Oxidation of pharmaceutically active compounds during water treatment

Oxidation of Pharmaceutically Active Compounds During Water
Treatment [Project #4066]
ORDER NUMBER: 4066

DATE AVAILABLE: Fall 2010

PRINCIPAL INVESTIGATORS:
Timothy J. Strathmann, Lanhua Hu, and Heather M. Martin
OBJECTIVES:
The principal objectives of this project were to characterize the oxidation reactions of
representative pharmaceutically active compounds (PhACs) and other wastewater organic
micropollutants with potassium permanganate (KMnO4) and potassium ferrate (K2FeO4)
salts, and to assess the potential use of these reactions to degrade problematic
micropollutants during drinking water treatment operations.
BACKGROUND:
A number of studies in recent years have reported on the ubiquitous occurrence of PhACs
and other classes of wastewater-derived organic micropollutants in drinking water
sources. A variety of PhACs have been detected, including hormones, antibiotics,
anticonvulsants, painkillers, lipid regulators, beta-blockers, antihistamines, and X-ray
contrast media. These micropollutants end up in water bodies principally because of
incomplete removal during conventional wastewater treatment processes.
Although environmental PhAC concentrations are low (nanogram to microgram per liter),
there remains serious public concerns about potential health effects associated with long-
term exposure to these chemicals in drinking water, especially considering the unknown
effects of complex pharmaceutical mixtures. Concern also exists about wastewater
discharges of some PhACs promoting increased bacterial antibiotic resistance and others
causing reproductive problems in exposed aquatic organisms. Because of these concerns,
it is important that researchers improve understanding of the fate and degradation of
PhACs in both natural aquatic environments and during conventional treatment processes,
and at the same time work to develop new technologies to more effectively remove
problematic micropollutants from drinking water and wastewater.
Chemical oxidation processes may be the most important drinking water treatment
method for removal of PhACs. Studies have investigated the oxidation of PhACs by
common drinking water disinfectants, including free and combined chlorine, ozone, and
chlorine dioxide. However, there remains a gap in the understanding of PhAC oxidation
and fate at many treatment facilities that employ potassium permanganate to control taste
and odor, remove color, control biological growth, and remove iron and manganese. This
project aims to fill this critical research gap and to evaluate the use of an emerging water
treatment oxidant, potassium ferrate, for treatment of PhACs in comparison to
permanganate.
APPROACH:
A laboratory study was performed to characterize the reactions between permanganate/
ferrate and a representative group of 18 PhACs (acetaminophen, atenolol, bezafibrate,
bisphenol A, caffeine, carbamazepine, chlortetracycline, ciprofloxacin, diclofenac, 17α-
ethinylestradiol, ibuprofen, iopromide, lincomycin, N-nitrosodimethylamine,
sulfamethizole, sulfamethoxazole, triclosan, and trimethoprim). The project team
examined the reactions in batch systems containing both laboratory solutions with well-
defined composition and natural source waters obtained from participating utilities. A
series of research tasks were performed to meet the project objectives.
1. Screening survey experiments were conducted to identify classes of PhACs most susceptible to permanganate and ferrate treatment processes and to assess the potential for adsorption of PhACs to MnO2(s) and Fe(OH)3(s) reaction products. 2. Kinetic parameters for permanganate and ferrate reactions with selected PhACs were measured in laboratory solutions with well-defined composition. 3. Predictive models were developed to account for the effects of varying water quality conditions (e.g., pH) on reaction rates. 4. PhAC oxidation byproducts formed during permanganate and ferrate treatment processes were identified using liquid chromatography-mass spectrometry techniques, and the effects of permanganate treatment on pharmaceutical potency of selected antibiotics were assessed using bacterial growth inhibition bioassays. Mechanisms for oxidation of individual PhACs were proposed based upon the identified reaction products. 5. The validity of the predictive models for permanganate reactions during PhAC treatment were tested in natural source waters provided by six participating utilities.
RESULTS/CONCLUSIONS:
Results of initial survey experiments indicated that permanganate and ferrate are selective
oxidants that can be expected to oxidize only a fraction of PhACs present in source
waters, and PhACs exhibited similar reactivity patterns with both oxidants. Of the 18
PhACs surveyed with permanganate, only 10 exhibited high or moderate reactivity, while
only 8 out of 12 compounds surveyed showed high or moderate reactivity with ferrate.
None of the surveyed compounds adsorbed to the MnO2(s) and Fe(OH)3(s) reaction
products appreciably, indicating that oxidation is the major mechanism for removal occurring during the experiments. Results from the survey experiments indicated that permanganate and ferrate are highly reactive with PhACs that possess olefin (carbamazepine) and phenolic (acetaminophen, bisphenol A, chlortetracycline, 17α-ethinylestradiol, and triclosan) moieties in their structures. PhACs exhibiting moderate reactivity with permanganate and/or ferrate possess secondary/tertiary aliphatic or aromatic amine groups in their structures (atenolol, ciprofloxacin, diclofenac, lincomycin, sulfamethoxazole, and trimethoprim). Kinetic studies performed in laboratory solutions prepared with well-defined composition (oxidant concentration, pH, temperature) demonstrate that PhAC reactions with permanganate and ferrate can be described by second-order rate laws (first order in both oxidant and PhAC concentration). The temperature dependence of PhAC reactions with permanganate follows the Arrhenius equation. Reaction rates are significantly affected by the solution pH, and the wide variations can be accounted for using kinetic models that consider the changing speciation of both the PhAC and the reacting oxidant. Model formulation assumes that individual oxidant species react with individual PhAC species with characteristic second-order rate constants. Tests show that permanganate and ferrate oxidize the target PhACs incompletely, leading to little or no organic carbon mineralization. A large number of reaction products were identified for five PhACs (carbamazepine, ciprofloxacin, lincomysin, trimethoprim, and 17α-ethinylestradiol) reacting with permanganate (and ferrate in the case of carbamazepine). Findings support the proposed reaction centers being olefin, phenolic, and amino groups. PhAC oxidation yields a variety of oxygen insertion products (alcohol, aldehyde, ketone, carboxylic acid) and ring cleavage products. Major oxidation pathways are proposed for each of the PhACs examined. Antibiotic growth inhibition bioassays performed with ciprofloxacin and trimethoprim demonstrate that permanganate treatment yields products that possess negligible antibiotic potency in relation to the parent antibiotics. This finding is crucial because it demonstrates that incomplete oxidation of PhACs will be sufficient to eliminate the biological activity of trace micropollutants present in drinking water sources. Additional research is needed to verify that the findings for ciprofloxacin and trimethoprim extend to a wider range of PhACs and target bioassays. In general, kinetic models for PhAC reactions with permanganate, which were developed and parameterized in laboratory solutions using high reagent concentrations (to simplify experiments), accurately predict the extent of PhAC oxidation in utility source waters treated with low dosages of permanganate that are more relevant to drinking water utilities (e.g., < 4 mg/L KMnO4). In some utility source waters, much greater oxidation of selected PhACs is observed than is predicted and is attributed to more complex auto-catalytic processes that occur when the MnO2(s) product forms. Thus, the predictive models can be seen as providing a conservative estimate for the amount of PhAC
oxidation expected during permanganate treatment, with auto-catalytic processes serving
to provide an additional level of treatment beyond that predicted.
APPLICATIONS/RECOMMENDATIONS:
The results obtained in this study demonstrate that permanganate and ferrate can be used
to selectively oxidize certain PhACs of concern to water utilities. Findings also improve
our comprehensive understanding of PhAC fate during water treatment operations at
facilities that already employ permanganate processes for other purposes. PhACs
containing phenolic, olefin, secondary and tertiary amine, and heterocyclic N structures
are expected to be oxidized to some degree during existing permanganate treatment
processes. Among PhACs commonly detected in drinking water sources, phenolic
endocrine disrupter compounds like 17α-ethinylestradiol and bisphenol A are believed to
pose the greatest risk at environmental concentration levels (ng/L – μg/L). Thus, their
high reactivity with permanganate suggests that facilities that employ this oxidant for
other purposes are also gaining an additional barrier to public health exposure. Also, it is
worth noting that carbamazepine, which is among the most widely detected PhACs, but is
resistant to oxidation by chlorine disinfectants, is rapidly oxidized by permanganate and
ferrate. Thus, in plants that use permanganate in addition to free or combined chlorine
disinfectants, permanganate oxidation is expected to be the major removal mechanism.
To help utilities predict the level of PhAC oxidation in their facility’s existing
permanganate treatment process, the project team developed a simple to use Microsoft
Excel file that incorporates the kinetic models for selected PhACs (see MULTIMEDIA
below). Utilities can also use the model in an iterative fashion to determine the
appropriate permanganate dosage and reactor residence time needed to oxidize a PhAC of
concern within their individual facility.
Although permanganate is not often used at wastewater treatment facilities, results from
this study suggest that it may be an effective option for selectively oxidizing reactive
PhACs in these facilities. More non-selective oxidants like those generated by advanced
oxidation processes will be inefficient within such organic-rich matrices. Permanganate,
which is a more selective oxidant, will persist for longer times enabling a significant
degree of oxidation for reactive PhACs. Treatment of PhACs at the source may be a
better overall strategy for addressing PhACs than discharging the compounds to receiving
waters that are also used as drinking water sources. Treatment of potent estrogenic
hormones and antibiotics with permanganate or other oxidants at the wastewater
treatment plant prior to discharge may also be warranted from an ecological viewpoint.
Ferrate salts are not currently used in field scale drinking water or wastewater treatment
operations, although its use in a variety of treatment processes is gaining increasing
attention among water quality researchers. Results presented here support the further
development of ferrate for water treatment applications. The project team recommends
further focus on developing on-site technologies for ferrate production to overcome
practical issues associated with transport/storage and long-term stability of ferrate
solutions.

MULTIMEDIA:
The CD-ROM with this report includes a Microsoft Excel file that can be used to predict
the removal of selected pharmaceutically active compounds in utility source waters.
Required user input includes the PhAC of interest, source water temperature and pH, and
either the dosed KMnO4 concentration and reaction time of interest or the measured
KMnO4 exposure (integrated concentration  time) in the source water.

PARTICIPANTS:
Six utilities participated in this project: City of Bloomington (Ill.) Water Department;
City of Decatur (Ill.) Water Department; Louisville (Ky.) Water Company; Passaic Valley
(N.J.) Water Commission; Southern Nevada Water Authority, and City of St. Louis (Mo.)
Water Division.

Source: http://www.waterrf.org/ExecutiveSummaryLibrary/4066_ExecutiveSummary.pdf