Application of manganese oxides under anoxic conditions to remove diclofenac from water

Accepted Manuscript Title: Application of manganese oxides under anoxic conditions to remove diclofenac from water Authors: Wenbo Liu, Alette A.M. Langenhoff, Nora B. Sutton, Huub H.M. Rijnaarts PII: DOI: Reference:

S2213-3437(18)30257-4 https://doi.org/10.1016/j.jece.2018.05.011 JECE 2374

To appear in: Received date: Revised date: Accepted date:

27-11-2017 16-4-2018 5-5-2018

Please cite this article as: Wenbo Liu, Alette A.M.Langenhoff, Nora B.Sutton, Huub H.M.Rijnaarts, Application of manganese oxides under anoxic conditions to remove diclofenac from water, Journal of Environmental Chemical Engineering https://doi.org/10.1016/j.jece.2018.05.011 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Application of manganese oxides under anoxic conditions to remove diclofenac from water

IP T

Wenbo Liu, Alette A M Langenhoff *, Nora B Sutton, and Huub H M Rijnaarts

Sub-department of Environmental Technology, Wageningen University & Research, 6708 WG,

SC R

Wageningen, the Netherlands

U

*Corresponding author.

N

Alette A M Langenhoff

Sub-department of Environmental Technology, Wageningen University & Research, 6708 WG,

A

Wageningen, the Netherlands

Other authors Wenbo Liu

ED

M

E-mail address: [email protected]; Tel.: +31 317 483339

Sub-department of Environmental Technology, Wageningen University & Research, 6708 WG,

PT

Wageningen, the Netherlands

CC E

E-mail address: [email protected]; Tel.: +31 317 483339

Nora B Sutton

Sub-department of Environmental Technology, Wageningen University & Research, 6708 WG,

A

Wageningen, the Netherlands E-mail address: [email protected]; Tel.: +31 317 483339

Huub H M Rijnaarts Sub-department of Environmental Technology, Wageningen University & Research, 6708 WG, Wageningen, the Netherlands E-mail address: [email protected]; Tel.: +31 317 483339 1

N

U

SC R

IP T

Graphic Abstract

A

Highlights

MnO2 successfully removes diclofenac under anoxic conditions



Diclofenac removal is stable at 10 – 30℃ and inhibited at 40℃



MnO2:diclofenac ratio, metal ions, PO43-, and humic acids affect diclofenac removal



Optimal application conditions identified for this promising new technology

CC E

PT

ED

M



A

Abstract

This study focuses on the potential of applying manganese oxides (MnO 2) under anoxic conditions

(absence of oxygen) to remove diclofenac (DFC). By investigating parameters that are important for application, including temperature, MnO2: DFC molar ratio, and co-solutes, the DFC removal potential is evaluated in terms of efficiency and observed initial kinetics (kobs, init). Four commonly-used kinetic

2

models are compared in this study and the best fitting one is employed. Overall, DFC removal and kobs,

init

both increase upon changing temperature from 10 to 30ºC and both decrease after further increasing

temperature to 40ºC. Increasing the MnO2: DFC molar ratio improves degradation, as this provides

more reactive surface sites for DFC conversions. However, DFC removal does not further increase

IP T

when increasing the MnO2:DFC from 2200:1 to 8900:1. The presence of metal ions inhibits DFC removal, possibly because the ions adsorb onto the reactive sites at the MnO2 surface and compete

SC R

with DFC. Phosphate has a diverse effect on DFC degradation: low concentrations inhibit and high concentrations promote removal. The presence of humic acids significantly promotes diclofenac

U

removal. These findings are a first step towards further developing pharmaceutical removal

M

A

N

technology using MnO2 under anoxic conditions.

Keywords: manganese oxides; anoxic abiotic pharmaceutical removal; temperature; MnO 2:diclofenac

A

CC E

PT

ED

ratio; co-solutes; application

3

1.

Introduction

Diclofenac ([2-(2,6-Dichloroanilino)phenyl]acetic acid, DFC) is a commonly used pharmaceutical known to have the highest acute toxicity to wildlife among the non-steroidal anti-inflammatory drugs

IP T

[1, 2]. In Europe, mean DFC consumption is around 1300 μgcapita-1d-1 [3]. After consumption, DFC

SC R

will enter the municipal wastewater via urine and feces. In conventional wastewater treatment plants

(WWTPs) like activated sludge processes, DFC removal efficiency is 40% - 50% [4]. The incomplete

N

eventually drinking water in the range of 0 – 1000 ng·L-1 [5].

U

removal of DFC in WWTPs results in release of this compound to surface water, groundwater, and

A

DFC can damage renal and gastrointestinal tissues of vertebrates, leading to death of these

M

organisms [2]. Usually, the effective concentration for DFC toxicity is 100 - 300 μg·L-1 [2, 6] , which is much higher than the concentration detected in the aquatic environment. However, a recent study

ED

shows that DFC can cause tissue damage at concentrations as low as 250 ng·L-1 [2], which is within the

PT

environmentally-relevant range . Due to toxic effects of DFC to the environment and ecosystem, the European Union (EU) has added DFC into the "Watchlist"[7]. The European Community suggests for

CC E

the environmental quality threshold for good water quality a value of 100 ngL-1 (annual average) [8]. Various technologies, such as advanced oxidation processes, ozonation, sorption to activated carbon or

A

algae treatment systems, can remove DFC, but these technologies often either produce toxic intermediates or are too costly [9, 10].

Due to their high removal efficiencies of recalcitrant compounds, sustainability, and lower construction and operation costs, water treatment technologies under anoxic conditions (absence of

4

oxygen) are becoming more and more popular in removal of pharmaceuticals like DFC [11]. In addition, described pharmaceutical removal processes can be more attractive under anoxic conditions than under oxic conditions. For example, using manganese oxide (MnO 2) to remove pharmaceuticals can achieve high removal efficiencies and may produce fewer toxic intermediates [12, 13]. The

IP T

pharmaceutical removal process with MnO2 may be the similar under oxic and anoxic conditions, namely the compounds will be adsorbed onto the MnO2 surface first and then be oxidized. However,

SC R

the specific removal mechanisms could be different, leading to higher DFC removal efficiency under

anoxic conditions than under oxic conditions. Our previous study showed that DFC removal improves

U

from 59% under oxic conditions to 78% under anoxic conditions, indicating that anoxic conditions are

N

favourable for DFC removal with MnO2 [14]. The results also show that the adsorption has no

A

contribution to the final DFC removal, even though it is an important mechanistic step [14]. In other

M

studies, anoxic conditions either inhibit or have no effects on removal with MnO 2 [15, 16]. This unique

ED

finding opens the way for development of a new cost-efficient and sustainable pharmaceutical removal process based on applying MnO2 under anoxic conditions, which has been insufficiently studied. In

PT

addition, our study shows that the properties of pharmaceuticals and MnO 2 might have greater

CC E

influence on the pharmaceutical removal under anoxic conditions than under oxic conditions. Furthermore, it is still unclear how anoxic conditions will affect the removal mechanisms. Thus, the

A

effects of operational parameters on the DFC removal are unpredictable under anoxic conditions based on previous studies under oxic conditions. To develop anoxic DFC removal with MnO2 into an applicable technology, the operational parameters should be studied first. Here we focus on temperature, the MnO2: DFC molar ratio, and the presence of co-solutes (metal ions, phosphate, humic acids). 5

Reactions between MnO2 and organic compounds are endothermic, and therefore increasing temperature can accelerate removal rates [13, 17]. However, observed initial rate constant of oxytetracycline removal with MnO2 under oxic conditions is reported to be lower at 40ºC than at 5 to 30ºC [17]. These contradicting results to theory and the lack of investigations under anoxic conditions

IP T

indicates the necessity of further investigating temperature effects. From an application perspective, the results will partially determine the applicability of MnO2 technologies in different seasons and

SC R

climates.

Another important parameter in the application is the molar ratio between MnO 2 and the

U

pharmaceutical to be degraded. Previous studies show that increasing the MnO 2:chlortetracycline from

N

4:1 to 32:1 leads to faster removal [13]. However, no significant change is observed in triclosan,

M

A

carbadox, and ciprofloxacin removal with increase of MnO 2:pharmaceutical ratio from 10:1 to 300:1 [18]. In addition, the ratios studied previously are usually low while they can be much higher in

ED

application due to the low concentration of pharmaceuticals. From an application perspective,

PT

optimizing MnO2: DFC molar ratio can decrease operation costs, by minimizing MnO 2 addition to maximize DFC removal.

CC E

Co-solutes in water including metal ions, anions (like phosphate), and organic compounds are

commonly found in water contaminated with pharmaceuticals. The EU allows 3.57μM iron and

A

0.91μM manganese in drinking water [19], indicating that Mn and Fe could be present as a co-solutes in water treatment. In urban wastewater effluent, the phosphate concentration should be < 2 mg PL-1 while the organic compounds (expressed as Chemical Oxidation Demand, COD) are allowed at 125 mg O2L-1 based on EU regulation [20]. All these co-solutes can affect pharmaceutical removal with MnO2 in WWTP effluents under oxic conditions. Previous studies under oxic conditions showed that 6

pharmaceutical removal could be inhibited by metal ions and phosphate (PO43-) [13, 21] while humic acid representing organic matters can both promote and inhibit pharmaceutical degradation

[21, 22].

However, none of these parameters have been investigated under anoxic conditions. In this study, four metal ions (Mn2+, Ca2+, Mg2+, and Fe3+), PO43-, and different concentrations of humic acid are selected

IP T

to investigate the effects of co-solutes on the application of MnO2 under anoxic to remove DFC.

SC R

Here, we investigate the efficiencies and reaction kinetics of DFC removal from water by MnO2 under anoxic conditions. The effects of temperature, MnO2: DFC molar ratio, and co-solutes are

investigated to evaluate the application potential and optimal conditions. Removal mechanisms under

U

anoxic conditions are affected by the pharmaceutical chemical structure and properties of both

N

pharmaceuticals and MnO2 [14, 23]. The results provide the basis for developing a new treatment

2.1 Chemical reagents

ED

2. Materials and methods

M

A

technology for removal of trace organic contaminants from WWTP effluent.

PT

Diclofenac sodium salt (DFC) was purchased from Sigma-Aldrich (Table 1). Other chemical

CC E

reagents used in the experiments were purchased from either Sigma-Aldrich or Merck KGaA (Germany). All the chemicals had > 97% purity. The liquid used in these experiments was at either high

A

performance liquid chromatographic grade or ultra-performance liquid chromatographic grade. Millipore water purification system was used to prepare ultrapure water (18.3 Mcm, TOC=18 ppb). Anoxic water was prepared by boiling demineralized water (demiwater) for 5 min and cooling under N2-flow to the room temperature. All solutions were prepared with anoxic water unless specified otherwise. DFC stock solution was prepared in ultrapure water to reach a final concentration of 157 7

µM in a 50-mL amber glass bottle. This stock was stored at -20ºC to avoid potential decomposition. Stock solution (10 mM) of CaCl2, MgCl2, MnCl2, and FeCl3 were prepared with anoxic water. The solution of PO43- was prepared by mixing the same amount of Na2HPO4 and NaH2PO4 (c/c =1:1). Humic acid (HA) was purchased from Sigma-Aldrich (product No. 53680). HA stock solution was

U

SC R

stock was stored in a 50-mL glass bottle at 4ºC to avoid decomposition.

IP T

prepared at a final concentration of 1000 mg·L-1 in anoxic water with 1 – 3 drops of 1M NaOH. This

N

2.2 Synthesis of manganese oxides

A

Manganese oxides (MnO2) were freshly synthesized by a modified method as previously

M

described [14]. In brief, equal volumes of 0.4 M MnCl2 and 0.4 M KMnO4 were mixed and adjusted to

ED

pH 10 with 1M NaOH. The MnO2 was washed six times with anoxic demiwater via centrifugation (15 min at 5000 rpm) and decant. All the experiments in this study used freshly prepared MnO2. To keep

PT

MnO2 fresh, it was stored at 4 ºC as a suspension with N2 headspace and used within one week after

CC E

preparation. The MnO2 prepared is exactly the same used in previous studies, which indicate MnO 2 is amorphous with a surface area of ~180m2/g [14, 25].

A

2.3 Experimental setup

Batch experiments were prepared under anoxic conditions in 125-mL glass bottles. All bottles

were sealed with rubber stoppers and aluminum crimp-caps, settled at 30ºC without shaking, and covered with aluminum foil to avoid photodecomposition. The pH was maintained at 7 for all experiments with a 10 mM MOPS (4-morpholinepropanesulfonic acid) buffer solution. Sodium 8

chloride (NaCl) was added to adjust the ionic strength (I=0.01M). The final concentration of MnO2 in the bottle was 7 mM. The 50-ml buffer containing NaCl was added into the experimental bottles. The headspace of all bottles was exchanged with N2 to keep the system anoxic.

The experiments were started by adding 0.25 – 1 mL DFC stock solution (157 µM) into the

IP T

experimental bottles. Based on our previous study, the pharmaceutical removal with MnO 2 seems to

SC R

reach a plateau concentration after 24 hour reaction [14]. To confirm that, the experiments in this study last for 33 hours. All the experiments were carried out in triplicate and lasted for 33 hours in most cases. Only the experiments on effects of phosphate were carried out in duplicate. 1.2 mL mixture

U

samples were taken regularly during the experimental period. The samples were centrifuged

N

immediately at 10000 rpm for 10 min. The supernatant was collected and stored at -20ºC until analysis.

M

A

Sorption of DFC onto MnO2 were not taken into account because it is insignificant in removal based on previous study [14, 26], but it is still an important step in removal mechanism.

ED

Control experiments without MnO2 were prepared and conducted simultaneously with each batch

PT

of reactions. The effect of temperature was investigated by conducting the experiments at 10, 20, 30, or 40ºC. The effect of MnO2: DFC molar ratio was investigated from 480:1 to 8900:1 at 20ºC. The effect

CC E

of co-solutes was investigated at 30ºC with different concentrations of metal ions (Mn2+, Ca2+, Mg2+, Fe3+ at 0.1 and 1 mM), the anion (PO43- at 0.5, 1, and 116 mgPL-1), and dissolved organic compound

A

(HA at 5, 10, 15, and 25 mgL-1). In the experiments with HA, DFC was added before adding HA. Samples were taken before and after adding HA at the beginning of the experiments.

2.4 Analysis

DFC was measured by ultra-performance liquid chromatography with a diode array detector 9

(Ultimate 3000, Thermo co. Ltd.). 10 μL of samples were automatically injected into a CSH phenylHexyl column (1.7μm, 130 Å, 2.1 × 150 mm, Waters co., USA).

The detection limits of this analysis

is 50 μg·L-1. Before analysis, fenoprofen was added at a final concentration of 5 mgL-1 as internal standard. This allows us to identify and correct potential analytical errors. The details of analytical

IP T

methods and setting of the machine were described previously [14, 27].

SC R

Phosphate concentration was analyzed by colorimetric method using cuvettes from Hach Lange (LCK 349). All metal ions were analyzed by an inductively coupled plasma atomic emission

spectroscopy (ICP-OES, Perkin Elmer Instruments, USA). pH was determined by a pH meter

N

U

(PHM210, MeterLab, Radiometer Analytical).

A

2.5 Kinetics study

M

In literature, different models have been used to describe organic compound removal with MnO 2:

ED

pseudo-first-order model, pseudo-first-order model with data from the first several hours of the experiments, pseudo-second-order model, and mechanism-based kinetics models [13, 17, 18]. The

PT

pseudo-first-order kinetic model has been used to evaluate the reaction kinetics of micropollutants and

CC E

other organic compounds with MnO2 [28]. However, in experimental results deviation from this model after a certain reaction time was also reported. To better predict long-term kinetics of organic compound removal with MnO2, the pseudo-second-order model was used [13]. The pseudo-second-

A

order model usually gave a better fit during the long period. Similarly, the mechanism-based model was proposed to fit data obtained in long-term experiments. For example, Rubert and Pedersen [17] developed the mechanism-based kinetics model to better describe oxytetracycline removal with MnO2. Similarly, Zhang, Chen and Huang [18] described another kinetic model based on the removal

10

mechanism to describe antibacterial agent removal with MnO2. In this study, all four models were validated with the results obtained at 30ºC with 7 mM MnO2, 3.14 µM DFC, and no co-solutes, namely (1) pseudo-first-order kinetic model; (2) pseudo-first-order kinetic model with first 5 hour data; (3) pseudo-second-order model; and (4) mechanism-based kinetic model from Zhang, Chen and Huang

IP T

[18] (Section S1).

SC R

3. Results and discussion

3.1 Kinetic Results

U

Our results show that diclofenac (DFC) can be removed by applying MnO 2 under anoxic

N

conditions. The pseudo-first-order model fits the data poorly (R2=0.84) while other kinetic models fit

A

the data better and are useful for evaluating our kinetic studies (R2=0.96 – 0.99) (Figure S1, Table S1).

M

The results show that anoxic MnO2 mediated DFC removal follows a two-stage kinetic process under

ED

all experimental conditions. Most DFC conversion is obtained in the first 9 hours (following the pseudo-first-order reaction) and only about 10% DFC removal occurs in the subsequent 24 hours (from

PT

t=9 h to t=33 h, following the pseudo-zero-order reaction). A pseudo-first-order model was used to

CC E

calculate the observed removal rate constants (kobs,init) (Table 2, section S1). The data from the first 5 hours were used for the model calculation, as the first 5 hours showed linear removal. Without MnO2,

A

no DFC removal is observed (Table S2, Figure S2).

The kobs,init of these experiments show that DFC removal with MnO 2 under anoxic conditions depends highly on the conditions, which is similar to studies under oxic conditions [28-30].

11

Based on

the results, the first 5 hour data are linear and fit the pseudo-first-order model, indicating that the removal mainly occurred at the beginning of the reaction. This is also observed in previous studies on removal of other pharmaceuticals like chlortetracycline [13] and oxytetracycline [17]. The reason for the fast initial removal may be that at the beginning of the experiment, the MnO2 is fresh and thus not

IP T

occupied by either DFC or its transformation products. As a result, there are abundant reactive surface

SC R

sites for adsorption and oxidation.

3.2 Effect of temperature

Results show that the application of MnO2 to remove DFC under anoxic conditions is generally an

N

U

endothermic process (Table 2, Figure 1). When increasing the temperature from 10 to 30ºC, DFC

A

removal increases from 85 ± 0.4% to 90 ± 0.7%, and the kobs,init increases from 0.137 ± 0.007 to 0.213

M

± 0.006 h-1,. However, when the temperature is further increased to 40ºC, both removal and kobs, init drop dramatically to respectively 38 ± 1.5% and 0.041 ± 0.001 h -1 (Figure 1, Table 2). When the temperature

remained low.

PT

init

ED

is subsequently reduced from 40ºC to 20ºC, only a slight increase in total removal is observed and kobs,

CC E

At 40 ºC, MnO2 particle aggregation is observed (Figure S3(a-d), Figure S4). To investigate whether this aggregation plays a role in the unexpected drop of DFC removal at 40ºC, an additional experiment was performed (Figure 1, dashed line). The experiment was started at 40ºC, allowing

A

aggregates to form. Then the batch bottles were transferred to 20ºC at t=9 h. The temperature of these bottles dropped to 20 ºC within 1 hour. The results show that the MnO2 aggregation remains after the temperature drop, indicating that the aggregation process is irreversible. In addition, only 22 ± 1.2% DFC is removed from t=9 h to t= 33 h, which is lower than the 49 ± 1.3% DFC removal in this time

12

period if the experiment is started at 20ºC. The results indicate that the MnO 2 aggregation is an important reason for the decrease in DFC removal at 40ºC.

IP T

Increasing temperature will lead to increase of DFC removal with MnO 2 under anoxic conditions. These results indicated that the reaction between DFC and MnO2 under anoxic conditions is

SC R

endothermic, similar to that under oxic conditions [13, 21]. Also, the lower removal at high temperature is also observed under oxic conditions. For example, Rubert and Pedersen [17] find that

U

oxytetracycline removal with MnO2 under oxic conditions is lower at 40ºC than at 5 to 30ºC. In our

N

experiments, aggregation of MnO2 particles is observed at 40ºC which can decrease the amount and

A

availability of reactive surface sites on MnO2 for DFC removal. This may be a reason for the reduced

M

DFC removal at high temperature. The aggregation may be caused by two processes: Ostwald ripening

ED

and/or aging. Ostwald ripening, also called coarsening or competitive growth, occurs in the suspension where large particles grow at the cost of small particles [31]. Amorphous particles (or small crystals)

PT

dissolve and re-precipitate on the largest crystals initially present. As a result, the amorphous particles

CC E

or small crystals grow to relatively large crystals [32]. Ostwald ripening is promoted by elevated temperature, as demonstrated in previous studies [33, 34]. The aging process is the physical change of

A

minerals over time. Aggregation of MnO2 is observed to be a fast process at 40ºC (hours – days), and a slow process at 20ºC (6 months, the duration of our experiments, Figure S3 (e, f)). This indicates that MnO2 aging process is accelerated when changing the temperature to 40ºC, which quickly changes MnO2 morphology irreversibly [35]. As a result, active MnO2 precipitates into a more stable form, leading to less DFC removal [36], and the reactivity cannot be restored when changing again to lower 13

temperatures, i.e. 20ºC (Figure 1, dashed line). Overall, the aggregation of MnO 2 particles leads to less DFC removal.

3.3 Effect of MnO2: DFC molar ratio

IP T

The MnO2: DFC molar ratio determines the extent of DFC removal under anoxic conditions. Results show that a higher MnO2: DFC molar ratio leads to higher and faster removal (Table 2, Figure

SC R

2). DFC removal increases from 52 ± 5.3% at MnO 2: DFC molar ratio of 480:1 to 87 ± 2.3% at MnO2: DFC molar ratio of 2200:1, while the kobs,init increases from 0.046 ± 0.004 to 0.197 ± 0.006 h-1. Further increasing the MnO2: DFC molar ratio from 2200:1 to 8900:1, both the removal efficiency and kobs,init

M

A

N

U

decrease and the kobs,init values levels off to about 0.147 ± 0.031 h-1.

When the MnO2 particles are applied in a reactor for treating WWTP effluent, the amount of

ED

MnO2 is in the mM range while the concentrations of DFC and other pharmaceuticals are extremely

PT

low, in the nM to μM range [12, 37]. Therefore, high MnO2: DFC molar ratio were used in this study to mimic those reactor conditions. Results show that increasing the MnO 2: DFC molar ratio improves

CC E

pharmaceutical removal with MnO2. Increasing the amount of MnO2 provides relatively more reactive surface sites. As a result, DFC can better attach to the MnO2 surface, the first step in its removal. After

A

adsorption, the DFC is removed via chemical oxidation. During this process, increasing surface reactive sites will not improve the oxidation of DFC. In addition, due to the properties of pharmaceuticals and MnO2, adsorption will not be improved by simply increasing the reactive surfaces sites through increasing MnO2: DFC molar ratio. On the contrary, both the removal and kobs,init decrease

14

when the MnO2: DFC molar ratio increases from 2200:1 to 8900:1. Similar results are reported previously in the study on triclosan removal with MnO 2 under oxic conditions [18].

A possible explanation is self-inhibition by DFC reaction products and Mn2+ during DFC removal. Previous research shows that the accessibility of MnO2 surface sites may decrease, causing to be less reactive [17]. That is because pharmaceuticals like DFC can occupy the reactive

IP T

the surface

SC R

sites, and their intermediates and Mn2+ from the reaction can also reduce reactivity. Higher MnO2: DFC molar ratios will lead to faster reactions between MnO2 and DFC at the beginning. As a result, the accessibility of MnO2 surface sites decrease faster and the reaction rate decreases as well. In addition,

U

large amount of MnO2 could maintain the relative proportions of Mn2+ [18],

A

N

pharmaceutical removal with MnO2.

which strongly inhibits

M

3.4 Effect of metal ions

ED

Under anoxic conditions, inhibition of DFC removal is observed in the presence of four metal ions at two different concentrations (Table 2, Figure 3). In the absence of metal ions, 90 ± 0.7% of the

PT

DFC is removed after 33 hours. In the presence of 0.1 mM Mn 2+, Ca2+, Mg2+, or Fe3+, only 80 – 84% of

CC E

DFC is removed. Results also show that inhibition effects in the kinetics follows the order of “metal free” < Ca2+ < Mg2+ ≈ Fe3+ < Mn2+. Ca2+ has the least inhibition on DFC removal, with a decrease in kobs,init from 0.213 ± 0.006 to 0.126 ± 0.013 h-1. Increasing the concentration of metal ions to 1 mM

A

results in further inhibition in the order of “metal free” < Fe 3+ < Ca2+ ≈ Mg2+ < Mn2+. Only 58 ± 0.4% of DFC is removed in the presence of Mn2+, followed by 77 ± 1.4% with Mg2+ and 74 ± 1.0% with Ca2+. The least inhibitory effect is observed in the presence of Fe 3+ with about 88 ± 1.3% of DFC removed, which is close to the removal (90%) in the absence of metals. The inhibition in the kinetics

15

follows the same order as the removal efficiency, namely “metal free” < Fe3+ < Ca2+ ≈ Mg2+ < Mn2+ (Table 2).

IP T

The reaction between MnO2 and DFC is closely related to the MnO2 surface as described previously [14]. Similar to other surface reactions, the interaction between metal ions and MnO 2

SC R

surface is an important factor affecting the removal of DFC [38]. The inhibitory effect of metal ions is most likely due to competition by ions with DFC for reactive surface sites on MnO 2. Since the pH

U

characterizing the point-of-zero-charge of MnO2 (birnessite) is 0.97 [39], MnO2 is negatively charged

N

at neutral pH [40]. Similarly, DFC is also negatively charged (pKa=4.15) under the experimental

A

conditions [41]. Therefore, the positively charged metal ions are more easily adsorbed onto MnO 2

M

surface via electrostatic interactions than DFC [26, 42]. This interaction forms a complex, resulting in

ED

fewer reactive surface sites available for DFC removal. The adsorption of metal ions onto MnO 2 is related to the radius of hydrated ions [43, 44]. At low concentration (0.1 mM), abundant reactive sites

PT

for DFC removal are present, even though some reactive surface sites are used by metal ions.

CC E

Therefore, there is little difference between metal ions in their inhibition of DFC removal. However, at a higher concentration (1 mM), the inhibition effect of metal ions follows the order of Mn 2+ > Ca2+ ≈ Mg2+ > Fe3+. This order is roughly the same as the adsorption affinity described previously [45]. The

A

lowest inhibition observed for Fe3+ is probably because the radius of Fe3+ (0.064 nm) is the smallest among the four metal ions (Mn2+ : 0.080 nm, Ca2+ : 0.103 nm, Mg2+ : 0.070 nm). It is reported that the smaller the ionic radius the lower the adsorption affinity [46]. In addition, even though Fe3+ was added

16

as ion into the bottle, some of them will precipitate at pH 7 in the experiments, which also decreases inhibition by Fe.

3.5 Effect of phosphate

IP T

Phosphate (PO43-) is a common pollutant in wastewater which affects removal of micropollutants and other organic compounds with MnO2 [13, 47, 48]. Of the studies carried out under oxic conditions,

SC R

a limited number of studies have looked into the effects of PO 43-, and no studies are reported for anoxic conditions [13, 47]. Therefore, we investigate the effects of phosphate (PO43-) on the removal of DFC with MnO2 under anoxic conditions (Table 3). At low PO43- concentration (< 1 mg PL-1 ), DFC

N

U

removal decreases from 76% at 0 mg P·L-1 to 46% at 0.5 mg PL-1. When the PO43- concentration

A

further increases to 1 mg PL-1, DFC removal increases slightly to 54%. Complete DFC removal is

ED

M

obtained at 116 mg PL-1.

PT

Similar to the metal ions, effects of PO43- on DFC removal with MnO2 are also related to the MnO2 surface reaction [38]. PO43- can adsorb onto the MnO2 surface, which decreases the reactive

CC E

surface sites for DFC removal [49, 50]. On the other hand, PO43- can decrease the inhibition by Mn2+ generated from DFC removal by forming Mn3(PO4)2 [51]. This mitigates inhibition of DFC removal by

A

Mn2+, leading to higher conversion efficiencies at higher PO43- levels. In addition, the structure of Mn3(PO4)2 can help in Mn(III) stabilization. As Mn(III) can improve the oxidizing powder of Mn(IV) [18], stabilization of Mn(III) by PO43- also results in enhanced DFC removal. When the PO43concentration is low, inhibition by PO43 on DFC removal dominates, resulting in lower DFC removal efficiencies. However, further increasing PO43- concentrations decrease inhibition by Mn2+,and even 17

stimulation due to Mn(III) stabilization occurs, resulting in more DFC removal. Even though the tested high PO43- concentrations are rarely observed in wastewater effluent, the PO 43- concentration in buffer can be as high as 310 mg PL-1 in lab-scale experiments using MnO2 [26, 52]. Based on this study, the effects of PO43- buffer such lab-scale studies should be considered.

IP T

3.6 Effect of humic acid

SC R

Humic acid (HA), chosen as a representative of organic matter, significantly promotes DFC

removal under anoxic conditions (Table 4). Based on the fast DFC removal observed in the presence of humic acid (HA) in a pre-test (data are not shown), initial DFC concentration was increased to 5.57

N

U

µM. Even when HA concentration is only 5 mg·L-1, 80 ± 0.4% DFC is removed within 1 hour. When

A

HA concentration increases from 10 mg·L-1 to 20 mg·L-1, DFC removal increases from 87 ± 0.3% to

ED

M

96 ± 0.1% within 1 hour.

PT

In this study, DFC removal is promoted in the presence of HA. Under oxic conditions, it is reported that HA can form a complex with Mn2+ produced from MnO2 during DFC removal [47]. Thus,

CC E

under anoxic conditions, the strong binding ability of HA leads to more attractive sorption/complexation sites for Mn2+ than MnO2 reactive sites, reducing blockage of the reactive

A

surface site by Mn2+. As a result, more sites are made available for DFC removal in the presence of HA. In addition, when HA can adsorbed onto MnO2 surface [52], the DFC might be adsorbed at the same time. Previous research also reports that DFC can form supramolecules with HA via a strong HAchloride interaction [53]. Therefore, the MnO2-HA-DFC leads to a decrease of DFC in the liquid phase.

18

3.7 Application potential

Since water treatment under anoxic conditions (absence of oxygen) can remove high organic loads at low cost, as well as recover energy in the form of methane, they are currently growing in application. In addition, some resistant pollutants like halogenated aromatic compounds are more

IP T

easily degraded under anoxic conditions [11]. As a result, pharmaceutical removal as a post-treatment

SC R

processes is expected to receive anoxic effluent from wastewater treatment plants. Applying MnO 2 under anoxic conditions to remove pharmaceuticals like DFC is a promising and efficient process.

Compared to pharmaceutical removal technologies under oxic conditions, anoxic conditions do not

U

require aeration to ensure oxygen supply. This significantly decreases the costs of operation. Our study

N

presents basic boundary conditions for application, providing the first step towards transferring this

M

A

process from the lab into a feasible technology. Based on the results, the optimal conditions for applying MnO2 under anoxic conditions to remove DFC are a neutral pH, moderate temperature (10 –

ED

30ºC), MnO2: DFC molar ratio ≈ 2200:1, no metal ions, no PO 43-, and in the presence of organic matter

PT

(humic acids).

Experiments at varied temperatures show that DFC removal efficiency is high and stable at the

CC E

range 10 – 30ºC. In most temperate and subtropical climate areas, the water or wastewater temperature is within this range all the year round. Therefore, no extra heating or temperature-control systems are

A

necessary, leading to less cost in construction and operation.

Supplying MnO2 during this process is another potential cost for operation. As mentioned in this

paper, MnO2: DFC molar ratio needed in the application can be high under both oxic and anoxic conditions due to the low pharmaceutical concentrations and possible reactor form (like bed filter). In

19

this study, results provide information on optimizing MnO 2 dosage based on pharmaceutical removal. These results show that the application at a certain MnO2: DFC molar ratio (≥2200) is possible to maintain the high pharmaceutical removal efficiency for a long time.

We studied the effects of typical co-solutes to assess the application of this process both in

IP T

drinking water treatment and in wastewater treatment processes. The effects of these co-solutes on The

SC R

four selected metal ions are all ubiquitously present in the aquatic environment, especially in drinking water. The two concentrations tested in this study are at a similar level to what is found in drinking water based on World Health Organization [54, 55] and EU [19] regulations. The results of our study

U

show that the presence of metal ions will prevent the abiotic DFC removal with MnO2 under anoxic

A

M

MnO2 will be a wise option in application.

N

conditions. Therefore, incorporation of metal removal processes and pharmaceutical removal with

Both PO43- and organic compounds, like HA, are commonly observed co-solutes in wastewater.

ED

The diverse effects of PO43- on DFC removal indicate that expected PO43- concentrations should be

PT

taken into account in design and operation of full application technologies. HA can promote DFC removal. In addition, the HA can bind the metals including Mg, Ca, and Mn. This reduces the

CC E

inhibition of metal ions on pharmaceutical removal with MnO2. As a result, the application of manganese oxides under anoxic conditions as a tertiary treatment could be advantageous. The organic

A

matter remaining after wastewater treatment processes contain often high levels of HA and may, therefore, promote DFC removal.

Finally, it should be considered that Mn2+ will be produced in the process. As mentioned above, the highest DFC removal is obtained at 30ºC, MnO2: DFC molar ratio ≥ 2200, no co-solutes or only

20

HA. Under these experimental conditions, the Mn2+ generated during DFC removal is about 14.5 μM (0.80 mg/L). This value is much higher than the allowed concentration (0.1 mg/L) or health-based concentration (0.4 mg/L) [54, 55]. Luckily, the Mn2+ can be re-oxidize to MnO2 by bacteria, and then reused for pharmaceutical removal [26, 42]. Thus, the cycling of Mn will prevent Mn2+ from

IP T

contaminating the water when pharmaceutical removal with MnO2 is applied. Therefore, the

SC R

regeneration of MnO2 should be considered as this technology is translated into application.

In this study, about 87% DFC (2.73µM) is removed within 33 hours at 20 ºC under anoxic conditions. The removal is lower than the study carried out by Forrez, Carballa [26], in which the DFC

U

removal is nearly complete (97%) within 20 hours. However, while adsorption is not significant in our

N

study, in Forrez et al [26], sorption accounts for 25% of the DFC removal. The experimental conditions

M

A

could be the main reason for this difference. It has been reported that lower pH will lead to higher pharmaceutical removal with MnO2 under both oxic and anoxic conditions [14, 26]. So it is reasonable

ED

that DFC removal in that study at pH 4.7 is higher than in our study at pH 7. In addition, the DFC could

PT

form different oxidation products under anoxic conditions [14], probably leading to the different removal in this study and in literature. Nonetheless, the DFC removal in this study is similar to

CC E

previous studies on different pharmaceuticals. For example, removal of carbamazepine within 30 hour is about 90% at pH 2.72 under oxic condition [52], while the removal of chlortetracycline within 30

A

hours is also about 90% at pH 5.0 [13]. Based on the comparison, it is clear that applying MnO2 under anoxic conditions to remove DFC can achieve relatively high efficiency at neutral pH and it is promising in application.

21

4. Conclusion

The application of manganese oxides (MnO2) under anoxic conditions to remove diclofenac (DFC) is effective. This study identifies factors that could influence pharmaceutical removal, and is thus the first step towards translating the process into the application. Based on the experimental

IP T

results, the optimal operational conditions are neutral pH, moderate temperature (10 – 30ºC), MnO2:

SC R

DFC molar ratio around 2200:1, no metal ions, no PO 43-, and in the presence of humic acids. Results show the diverse effects of temperature, MnO 2 dosage, and phosphate on the DFC removal efficiency

and the observed initial reaction rate constant (kobs,init). While there is obvious inhibition by metal ions,

U

the significant promotion of humic acid to DFC removal may be used to compensate for this inhibitory

N

effect, including mitigating the inhibitory effect of Mn2+ formed during the process. In addition, Mn2+

M

A

produced in this process can be oxidized back to MnO2 using permanganate, manganese oxidizing bacteria, or other oxidation processes. The agents used in the re-oxidation of Mn2+ to MnO2, can also

ED

remove pharmaceuticals [56, 57]. The cycling of Mn decreases the concentration of Mn2+ in the

PT

effluent and recovers valuable manganese for reuse.

In summary, anoxic MnO2 mediated removal of pharmaceuticals from water is a potentially

CC E

interesting, sustainable and promising technology. In the future, it is critical to identify the transformation products in order to further translating this research into a valid technology. Finally,

A

extending our research to pharmaceuticals other than DFC, and to develop the technology further to scaling up and application.

Acknowledgements 22

This work is finical supported by China Scholarship Council (CSC, File No. 201308610161), and Wageningen University. The authors thank Menghan Sun and Lydia Senanu for assistance with the experiments and Hans Beijleveld, Livio Carlucci, Ilse Gerrits, and Jean Slangen for their help in

A

CC E

PT

ED

M

A

N

U

SC R

IP T

analysis, and colleagues at our department for their valuable suggestions.

23

Reference

[1] N.J. Silvy, The wildlife techniques manual, Johns Hopkins University Press, USA, 2012. [2] L. Lonappan, S.K. Brar, R.K. Das, M. Verma, R.Y. Surampalli, Diclofenac and its transformation products: Environmental occurrence and toxicity - A review, Environ. Int. 96 (2016) 127-138.

IP T

[3] A.C. Johnson, E. Dumont, R.J. Williams, R. Oldenkamp, I. Cisowska, J.P. Sumpter, Do

concentrations of ethinylestradiol, estradiol, and diclofenac in European rivers exceed proposed EU environmental quality standards?, Environ. Sci. Technol. 47 (2013) 12297-12304.

Environ. Int. 69 (2014) 28-39.

SC R

[4] N. Vieno, M. Sillanpaa, Fate of diclofenac in municipal wastewater treatment plant - A review,

[5] E.J. Tiedeken, A. Tahar, B. McHugh, N.J. Rowan, Monitoring, sources, receptors, and control measures for three European Union watch list substances of emerging concern in receiving waters – A

U

20 year systematic review, Sci. Total Environ. 574 (2017) 1140-1163.

N

[6] S. Ribeiro, T. Torres, R. Martins, M.M. Santos, Toxicity screening of diclofenac, propranolol, sertraline and simvastatin using Danio rerio and Paracentrotus lividus embryo bioassays, Ecotoxicol.

A

Environ. Saf. 114C (2015) 67-74.

M

[7] European Union, Directive 2013/39/EU of the European Parliament and of the Council, http://eurlex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2013:226:0001:0017:EN:PDF, 2013 (accessed 01 May 2017).

ED

[8] European Commission, COM(2011)876 of European Commission, http://ec.europa.eu/smart-

PT

regulation/impact/ia_carried_out/docs/ia_2012/com_2011_0876_en.pdf, 2011 (accessed 01 May 2017). [9] W.B. Liu, N.B. Sutton, H.H.M. Rijnaarts, A.A.M. Langenhoff, Pharmaceutical removal from water with iron- or manganese-based technologies: A review, Crit. Rev. Environ. Sci. Technol. 46 (2016)

CC E

1584-1621.

[10] R. Mailler, J. Gasperi, Y. Coquet, S. Deshayes, S. Zedek, C. Cren-Olive, N. Cartiser, V. Eudes, A. Bressy, E. Caupos, R. Moilleron, G. Chebbo, V. Rocher, Study of a large scale powdered activated carbon pilot: Removals of a wide range of emerging and priority micropollutants from wastewater

A

treatment plant effluents, Water Res. 72 (2015) 315-330. [11] A.-K. Ghattas, F. Fischer, A. Wick, T.A. Ternes, Anaerobic biodegradation of (emerging) organic contaminants in the aquatic environment, Water Res. 116 (2017) 268-295. [12] M. Huguet, M. Deborde, S. Papot, H. Gallard, Oxidative decarboxylation of diclofenac by manganese oxide bed filter, Water Res. 47 (2013) 5400-5408. [13] G. Chen, L. Zhao, Y.-h. Dong, Oxidative degradation kinetics and products of chlortetracycline by manganese dioxide, J. Hazard. Mater. 193 (2011) 128-138. 24

[14] W. Liu, N.B. Sutton, H.H.M. Rijnaarts, A.A.M. Langenhoff, Anoxic conditions are beneficial for abiotic diclofenac removal from water with manganese oxide (MnO2), Environ. Sci. Pollut. Res. (2018) https://doi.org/10.1007/s11356-018-1569-2. [15] H.C. Zhang, C.H. Huang, Oxidative transformation of triclosan and chlorophene by manganese oxides, Environ. Sci. Technol. 37 (2003) 2421-2430. [16] Y. Li, D. Wei, Y. Du, Oxidative transformation of levofloxacin by δ-MnO2: Products, pathways and toxicity assessment, Chemosphere 119 (2015) 282-288.

IP T

[17] K.F. Rubert, J.A. Pedersen, Kinetics of oxytetracycline reaction with a hydrous manganese oxide, Environ. Sci. Technol. 40 (2006) 7216-7221.

[18] H. Zhang, W.R. Chen, C.H. Huang, Kinetic modeling of oxidation of antibacterial agents by

SC R

manganese oxide, Environ. Sci. Technol. 42 (2008) 5548-5554.

[19] European Union, Directive 98/83/EC of the European Parliament and of the Council,

http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:1998:330:0032:0054:EN:PDF, 1998 (accessed 01 May 2017).

U

[20] European Union, Directive 91/271/EEC of the European Parliament and of the Council, http://eurlex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:31991L0271&from=EN, 1991 (accessed 01

N

May 2017).

A

[21] K. Lin, W. Liu, J. Gant, Oxidative removal of bisphenol A by manganese dioxide: efficacy,

M

products, and pathways, Environ. Sci. Technol. 43 (2009) 3860-3864. [22] K.-H. Kang, J. Dec, H. Park, J.-M. Bollag, Effect of phenolic mediators and humic acid on cyprodinil transformation in presence of birnessite, Water Res. 38 (2004) 2737-2745.

ED

[23] J. Gao, C. Hedman, C. Liu, T. Guo, J.A. Pedersen, Transformation of sulfamethazine by manganese oxide in aqueous solution, Environ. Sci. Technol. 46 (2012) 2642-2651.

PT

[24] P.A. Todd, E.M. Sorkin, Diclofenac Sodium, Drugs. 35 (1988) 244-285. [25] A.A.M. Langenhoff, D.L. Brouwersceiler, E. Jhl, J.J. Quist, W. Jgpn, Z. Ajb, G. Schraa, Microbial

CC E

reduction of manganese coupled to toluene oxidation, FEMS Microbiol. Ecol. 22 (1997) 119-127. [26] I. Forrez, M. Carballa, K. Verbeken, L. Vanhaecke, T. Ternes, N. Boon, W. Verstraete, Diclofenac oxidation by biogenic manganese oxides, Environ. Sci. Technol. 44 (2010) 3449-3454. [27] Y. He, N.B. Sutton, H.H.H. Rijnaarts, A.A.M. Langenhoff, Degradation of pharmaceuticals in wastewater using immobilized TiO2 photocatalysis under simulated solar irradiation, Appl. Catal. B-

A

Environ. 182 (2016) 132-141. [28] L.Y. Jiang, J.M. Chen, R.Y. Zhu, C. Huang, H. Ji, Degradation kinetics and estrogenic activity of 17β-estradiol removal by aqueous manganese dioxide, J. Environ. Sci. Heal. A 45 (2010) 938-945. [29] X. Liao, C. Zhang, Y. Liu, Y. Luo, S. Wu, S. Yuan, Z. Zhu, Abiotic degradation of methyl parathion by manganese dioxide: Kinetics and transformation pathway, Chemosphere 150 (2016) 9096.

25

[30] Z. Lu, J. Gan, Oxidation of nonylphenol and octylphenol by manganese dioxide: kinetics and pathways, Environ. Pollut. 180 (2013) 214-220. [31] L. Ratke, P.W. Voorhees, Growth and coarsening: Ostwald ripening in material processing, Springer Berlin Heidelberg, USA, 2013. [32] J.F. Pankow, Aquatic chemistry concepts, Taylor & Francis, USA, 1991. [33] G. Madras, B.J. McCoy, Temperature effects during Ostwald ripening, J. Chem. Phys. 119 (2003) 1683-1693.

ostwald ripening, Metallurgical and Materials Transactions A, 30 (1999) 2341-2348.

IP T

[34] V.A. Snyder, J. Alkemper, P.W. Voorhees, N. Akaiwa, The influence of temperature gradients on

[35] S. Stroes-Gascoyne, J.R. Kramer, W.J. Snodgrass, Preparation, characterization and aging of δ-

SC R

MnO2, for use in trace metal speciation studies, Appl. Geochem. 2 (1987) 217-226.

[36] J.W. Murray, The surface chemistry of hydrous manganese dioxide, J. Colloid Interf. Sci. 46 (1974) 357-371.

[37] M. Huguet, V. Simon, H. Gallard, Transformation of paracetamol into 1,4-benzoquinone by a

U

manganese oxide bed filter, J. Hazard. Mater. 271 (2014) 245-251.

N

[38] K.S. Birdi, Surface and colloid chemistry principles and applications, CRC Press, USA, 2010.

A

[39] W.-f. Tan, S.-j. Lu, F. Liu, X.-h. Feng, J.-z. He, L.K. Koopal, Determination of the point-of-zero charge of manganese oxides with different methods including an improved salt titration method, Soil

M

Sci. 173 (2008) 277-286.

[40] L. Zhong, J. Yang, L. Liu, B. Xing, Oxidation of Cr(III) on birnessite surfaces: The effect of

ED

goethite and kaolinite, J. Environ. Sci. 37 (2015) 8-14. [41] I. Vergili, Application of nanofiltration for the removal of carbamazepine, diclofenac and

PT

ibuprofen from drinking water sources, J. Environ. Manage. 127 (2013) 177-187. [42] W. Liu, A.A.M. Langenhoff, N.B. sutton, H.H.M. Rijnaarts, Biological regeneration of manganese (IV) and iron (III) for anaerobic metal oxide-mediated removal of pharmaceuticals from water,

CC E

Chemosphere, (unpublished result). [43] M.J. Gray, M.A. Malati, Adsorption from aqueous solution by δ-manganese dioxide II. Adsorption of some heavy metal cations, J. Chem. Technol. Biot. 29 (1979) 135-144. [44] R.M. Mckenzie, Proton release during adsorption of heavy metal ions by a hydrous manganese

A

dioxide, Geochimica Et Cosmochimica Acta, 43 (1979) 1855-1857. [45] J.W. Murray, The interaction of metal ions at the manganese dioxide-solution interface, Geochimica et Cosmochimica Acta, 39 (1975) 505-519. [46] O. Schmoll, Protecting groundwater for health: Managing the quality of drinking-water sources, World Health Organization, Geneva, 2006. [47] C.K. Remucal, M. Ginder-Vogel, A critical review of the reactivity of manganese oxides with organic contaminants, Environmental science. Processes & impacts, 16 (2014) 1247-1266. 26

[48] J.E. Grebel, J.A. Charbonnet, D.L. Sedlak, Oxidation of organic contaminants by manganese oxide geomedia for passive urban stormwater treatment systems, Water Res. 88 (2016) 481-491. [49] B. Pan, F. Han, G. Nie, B. Wu, K. He, L. Lu, New Strategy to enhance phosphate removal from water by hydrous manganese oxide, Environ. Sci. Technol. 48 (2014) 5101-5107. [50] L. Zhang, Y. Gao, Y. Xu, J. Liu, Different performances and mechanisms of phosphate adsorption onto metal oxides and metal hydroxides: a comparative study, J. Chem. Technol. Biot. 91 (2016) 12321239.

IP T

[51] K. Jin, J. Park, J. Lee, K.D. Yang, G.K. Pradhan, U. Sim, D. Jeong, H.L. Jang, S. Park, D. Kim, N.-E. Sung, S.H. Kim, S. Han, K.T. Nam, Hydrated manganese(II) phosphate (Mn 3(PO4)2·3H2O) as a water oxidation catalyst, J. Am. Chem. Soc. 136 (2014) 7435-7443.

manganese oxides, Environ. Sci. Pollut. Res. 19 (2012) 4206-4213.

SC R

[52] Y. He, J. Xu, Y. Zhang, C. Guo, L. Li, Y. Wang, Oxidative transformation of carbamazepine by

[53] A. Margon, A. Pastrello, D. Mosetti, P. Cantone, L. Leita, Interaction between diclofenac and soil humic acids, Soil Sediment Contam. 18 (2009) 489-496.

U

[54] World Health Organization, Guidelines for drinking-water quality, third edition, World Health

N

Organization, Geneva, 2008.

[55] J. Cotruvo, J. Bartram (Eds.), Calcium and Magnesium in drinking-water: Public health

A

significance, World Health Organization, Geneva, 2009.

M

[56] K.M. Furgal, R.L. Meyer, K. Bester, Removing selected steroid hormones, biocides and pharmaceuticals from water by means of biogenic manganese oxide nanoparticles in situ at ppb levels, Chemosphere 136 (2015) 321-326.

ED

[57] L. Hu, H.M. Martin, O. Arce-Bulted, M.N. Sugihara, K.A. Keating, T.I. Strathmann, Oxidation of carbamazepine by Mn(VII) and Fe(VI): Reaction kinetics and mechanism, Environ. Sci. Technol. 43

A

CC E

PT

(2009) 509-515.

27

SC R

IP T

Figures captions

Figure 1. DFC removal by applying MnO2 under anoxic conditions at different temperatures. The solid

U

lines represent the experiments maintained at one temperature while the dashed line represents the experiment in which the temperature changed. Experimental conditions: [MnO 2]0=7 mM, [DFC]0=3.14

CC E

PT

ED

M

A

N

µM, pH~7, I=0.01 M. Error bars are standard deviations of triplicate experiments.

Figure 2. DFC removal by applying different MnO2: DFC molar ratio under anoxic conditions. Experimental conditions: 20ºC, pH~7, I=0.01 M. Error bars are the standard deviation of triplicate

A

experiments.

28

IP T SC R

Figure 3. DFC removal by applying MnO2 under anoxic conditions in the presence of different metal ions at (a) 0.1 mM; and (b) 1 mM. Experimental conditions: [MnO 2]0=7 mM; [DFC]0= 3.14 µM, 30 ºC,

A

CC E

PT

ED

M

A

N

U

pH~7, I=0.01 M. Error bars are standard deviations of triplicate experiments.

29

Tables

Table1. Chemical structure and properties of diclofenac sodium (DFC) a Empirical Formula

C14H10Cl2NNaO2

Molecular Weight

318.13 gmol-1

pKa

4.15

logKow

1.51

Solubility in water

50mgmL-1

The data are collected from literature [24] and Sigma-Aldrich.

IP T

15307-79-6

SC R

a

CAS No.

Table 2. Calculated observed removal rate constants (kobs,init) for DFC removal by applying MnO 2 under anoxic conditions based on a pseudo-first-order kinetic model with data from the first 5 hours

U

under different experimental conditions (pH~7, I=0.01M).

MnO2: DFC molar

[DFC]0

(mM)

(µM)

Matrix

ratio

N

T(ºC）

[MnO2]0

kobs,init (h-1)

R2

([MnO2]0:[DFC]0)

7

3.14

2200:1

0.137 ± 0.007

0.99

20

10 mM MOPS

7

3.14

2200:1

0.197 ± 0.006

0.96

30

10 mM MOPS

7

3.14

2200:1

0.213 ± 0.006

0.98

40

10 mM MOPS

7

3.14

2200:1

0.041 ± 0.001

0.83

20

10 mM MOPS

1.5

3.14

480:1

0.046 ± 0.004

0.81

20

10 mM MOPS

3

3.14

950:1

0.075 ± 0.014

0.94

20

10 mM MOPS

6

3.14

1900:1

0.144 ± 0.008

0.97

10 mM MOPS

7

3.14

2200:1

0.197 ± 0.006

0.96

10 mM MOPS

7

1.57

4500:1

0.177 ± 0.007

0.94

10 mM MOPS

7

0.79

8900:1

0.147 ± 0.031

0.93

20

ED

CC E

20

PT

20

A

10 mM MOPS

M

10

10 mM MOPS, no metal ions

7

3.14

2200:1

0.213 ± 0.006

0.98

30

0.1 mM MnCl2 + 10 mM MOPS

7

3.14

2200:1

0.116 ± 0.010

0.99

30

0.1 mM CaCl2 + 10 mM MOPS

7

3.14

2200:1

0.142 ± 0.004

0.97

30

0.1 mM MgCl2 + 10 mM MOPS

7

3.14

2200:1

0.126 ± 0.011

0.96

30

0.1 mM FeCl3 + 10 mM MOPS

7

3.14

2200:1

0.126 ± 0.013

0.96

30

1 mM MnCl2 + 10 mM MOPS

7

3.14

2200:1

0.066 ± 0.007

0.93

30

1 mM CaCl2 + 10 mM MOPS

7

3.14

2200:1

0.094 ± 0.004

0.96

30

1 mM MgCl2 + 10 mM MOPS

7

3.14

2200:1

0.090 ± 0.011

0.95

30

1 mM FeCl3 + 10 mM MOPS

7

3.14

2200:1

0.151 ± 0.014

0.98

A

30

30

Table 3. DFC removal by applying MnO2 under anoxic conditions at different phosphate concentrations after 33 hours. Experimental conditions: T=30℃, [MnO2]0=7 mM; [DFC]0= 3.14 µM, pH~7, I=0.1M. PO43- (mg PL-1)

DFC Removal

0

76%

0.5

46%

1

54%

116

100%

Experimental conditions: [MnO2]0= 7 mM; [DFC]0= 5.57 µM, 30℃, pH~7, I=0.01 Ma.

IP T

Table 4. DFC removal by applying MnO2 under anoxic conditions at different HA concentrations. t=0h, no HA

t=0h, with HA

t=0.5h

t=1h

0b

0%

0%

n.a.c

25 ± 3.9%

5

0%

33 ± 0.6%

71 ± 1.2%

80 ± 0.4%

10

0%

49 ± 3.6%

76 ± 0.8%

87 ± 0.3%

15

0%

48 ± 0.7%

90 ± 0.2%

95 ± 0.4%

20

0%

75 ± 2.2%

91 ± 0.6%

96 ± 0.1%

U

SC R

HA (mg·L-1)

a

N

The data presented with the standard deviation

b

[DFC]0= 3.14 µM.

n.a.=no analyzed

A

CC E

PT

ED

M

A

c

31