2 mM H2O2 at pH 10.5, Scientific and Technology Research Council) for the
respectively, at an ozone input of 800 mg/l (20 min financial support under project: ‘ICTAG—C-042’.
ozonation).
† Pharmaceutical wastewater containing raw, ozonated or
perozonated formulation effluent was biotreated to References
simulate conditions of a pharmaceutical treatment plant
that would receive one of these effluents. Biological Akmehmet Balcioglu, I., Arslan, I., 2001. Partial oxidation of reactive
dyestuffs and a synthetic textile dye-bath by the O3 and O3/H2O2
COD removal rates and efficiencies were appreciably processes. Water Sci. Technol. 43 (2), 221–228.
improved in particular when penicillin effluent was APHA–AWWA–WPCF, 1998. Standard Method for the Examination of
pre-treated with the optimized O3/H2O2 processes. Water and Wastewater, 20th ed American Public Health Association,
However, the pharmaceutical wastewater could not be Washington, DC.
Arslan-Alaton, I., Akmehmet Balcioglu, I., 2002. The effect of pre-
completely degraded even with the most effective ozonation on the H2O2/UV-C treatment of raw and biologically
chemicalCbiological treatment combination and an pre-treated textile industry wastewater. Water Sci. Technol. 45 (12),
inert, non-biodegradable COD fraction always remained 297–304.
I.A. Alaton et al. / Journal of Environmental Management 73 (2004) 155–163 163

Beltran, F.J., Garcia-Araya, J.F., Alvarez, P.M., Rivas, J., 1998. Kümmerer, K., Al-Ahmad, A., Mersch-Sundermann, V., 2000. Biodegrad-
Aqueous degradation of atrazine and some of its main by-products ability of some antibiotics, elimination of the genotoxicity and affection
with ozone/hydrogen peroxide. J. Chem. Technol. Biotechnol. 71, of wastewater bacteria in a simple test. Chemosphere 40, 701–710.
345–355. Mitani, M.M., Keller, A.A., Bunton, C.A., Rinker, R.G., Sandall, O.C.,
Benitez, F.J., Acero, J.L., Gonzalez, T., Garcia, J., 2001. Ozonation and 2002. Kinetics and products of reactions of MTBE with ozone and
biodegradation processes in batch reactors treating black table olives hydrogen peroxide in water. J. Haz. Mat. B89, 197–212.
washing wastewaters. Ind. Eng. Chem. Res. 40, 3144–3151. Reinthaler, F.F., Posch, J., Feierl, G., Wüst, G., Haas, D., Ruckenbauer, G.,
Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical Mascher, F., Marth, E., 2003. Antibiotic resistance of E. coli in sewage
review of data constants for reactions of hydrated electrons, hydrogen and sludge. Water Res. 37, 1685–1690.
atoms and hydroxyl radicals (OH%/O%K) in aqueous solution. J. Phys. Richardson, M.L., Brown, J.M., 1985. The fate of pharmaceutical
Chem. Ref. Data 17, 513–886. chemicals in the aquatic environment. J. Pharm. Pharmacol. 37, 1–12.
Editorial, ., 2000. Drugs in the environment. Chemosphere 40, 691–699. Scott, J.P., Ollis, D.F., 1995. Integration of chemical and biological
Glaze, W.H., Kang, J.W., Chapin, D.H., 1987. The chemistry of water oxidation processes for water treatment: review and recommendations.
treatment processes involving ozone, hydrogen peroxide and ultraviolet Environ. Prog. 14, 88–103.
irradiation. Oz. Sci. Eng. 9, 335–352. Staehelin, J., Hoigné, J., 1982. Decomposition of ozone in water: rate of
Gulyas, H., von Bismarck, R., Hemmerling, L., 1995. Treatment of initiation by hydroxide ions and hydrogen peroxide. Environ. Sci.
industrial wastewaters with ozone/hydrogen peroxide. Water Sci. Technol. 16, 676–681.
Technol. 32, 127–134. Stockinger, H., Kut, O.M., Bourne, J.R., Heinzle, E., 1995. Removal of
Halling-Sorensen, B., 2000. Algal toxicity of antibacterial agents used in chloro and nitro aromatic wastewater pollutants by ozonation and
intensive farming. Chemosphere 40, 731–739. biotreatment. Environ. Sci. Technol. 29, 2016–2022.
Hess, T., Schrader, P.S., 2002. Coupled biotic–abiotic mineralization of Ternes, T.A., Stüber, J., Herrmann, N., McDowell, D., Ried, A.,
2,4,6-trinitrotoluene (TNT). J. Envon. Qual. 31, 736–744. Kampmann, M., Teiser, B., 2003. Ozonation: a tool for removal of
IOA Standardisation Committee—Europe, 001/87-F, 1987. Iodometric pharmaceuticals, contrast media and musk fragrances from waste-
method for the determination of ozone in a process gas, Brussels. water?. Water Res. 37, 1976–1982.
ISO 8192, 1986. Water Quality-Oxygen Demand Inhibition Assay by Vel Leitner, N.K., Papailhou, A.L., Croué, J.P., Peyrot, J., Doré, M., 1994.
Activated Sludge. Oxidation of MTBE and ETBE by ozone and combined ozone/hydro-
Jochimsen, J.C., Jekel, M.R., Hegemann, W., 1997. Combined oxidative gen peroxide. Oz. Sci. Eng. 16, 345–355.
and biological treatment of separated streams of tannery wastewater. Wollenberger, L., Halling-Sorensen, B., Kusk, K.O., 2000. Acute and
Water Sci. Technol. 35, 337–345. chronic toxicity of veterinary antibiotics to Daphnia magna. Chemo-
Kümmerer, K., Steger-Hartmann, T., Meyer, M., 1997. Biodegradability of sphere 40, 723–730.
the anti-tumor agent ifosfamide and its occurrence in hospital effluents Zwiener, C., Frimmel, F.H., 2000. Oxidative treatment of pharmaceuticals
and communal sewage. Water Res. 31 (11), 2705–2710. in water. Water Res. 34 (6), 1881–1885.

You might also like