Accepted Manuscript

II
Removal of phosphonates from industrial wastewater with UV/Fe , Fenton and UV/
Fenton treatment

Eduard Rott, Ralf Minke, Ulusoy Bali, Heidrun Steinmetz

PII: S0043-1354(17)30483-9
DOI: 10.1016/j.watres.2017.06.009
Reference: WR 12963

To appear in: Water Research

Received Date: 8 February 2017

Revised Date: 2 June 2017
Accepted Date: 4 June 2017

Please cite this article as: Rott, E., Minke, R., Bali, U., Steinmetz, H., Removal of phosphonates from
II
industrial wastewater with UV/Fe , Fenton and UV/Fenton treatment, Water Research (2017), doi:
10.1016/j.watres.2017.06.009.

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.
 ACCEPTED MANUSCRIPT

1 Removal of phosphonates from industrial wastewater with

2 UV/FeII, Fenton and UV/Fenton treatment

3 Eduard Rotta,*, Ralf Minkea, Ulusoy Balib, Heidrun Steinmetzc

PT
a
4 Institute for Sanitary Engineering, Water Quality and Solid Waste Management, University

5 of Stuttgart, Bandtäle 2, 70569 Stuttgart, Germany

RI
b
6 Environmental Engineering Department, Cumhuriyet University, 58140 Sivas, Turkey

SC
c
7 Chair of Resource Efficient Wastewater Technology, University of Kaiserslautern, Paul-

8 Ehrlich-Str. 14, 67663 Kaiserslautern, Germany

U
AN
9 * Corresponding author. Tel.: +49 711 68560497; fax: +49 711 68563729; E-mail address:

10 [email protected] (E. Rott)

M

11 Graphical abstract
D

Catalyst: H2SO4/ UV-FSR: UV free surface reactor

Industry: Catalyzed photolysis:
TE

FeSO4 NaOH ST: Sedimentation tank

UV lamp NT: Neutralization tank
H2SO4/NaOH
Phosphonate reduced
Cooling wastewater Receiving
Concentrate ST
system NT water
EP

UV-FSR
Fenton reagent:
Sludge
Paper/ (Photo-)Fenton: FeSO4/H2O2
Ca(OH)2/NaOH
textile UV lamp
production H2SO4
C

Rinsing Organically Wastewater

processes polluted ST ST treatment
in industry wastewater UV-FSR NT Phosphonate plant
AC

Other reduced
industries wastewater
Sludge (acidic) Sludge (neutral)
12

13 Abstract

14 Phosphonates are an important group of phosphorus-containing compounds due to their

15 increasing industrial use and possible eutrophication potential. This study involves

16 investigations into the methods UV/FeII, Fenton and UV/Fenton for their removal from a pure
1
 ACCEPTED MANUSCRIPT
17 water matrix and industrial wastewaters. It could be shown that the degradability of

18 phosphonates by UV/FeII (6 kWh/m³) in pure water crucially depended on the pH and was

19 higher the less phosphonate groups a phosphonate contains. The UV/FeII method is

20 recommended in particular for the treatment of concentrates with nitrogen-free phosphonates,

21 only little turbidity and a low content of organic compounds. Using Fenton reagent, the

PT
22 degradation of polyphosphonates was relatively weak in a pure water matrix (<20%

transformation to o-PO43–). By means of the Photo-Fenton method (6 kWh/m³), those

RI
23

24 phosphonates with the smallest numbers of phosphonate groups were easier degraded as well

SC
25 at pH 3.5 in a pure water matrix (o-PO43– formation rates of up to 80%). Despite an

26 incomplete transformation of organically bound phosphorus to o-PO43– with Fenton reagent in

27
U
an organically highly polluted wastewater (max. 15%), an almost total removal of the total P
AN
28 occurred. The most efficient total P elimination rates were achieved in accordance with the
M

29 following Fenton implementation: reaction → sludge separation (acidic) → neutralization of

30 the supernatant → sludge separation (neutral). Accordingly, a neutralization directly after the
D

31 reaction phase led to a lower total P removal extent.

TE

32 Keywords: Phosphonates, metal-catalyzed photolysis, Photo-Fenton, wastewater treatment

EP

33 1 Introduction
C

34 1.1 Motivation
AC

35 Phosphonates are versatile complexing agents due to their excellent complex stability and

36 their substoichiometric effectiveness as ‟thresholders” (Gledhill and Feijtel, 1992). They are

37 used in the paper and textile industries, are part of household and industrial detergents as well

38 as cosmetics, are used as antiscalants in membrane technology, as cooling stabilizers in

39 cooling systems and are also found in oil production, cement, electroplating, and medical

40 applications. In 1998, global consumption of phosphonates was 56,000 tons (Davenport et al.,

2
 ACCEPTED MANUSCRIPT
41 2000). A total consumption of 94,000 tons in 2012 shows a significant increase in the

42 consumption of these compounds (EPA, 2013).

43 Fig. 1 gives an overview of the chemical structures of the quantitatively most important

44 phosphonates PBTC, HEDP, NTMP, EDTMP and DTPMP. The covalent C-P bond in the

45 phosphonate groups (C-PO(OH)2) of the phosphonates ensures high molecular stability

PT
46 (Gledhill and Feijtel, 1992). Phosphonates can be divided into the group of nitrogen-free

RI
47 phosphonates, also characterized by carboxyl and hydroxyl groups, and into the group of

48 aminophosphonates. Phosphonates with more than one phosphonate group are also referred to

SC
49 as polyphosphonates. Phosphonates have a high water-solubility, a low solubility in organic

50 solvents and a very low volatility (Nowack, 2003; Gledhill and Feijtel, 1992).

U
AN
O O O
O OH OH HO
OH P P P
HO O P O OH O OH HO O
O O OH OH
HO P N HO P N N P OH
H3C HO P N N P OH N
HO OH HO HO OH
OH O OH HO O
M

P HO OH
O O P P P P P
HO OH HO O HO OH
OH OH OH O OH O O
PBTC HEDP NTMP EDTMP DTPMP
2-Phosphono- 1-Hydroxyethylidene- Nitrilotrimethyl- Ethylenediaminetetra- Diethylenetriaminepenta-
D

butane-1,2,4- (1,1-diphosphonic phosphonic acid (methylene phosphonic acid) (methylene phosphonic acid)
tricarboxylic acid acid)
51
TE

52 Fig. 1: Chemical structures of phosphonates (based on ACS, 2016).

53 Phosphonates are subject to natural elimination mechanisms (Nowack and Stone, 2000;
EP

54 Matthijs et al., 1989; Schowanek and Verstraete, 1990), which speak for long-term release of
C

55 bioavailable ortho-phosphate (o-PO43–) in water, thus a contribution of phosphonates to the

AC

56 eutrophication of water bodies cannot be excluded (Studnik et al., 2015; Grohmann and

57 Horstmann, 1989). Direct discharge of phosphonate-containing cooling water or membrane

58 concentrate, as well as insufficiently purified municipal wastewater, can also lead to increased

59 phosphorus concentrations in water bodies. Furthermore, phosphonates are associated with

60 the remobilization of toxic heavy metals bound in sediments (Gledhill and Feijtel, 1992;

61 Bordas and Bourg, 1998; Nowack, 2003). Industrial effluents are partly directed to central

3
 ACCEPTED MANUSCRIPT
62 wastewater treatment plants, either unpurified or pre-cleaned (indirect discharge). There,

63 phosphonates can interfere with phosphate precipitation by masking the precipitant

64 (Horstmann and Grohmann, 1984). In biological wastewater treatment plants, phosphonates

65 are neither degraded aerobically (Horstmann and Grohmann, 1988) nor anaerobically

66 (Nowack, 1998). However, phosphonates are removed to degrees of 50 to at least 95%

PT
67 (Metzner, 1990; Müller et al., 1984; Nowack, 1998 and 2002). Their elimination in

RI
68 wastewater treatment plants is therefore primarily ascribed to adsorption onto activated sludge

69 and metal precipitates (Nowack, 2002). Accordingly, these compounds are brought into the

SC
70 environment when the contaminated sewage sludge is applied in agriculture. But still low

71 concentrations of phosphonates may leave wastewater treatment plants via the water path

72
U
through the effluent. Removal of phosphonates from industrial wastewater before its direct or
AN
73 indirect discharge is therefore particularly desirable.
M

74 Phosphonate-contaminated industrial wastewaters can have very different compositions: on

75 the one hand, usually clear, organically only slightly polluted concentrates with typically high
D

76 water hardness and anion concentrations (e.g. cooling water, membrane concentrate), and
TE

77 organically polluted wastewaters, e.g. wastewaters from industrial rinsing processes and

78 wastewaters from the textile and paper industries.

EP

79 Studies about the elimination of phosphonates (elimination by flocculation, filtration, O3,

C

80 MnII/O2, UV/FeII) (Fischer, 1992; Klinger et al., 1998; Nowack and Stone, 2000; Boels et al.,
AC

81 2010; Zhou et al., 2014; Lesueur et al., 2005) generally comprise experiments with pure water

82 spiked with phosphonates or a very small phosphonate spectrum, so that the influence of the

83 wastewater matrix effects on the elimination processes is largely unknown. So far, there is

84 still no established procedure for the selective removal of phosphonates from wastewater.

85 Phosphonates are oxidized to a greater extent, especially in the presence of metal ions

86 (MnII/O2, UV/FeII, FeII/H2O2) (Nowack and Stone, 2000; Matthijs et al., 1989; Pirkanniemi et

4
 ACCEPTED MANUSCRIPT
87 al., 2003). This study therefore investigates the removal of phosphonates from industrial

88 wastewaters by the promising UV/FeII and (UV/)FeII/H2O2 methods exploiting these

89 properties.

90 1.2 Metal-catalyzed photolysis (UV/FeII)

PT
91 Matthijs et al. (1989), Gledhill and Feijtel (1992) and Fischer (1993) have shown that

92 phosphonates are degraded to o-PO43– in the presence of metals and UV radiation or sunlight.

RI
93 Of all the metals studied (FeIII, CrIII, ZnII, CuII, CaII), iron proved to be the most efficient

SC
94 catalyst. In the experiments of Lesueur et al. (2005), after 1.5 h at least 80% of the organically

95 bound phosphorus of the aminophosphonates NTMP, EDTMP, DTPMP and HDTMP (1

U
96 mg/L) dissolved in pure water in a UV immersion lamp reactor (150 W, medium pressure) at
AN
97 pH 3 and 5–6 in the presence of 0.2 mg/L FeII was converted to o-PO43– (due to the complex

98 analysis of phosphonates, only the total P or their oxidation to o-PO43– are determined in most
M

99 studies, not the concentrations of the phosphonates themselves). Without iron, the reaction
D

100 proceeded more slowly, but resulted in a pronounced o-PO43– formation of between 70 and
TE

101 90% at the same pH values. The reaction proceeded significantly slower at pH 10, so that the

102 o-PO43– formation extent did not exceed the 75% mark even after 1.5 h for all phosphonates.
EP

103 The experiments showed that aminophosphonates can be degraded even without a catalyst at a

104 sufficiently high UV power, the use of a catalyst, however, enhances the degradation of
C

105 phosphonates significantly and may also decrease the required UV energy consumption.
AC

106 Matthijs et al. (1989) suggested that the degradation mechanism is a photoinduced ligand-to-

107 metal charge transfer, where FeIII-complexes are subject to photolysis. Lesueur et al. (2005),

108 however, did not rule out that FeII-phosphonate complexes can also be photolytically

109 degraded.

5
 ACCEPTED MANUSCRIPT
110 So far, there are no current studies about the removal of phosphonates with the UV/FeII

111 method. All the studies mentioned have in common that the phosphonates have only been

112 spiked in pure water, so that no conclusions can be drawn about the technical feasibility of the

113 UV/FeII method for selective phosphonate removal from wastewater.

1.3 Fenton (FeII/H2O2) and Photo-Fenton (UV/FeII/H2O2)

PT
114

115 Phosphonates have a high stability against decomposition by H2O2, whereby they are able to

RI
116 function as bleach stabilizers under strongly alkaline conditions and at very high

SC
117 temperatures, such as those present in bleaching liquors of the paper and textile industries. In

118 contact with H2O2, aminophosphonates are readily oxidized to N-oxides which retain their

U
119 ability of complexation and peroxide stabilization (Croft et al., 1992; Carter et al., 1967). One
AN
120 way to oxidize H2O2 stable compounds is to convert the oxidizing agent into a much more

121 reactive form by means of catalysts as in the Fenton reaction. The complex reaction
M

122 mechanism of the Fenton reaction was described extensively by Ya Sychev and Isak (1995)
D

123 and can be summarized as follows:

TE

Fe2+ + H2O2 → Fe3+ + OH– + •OH (1)

EP

Fe3+ + H2O2 → Fe2+ + •O2H + H+ (2)

124 In the Fenton reaction, ferrous iron is oxidized to ferric iron by H2O2 to produce a highly
C

125 reactive hydroxyl radical (•OH) (Equation 1). Ferric iron can be reduced to ferrous iron again
AC

126 in the Fenton-like reaction releasing a hydroperoxyl radical (•O2H) (Equation 2). The

127 effectiveness of Fenton reagent can be enhanced by UV radiation. This is essentially due to

128 the reaction mechanisms in Equations 3 and 4 (Faust and Hoigné, 1990; Legrini et al., 1993).

6
 ACCEPTED MANUSCRIPT

Fe(OH)2+ + hv → Fe2+ + •OH (3)

H2O2 + hv → 2 •OH (4)

129 UV light stimulates the reduction of ferric iron resulting in the release of further hydroxyl

130 radicals to form ferrous iron. Thus, the accumulation of ferric iron and the premature

PT
131 termination of the Fenton reaction are counteracted. Furthermore, hydrogen peroxide is

RI
132 homolytically cleaved into two hydroxyl radicals under exposure to UV light.

133 The higher the reaction pH is, the more unreactive iron hydroxides are formed, which is why

SC
134 the Fenton reagent is generally dosed at low pH values. The precipitated sludge acts as an

U
135 adsorbent for reaction products as well as for stable compounds.
AN
136 Croft et al. (1992) investigated the mechanism of peroxide stabilization by

137 aminophosphonates in detail. They found that the peroxide stabilizing effect is not due to the
M

138 retardation of the Fenton reaction (Equation 1). On the contrary, they found that the

complexation of FeII by aminophosphonates contributed to an increase in the rate constant of

D

139

140 the Fenton reaction (75 L/mol/s without and 1·104–2·105 L/mol/s in the presence of
TE

141 aminophosphonates). The peroxide stabilizing effect is much more due to the fact that
EP

142 aminophosphonates stabilize the higher valence state of iron, hinder the effective reduction of

143 FeIII to FeII (Equation 2) and thus interfere with the Fenton cycle. Experiments on the removal
C

144 of phosphonates spiked in pure water by means of Fenton reagent were only carried out by
AC

145 Grohmann and Horstmann (1989) and by Fürhacker et al. (2005). Grohmann and Horstmann

146 (1989) described an experiment in which only 1% of a Fe3+-PBTC complex (100 µmol/L, 27

147 mg/L phosphonate) was converted to o-PO43– by equimolar FeII and H2O2. However, the pH

148 value underlying the experiment was not mentioned, so that these test results have only

149 limited validity. Fürhacker et al. (2005) showed that 1 mg/L HDTMP at pH 3 and 5.8 were

150 converted much more effectively to o-PO43– by means of Photo-Fenton reagent

7
 ACCEPTED MANUSCRIPT
151 (UV/FeII/H2O2) than by means of UV/FeII (same concentrations, only without H2O2). Thus

152 far, there have neither been conclusive studies about the removal of phosphonates with Fenton

153 reagent with pure water, nor with industrial wastewater, which means that the Fenton and the

154 Photo-Fenton method for phosphonate removal still requires considerable research.

PT
155 1.4 H2O2 dosage

156 The stoichiometrically required H2O2 concentration (cH₂O₂, 100%) can be calculated from the

RI
157 COD concentration (chemical oxygen demand for the oxidation of organic constituents in

SC
158 water) of the raw sample. The COD is determined by adding the much more reactive

159 potassium dichromate to the sample, the oxidation takes place under optimal conditions (high

U
160 temperature, acidic pH, catalyst, etc.) and the amount of potassium dichromate consumed is
AN
161 converted into oxygen equivalents. Such a conversion can also be carried out for H2O2. For

162 the reduction of 1 mol H2O2 (34 g/mol), 2 mol electrons are required (Equation 5). This is half
M

163 the amount of the substance which is required during the reduction of 1 mol O2 (32 g/mol)
D

164 (Equation 6). Taking this ratio into consideration, a factor of 2.125 results, which must be
TE

165 multiplied by the COD of the raw sample in order to determine the stoichiometrically required

166 H2O2 concentration (Equation 7). Stoichiometrically under-dosed and over-dosed H2O2
EP

167 concentrations can be set in the percentage ratio (VH₂O₂) to the stoichiometrically required

168 concentration. By means of Equation 7, this ratio can also be represented as a function of the
C

169 COD of the raw sample (Equation 8).

AC

H2O2 + H+ + 2 e– → H2O + OH– (5)

O2 + 4 H+ + 4 e– → 2 H2O (6)

8
 ACCEPTED MANUSCRIPT
g
mg 4e– 34 mol H2 O2 mg mg
cH₂O₂, 100% ቂ ቃ = – · g ·COD ቂ ቃ =2.125·COD ቂ ቃ (7)
L 2e 32 O L L
mol 2

mg
cH₂O₂ ቂ ቃ mg -1 mg
VH₂O₂ ሾ%ሿ= L
mg ·100%=cH₂O₂ ቂ ቃ ·100%·2.125 -1
·COD ቂ ቃ (8)
cH₂O₂, 100% ቂ ቃ L L
L

PT
170 If we assume that mineralization of organic matter converts C, H, P, and N into inorganic

RI
171 carbonate, water, ortho-phosphate, and nitrate, then the COD for the total oxidation of 3 mg/L

172 phosphonate (experiments of this work with a pure water matrix were always carried out with

SC
173 this concentration) can be calculated as follows: PBTC: 2.310 mg/L, HEDP: 1.398 mg/L,

U
174 NTMP: 2.088 mg/L, EDTMP: 2.751 mg/L, DTPMP: 3.098 mg/L (Supplementary data,
AN
175 Section A).

176 2 Materials and methods

M

177 2.1 Examined wastewaters

D

178 2.1.1 Overview

TE

179 For experiments with wastewaters, cooling tower effluent (concentrate) and wastewater from
EP

180 phosphonate production (organically polluted wastewater) were obtained. All samples were

181 taken randomly during operation. After sampling, all samples were stored at approx. 4°C
C

182 without further processing and were generally used within one week.
AC

183 High alkaline earth metal and COD concentrations can result in severe problems regarding the

184 detection of phosphonates in wastewater (Nowack, 1997; Klinger et al., 1997; Knepper,

185 2003). Only recently, Schmidt et al. (2014) described a method using the coupled system IC-

186 ICP-MS, which combines the advantages of low determination limits for polyphosphonates in

187 the range of 0.1 µg/L with the applicability to environmental samples. Whether this method

9
 ACCEPTED MANUSCRIPT
188 can also be applied to heavily polluted industrial wastewater and concentrates with high water

189 hardness still needs further investigation, which is why no phosphonate detection was applied

190 in this work. However, In all wastewater samples the total P of the raw samples (TP0) was

191 predominantly composed of the dissolved organic phosphorus fraction to which phosphonates

192 are attributed. Therefore, the determination of TP and o-PO43–-P was considered sufficient for

PT
193 all experiments.

RI
194 2.1.2 Cooling tower effluent of a power plant

195 The wastewater sample from a coal-fired power plant was taken directly from the cooling

SC
196 tower basin. The cooling tower is fed with flocculated river water. According to the operator,

U
197 the hardness stabilizer PBTC is used. The taken sample had only concentrations of 0.15 mg/L
AN
198 TP0 and 0.03 mg/L o-PO43–-P due to prolonged rainfall prior to sampling. Gartiser and Urich

199 (2002), however, assume typical phosphonate use concentrations in cooling systems of 1.5 to
M

200 20 mg/L. Accordingly, and to allow better comparability of the results with pure water

201 experiments in which phosphonate concentrations of 3 mg/L were used, the sample was
D

202 spiked with 3 mg/L PBTC (0.35 mg/L PBTC-P) resulting in a TP0 concentration of 0.5 mg/L.
TE

203 The cooling tower effluent (pH 7.4) was very clear (3 NTU) and colorless with a very low
EP

204 COD of 20 mg/L, a general water hardness of 35–40 dGH (160–180 mg/L Ca, 50–60 mg/L

205 Mg), an electrical conductivity of 2.1 mS/cm and an acid capacity to pH 4.3 of 1.8 mmol/L.
C

206 Further parameters were 380 mg/L chloride, 390 mg/L sulfate, 0.14–0.16 mg/L Fe, 1.3–1.5
AC

207 mg/L Al and <45 µg/L Mn.

208 2.1.3 Wastewater from phosphonate production

209 The phosphonate producer synthesizes technical solutions and solids of polyphosphonates,

210 primarily HEDP, NTMP, EDTMP, and DTPMP. Wastewater is mainly produced when the

211 reactors are rinsed. The wastewater sample taken had a very high TP0 concentration of

212 350 mg/L with an o-PO43–-P concentration of only approx. 10 mg/L. It must therefore be
10
 ACCEPTED MANUSCRIPT
213 assumed that phosphonates formed the largest proportion of the phosphorus compounds

214 present. The pH of the sample was 9 and the COD concentration was 4.8 g/L. Furthermore,

215 the following composition was typical for this kind of wastewater (minimum and maximum

216 values in three randomly taken samples): 14–28 mS/cm electrical conductivity, 110–750 NTU

217 turbidity, 5.4–18.7 mmol/L acid capacity to pH 4.3, 30–42 mg/L Ca, 2.8–3.4 mg/L Mg, 1.9–

PT
218 2.8 mg/L Fe, 1.7–2.3 mg/L Al, 47–61 µg/L Mn, 2.7–7.9 g/L chloride and 1.9–4.0 g/L sulfate.

RI
219 2.2 Reagents and chemicals

SC
220 For all stock solutions and dilutions, pure water was used produced on-site by means of an ion

221 exchanger (Seradest SD 2000) and a downstream filter unit (Seralpur PRO 90 CN).

U
222 FeSO4·7H2O (analytical grade) was purchased from Sigma-Aldrich (St. Louis, Missouri,
AN
223 U.S.A.), H2O2 solution (30%, pure) from AppliChem (Darmstadt, Germany), H2SO4 solution

224 (95–97%) from Merck (Darmstadt, Germany) and solid NaOH from VWR International
M

225 (Radnor, Pennsylvania, U.S.A.). PBTC was obtained from Zschimmer & Schwarz Mohsdorf
D

226 (Burgstädt, Germany) as a technical solution (50%, CUBLEN P 50). HEDP·H2O (≥95%) and
TE

227 NTMP (≥97%) were purchased as solids from Sigma-Aldrich. EDTMP (approx. 5.3% water

228 of crystallization) and DTPMP (approx. 16% water of crystallization) were synthesized by
EP

229 Zschimmer & Schwarz Mohsdorf. All phosphonate samples had no significant phosphate

230 impurity (ratio of phosphate-P to total P: <1%, from own measurements).

C

2.3 Experimental procedure

AC

231

232 2.3.1 Overview of the experiments

233 In Experiments 1 and 2, the degradability of the phosphonates PBTC, HEDP, NTMP,

234 EDTMP and DTPMP dosed to pure water in a concentration of 3 mg/L (0.35 mg/L PBTC-P,

235 0.90 mg/L HEDP-P, 0.93 mg/L NTMP-P, 0.85 mg/L EDTMP-P, 0.81 mg/L DTPMP-P) was

236 investigated by means of Fenton, UV/Fenton, UV/FeII, UV/H2O2 and UV only by varying the
11
 ACCEPTED MANUSCRIPT
237 FeSO4 and H2O2 concentrations as well as the pH. In Experiments 3 and 4, the UV/FeII and

238 the (Photo-)Fenton method were investigated with phosphonate-containing industrial

239 wastewater regarding the elimination of phosphorus and the required dosage concentrations

240 (variation of FeII-H2O2-ratio and H2O2 concentration). Depending on the experiment, several

241 samples were prepared in duplicate or triplicate and analyzed with single determinations. The

PT
242 results of duplicate or triplicate samples were averaged and the standard deviation was

RI
243 calculated (graphical representation with error bars).

244 Phosphonates were spiked using 3 g/L phosphonate stock solutions prepared in advance. All

SC
245 test series were carried out in 100 mL bottles on magnetic stirrers (250 rpm) filled with either

246 50 mL (samples with pure water) or 100 mL sample (samples with wastewater since due to

U
AN
247 the higher amount of analyzed parameters more volume was required). The stoichiometric

248 H2O2 concentration was calculated using the chemical oxygen demand (COD) of the raw
M

249 sample (see Section 1.4). In the case of low dosage concentrations (<0.2–0.3 g/L Fe),

250 FeSO4·7H2O was dosed from a solution prepared with pure water; otherwise, FeSO4·7H2O
D

251 was directly dosed into the sample bottles. The irradiation of samples with UV light was
TE

252 carried out as described in Section 2.3.2. Subsequent to a one-hour reaction phase, a partial

253 volume was taken from all samples in the still homogeneous state, which was analyzed for the
EP

254 o-PO43– concentration. In order to determine the sum of dissolved, precipitated and adsorbed
C

255 o-PO43–, the analysis was conducted without membrane filtration (WMF) (marked as ‟o-PO43–
AC

256 -PWMF”, for further descriptions see Section 2.4). After neutralization, the supernatants were

257 analyzed for the TP, o-PO43–-P or COD. For detailed descriptions of the experiments, see

258 Sections 2.3.3 to 2.3.6.

12
 ACCEPTED MANUSCRIPT
259 2.3.2 UV lamp

260 Four sample bottles, each equipped with magnetic rods, were randomly placed in a square

261 arrangement under the Sterisol NN 30/89 UV lamp (30 W total power, approx. 300 W/m²,

262 low-pressure, 11 W UV-C power at 254 nm) on four magnetic stirrers. The distance between

263 the sample surface and the UV lamp was generally about 10 cm. The irradiated sample

PT
264 surface was about 10 cm2. For a sample volume of 100 mL, the specific energy consumption

RI
265 per sample was therefore about 3 kWh/m³ (=300 W/m²·10 cm²·1 h/100 mL). Of this, the

266 active UV-C range accounted for 1.1 kWh/m³ (=3 kWh/m³·11 W/30 W). With a sample

SC
267 volume of 50 mL, the specific energy consumption corresponded to 6 kWh/m³, of which 2.2

268 kWh/m³ accounted for the UV-C range.

U
AN
269 2.3.3 FeII/H2O2 in a pure water matrix (Experiment 1)

270 1 liter of pure water was spiked with a phosphonate concentration of 3 mg/L, brought to pH
M

271 3.5 by means of H2SO4, and then intensively stirred and analyzed for the TP0 concentration.
D

272 FeSO4 was added at various concentrations to 50 mL samples of this solution (molar FeII-
TE

273 H2O2-ratios: 1:30–5:1 and one ratio in the range between 1:100 and 1:400 depending on the

274 phosphonate to take into account an equimolar FeII-phosphonate-ratio as well; the highest
EP

275 molar FeII-phosphonate-ratio tested was 2,000:1). Then, H2O2 was added to the samples in a

276 stoichiometrically eightfold excess (at this H2O2 concentration a maximum of o-PO43–
C

277 formation could be observed in preliminary experiments; Supplementary data, Section B)

AC

278 resulting in a maximum pH decrease down to pH 2.75 depending on the FeII concentration.

279 After stirring for 1 h in the open bottle (an additional pH drop of no more than 0.2 was

280 observed here; the pH range of this experiment was therefore 2.5–3.5), the o-PO43–-PWMF

281 concentration was determined in each sample. Each sample was prepared in duplicate.

13
 ACCEPTED MANUSCRIPT
282 2.3.4 UV/FeII, UV/H2O2, UV/FeII/H2O2 in a pure water matrix (Experiment 2)

283 1 liter of pure water was provided with a phosphonate concentration of 3 mg/L and FeSO4

284 several times in an equimolar ratio to the phosphonate. Batches without FeSO4

285 (c(FeII):c(phosphonate)=0) were also prepared. The initial pH (3.5, 5.0, 6.5, 8.0, 9.5) was

286 adjusted by means of H2SO4 or NaOH, the solution was then stirred intensively and analyzed

PT
287 for the TP0 concentration. Optionally, H2O2 was added in the stoichiometrically sixteen-fold

RI
288 concentration to 50 mL samples of this solution (at this H2O2 concentration a maximum of

289 o-PO43– formation could be observed in preliminary experiments; Supplementary data,

SC
290 Section B). After stirring for 1 h under the UV lamp, the o-PO43–-PWMF concentration was

291 determined in each sample. Due to the experimental setup (the UV lamp was mounted directly

U
AN
292 above the reaction vessel), a pH correction was not possible during this stirring phase.

293 However, the pH after 1 h was measured and depending on the tested variant (UV, UV/FeII,
M

294 UV/H2O2, UV/FeII/H2O2), different, mostly only small pH changes were observed. Thus, the

295 following pH ranges were tested in this experiment: 3.4–3.6, 4.6–5.0, 6.2–6.9, 7.0–8.0, 8.0–
D

296 9.5. Each sample was prepared in triplicate.

TE

297 2.3.5 UV/FeII, UV/H2O2, UV/FeII/H2O2 in cooling tower effluent (Experiment 3)

EP

298 The cooling tower effluent spiked with 3 mg/L PBTC was brought to pH 3.5 using H2SO4.

299 Optionally, 4.6 mg/L FeII and optionally 30 mg/L H2O2 were added to 100 mL samples of this
C

300 sample in order to test the following four variants: UV, UV/FeII, UV/H2O2, UV/FeII/H2O2
AC

301 (because 100 mL samples were used, the specific UV energy consumption was half as high

302 [3 kWh/m³] as in the experiments with pure water [6 kWh/m³]). Each of these variants was

303 prepared fourfold (four bottles with the same contents). After stirring for 1 h under the UV

304 lamp, the o-PO43–-PWMF concentration (homogeneous) was determined for each variant in all

305 four samples. The addition of chemicals and the contact time had not resulted in any

14
 ACCEPTED MANUSCRIPT
306 significant pH change (maximum drop down to pH 3.4) due to the strong buffer capacity of

307 the wastewater. Two of the initial four equally prepared samples of each variant were

308 immediately membrane-filtered (0.45 µm pore size, nylon filter) and the o-PO43– as well as

309 the TP concentrations of the filtrate (pH 3.5) were determined. The two other samples were

310 first neutralized by means of NaOH (pH 7.0), then membrane-filtered and now the o-PO43– as

PT
311 well as the TP concentrations of the filtrate were determined.

RI
312 2.3.6 FeII/H2O2 in organically polluted wastewater (Experiment 4)

SC
313 The raw sample of phosphonate production wastewater was brought to pH 2.5 by means of

314 H2SO4. After intensive mixing, it was analyzed for the TP0, o-PO43– and COD concentrations.

U
315 FeSO4 and H2O2 were added to several 100 mL samples of this sample resulting in a pH
AN
316 decreased to 2.2–2.5. Each sample was prepared in replicate. Each 100 mL sample was then

317 stirred in the open bottle for 1 h. Here, an insignificant additional pH drop of no more than
M

318 0.15 was observed. Immediately thereafter, the o-PO43–-PWMF concentration (homogeneous)
D

319 was determined in each sample. The further procedure involved two variants. One of the
TE

320 replicated samples was treated according to the neutralization variant 1. Here, the sample was

321 neutralized directly with NaOH and then sedimented for 15–24 h. The resulting supernatant
EP

322 (‟supernatant N1 (neutral)”) was then analyzed for parameters such as o-PO43–-PWMF and TP.

323 The other sample was treated according to the neutralization variant 2. Here, the sample was
C

324 first sedimented for 15–24 h without prior pH adjustment. 60 mL of the resulting supernatant
AC

325 (‟supernatant N2a (acidic)”) were transferred into an empty bottle. At that point, the o-PO43–

326 -PWMF and TP concentrations of this supernatant were determined. The transferred sample was

327 first neutralized by means of NaOH, and then sedimented for 15–24 h. The resulting

328 supernatant (‟supernatant N2b (neutral)”) was then analyzed for the o-PO43–-PWMF and TP

329 concentrations.

15
 ACCEPTED MANUSCRIPT
330 2.4 Analytical methods

331 All glass materials which came into contact with the sample were rinsed with hydrochloric

332 acid and pure water in advance. The total P (TP) determination was carried out according to

333 ISO 6878 (molybdenum blue method) by means of a one-hour peroxodisulfate digestion. ISO

334 6878 provides for the separation of the suspended sample components by membrane filtration

PT
335 in advance to the o-PO43– analysis. In order to determine the extent of organically bound

RI
336 phosphorus oxidized to o-PO43– during the UV/FeII and Fenton reactions, such membrane

337 filtration was not carried out (marked as ‟o-PO43–-PWMF”) (see also Supplementary data,

SC
338 Section C). Precipitated phosphate or adsorbed o-PO43–, which are separated from the sample

by a filtration, were thus measured together with dissolved o-PO43–. This is important, since

U
339

during the reaction sludge (predominantly Fe(OH)3) precipitates, so that o-PO43– formed in
AN
340

341 the reaction phase either adsorbs or reacts to non-soluble FePO4. The ascorbic acid solution
M

342 (reducing agent) required according to ISO 6878 was always added directly to the partial

343 volumes for the o-PO43–-PWMF determination in order to interrupt the oxidation processes.
D

344 Turbid and colored samples were additionally treated with a compensation solution according
TE

345 to ISO 6878. The extinctions were measured with the UV/VIS spectrophotometer JASCO V-

346 550.
EP

347 The pH was determined using the WTW pH electrode SenTix 81 in combination with the
C

348 WTW pH91 instrument. The chemical oxygen demand (COD) was determined using the
AC

349 Hach Lange cuvette rapid tests LCK 414 and LCK 514. In these tests, a certain quantity of

350 sample is added to a predefined mixture of sulfuric acid, potassium dichromate, silver sulfate

351 and mercury salt, heated up to 148°C for two hours in a thermostat (Hach Lange HT200S)

352 and then analyzed in a photometer (Hach Lange DR2800). H2O2-containing samples were

353 pre-treated with Aspergillus niger catalase (Sigma-Aldrich) (max. 20 mg/L) before the

16
 ACCEPTED MANUSCRIPT
354 phosphorus and COD analysis as hydrogen peroxide in the sample was found to be interfering

355 with the analysis (Talinli and Anderson, 1992).

356 3 Results and discussion

357 3.1 FeII/H2O2 in a pure water matrix (Experiment 1)

PT
358 In order to clarify the question of to what extent phosphonate-phosphorus is oxidized to

RI
359 o-PO43–-P by exposure to Fenton reagent for one hour without influence of UV radiation,

360 H2O2 was added to each phosphonate at pH 2.5–3.5 in pure water in stoichiometrically

SC
361 eightfold excess, while the FeII-H2O2-ratio was subject to a variation.

U
362 No significant reaction to o-PO43– was observed for all phosphonates when no FeII was dosed
AN
363 (sole H2O2 dosage: c(o-PO43–-PWMF)/c(TP0)<2%, not shown in Fig. 2 due to logarithmic

364 representation), which is plausible with regard to their use as bleach stabilizers in bleaching
M

365 liquors.
D
TE
C EP
AC

17
 ACCEPTED MANUSCRIPT
100

c(o-PO43–-PWMF)/c(Total P0) (%)

PBTC
90 HEDP
80 NTMP
70 EDTMP
60 DTPMP
50
40
30
20

PT
10
0
0.001 0.010 0.100 1.000 10.000
II
c(Fe )/c(H2O2) (mol/mol)
366

RI
367 Fig. 2: Oxidation of phosphonates (initial concentration of 3 mg/L) to o-PO43– at pH 2.5–3.5 (pH

368 3.5 was adjusted using H2SO4; the pH could decrease during the reaction phase to a

SC
369 minimum of 2.5) in water as a result of a one-hour treatment with Fenton reagent at a

stoichiometrically eightfold H2O2 dosage concentration with varying FeII-H2O2-ratios

U
370

371
AN
(cH₂O₂=8·2.125·COD0, COD0 according to Section 1.4). For circled symbols:

372 c(FeII):c(phosphonate)=1:1.
M

373 In Fig. 2, those points indicated by a circle correspond to an equimolar Fe-phosphonate-ratio

374 for each phosphonate. At the equimolar Fe-phosphonate-ratio, only a slight o-PO43– formation
D

375 was observed for nitrogen-free phosphonates. At higher FeII-H2O2-ratios – thus also at higher
TE

376 FeII-phosphonate-ratios – some of the iron was present in the excess to the phosphonate and

377 therefore not necessarily completely complexed. For polyphosphonates, this resulted in a
EP

378 maximum 20% conversion to o-PO43–. PBTC stuck out with the highest conversion rates of up
C

379 to 50%. Accordingly, in the presence of Fenton reagent, a preferred oxidation of the nitrogen-
AC

380 free phosphonate PBTC to o-PO43– occurred, whereas in the presence of MnII/O2 (Nowack

381 and Stone, 2000) and O3 (Klinger et al., 1998) aminophosphonates were preferably oxidized.

382 The very reactive hydroxyl radical which is active in the Fenton reaction is very short-lived

383 and therefore does not react selectively with sites of increased electron density such as O2 and

384 O3 (Lee and von Gunten, 2010). Furthermore, Croft et al. (1992) had shown that N-oxides of

385 the aminophosphonates react with hydroxyl radicals to form long-lived nitroxides in which

18
 ACCEPTED MANUSCRIPT
386 the C-P bonds remain intact, what could explain the low formation extent of o-PO43– in the

387 case of aminophosphonates. It is thus probable that the C-PO(OH)2 groups in the

388 phosphonates represent the most stable component in the molecule to hydroxyl radicals, and

389 thus are the limiting factor in the degradation rate of phosphonates by Fenton reagent.

390 Especially the nitrogen-free phosphonates with three carboxyl groups in the case of PBTC and

PT
391 an alcohol group in the case of HEDP have functional groups which are not present in

RI
392 aminophosphonates. In order to better understand the role of these functional groups with

393 regard to the stability of these phosphonates, more detailed investigations on the degradation

SC
394 pathways of the phosphonates would be required.

U
395 Molar FeII-H2O2-ratios >1 were associated with lower o-PO43– formation rates. A strong
AN
396 formation of hydroxyl radicals can occur by means of excessively high excess of iron, so that

397 the hydroxyl radicals react essentially only among themselves (recombination). Furthermore,
M

398 excess Fe2+ ions can act as free-radical scavengers (Barbusiński and Filipek, 2001). As a

399 result, there are less radicals which can oxidize the phosphonate.
D
TE

400 3.2 UV/FeII, UV/H2O2, UV/FeII/H2O2 in a pure water matrix (Experiment 2)

401 Fig. 3 summarizes the conversion rates of all phosphonates to o-PO43– at different pH ranges
EP

402 in a pure water matrix as a result of a one-hour treatment with various combinations of UV,

FeII (equimolar to phosphonate) and H2O2 dosed in a sixteenfold stoichiometric excess (this
C

403
AC

404 dosage concentration was chosen to obtain maximum conversion rates to o-PO43–). For this

405 experiment, the pH of the phosphonate-containing sample was adjusted after the possible FeII

406 dosing. Due to the subsequent H2O2 dosing, the pH value decreased only by a maximum of

407 0.3. Only after the one-hour reaction phase partially a stronger pH change was observed

408 predominantly in the alkaline range. The areas grayed out in Fig. 3 thus illustrate the

19
 ACCEPTED MANUSCRIPT
409 minimum and maximum pH value that could be measured in the corresponding samples

410 during the experiment.

100
90 (a) UV
80
70
60

PT
50
40
30
20

RI
10
1000
90 (b) UV/FeII

SC
80
70
60
50
c(o-PO43–-PWMF)/c(Total P0) (%)

U
40
30
AN
20
10
100
0
90 (c) UV/H2O2
M

80
70
60
50
D

40
30
TE

20
10
1000
90 (d) UV/FeII/H2O2
EP

80
70
60
50
C

40
30
AC

20
10
0
3 4 5 6 7 8 9 10
pH
PBTC HEDP NTMP EDTMP DTPMP
411

412 Fig. 3: Oxidation of phosphonates (initial concentration of 3 mg/L) to o-PO43– as a result of a

413 one-hour UV treatment (6 kWh/m³) with combinations of FeII (equimolar to phosphonate)

414 and H2O2 (cH₂O₂=16·2.125·COD0, COD0 according to Section 1.4) at the pH ranges 3.4–

20
 ACCEPTED MANUSCRIPT
415 3.6, 4.6–5.0, 6.2–6.9, 7.0–8.0, 8.0–9.5 in water. The initial pH (3.5, 5.0, 6.5, 8.0, 9.5) was

416 adjusted using H2SO4 and NaOH in the phosphonate and FeSO4 containing sample before

417 addition of H2O2. The pH ranges are therefore due to pH change after the H2O2 dosing

418 and during the reaction phase.

419 None of the phosphonates could be degraded to o-PO43– by UV irradiation alone (Fig. 3a),

PT
420 which underlines their UV resistance in the noncomplexed state. Lesueur et al. (2005) had

421 been able to detect an oxidation of aminophosphonates to o-PO43– even without a catalyst

RI
422 using a medium-pressure UV lamp at pH values between 3 and 10. In their experiments, a

SC
423 significantly higher specific UV energy consumption was used (112 kWh/m³ compared to the

424 experiment described here of 6 kWh/m³), which could explain the oxidation despite the

425 absence of a catalyst.

U
AN
426 In the presence of FeII, PBTC was very well degraded by UV radiation (up to 80%) while
M

427 polyphosphonates showed only low degradation rates below 30% (Fig. 3b). For PBTC,

428 highest degradation occurred in the acidic milieu (pH 3.4–5.0), while polyphosphonates at pH
D

429 3.4–3.6 were not significantly degraded to o-PO43– at all. With increasing pH, the
TE

430 degradability of PBTC decreased slightly and fell sharply in the alkaline range. At this pH

431 range, a precipitation of metal hydroxides was observed for both PBTC and HEDP. However,
EP

432 due to the o-PO43– analysis without previous membrane filtration (o-PO43–WMF), o-PO43– that
C

433 adsorbed onto Fe(OH)2 or Fe(OH)3 was also determined. The precipitation of iron hydroxide
AC

434 suggests that PBTC and HEDP were partly present in the noncomplexed state. In this state,

435 phosphonates are not degradable solely by UV radiation of 6 kWh/m³ (Fig. 3a). Approx. 10%

436 of NTMP was converted to o-PO43– almost constantly over the entire pH range. The

437 degradation rate of these three phosphonates between pH 5 and 8 decreased according to their

438 number of phosphonate groups (in parentheses): PBTC (1) >> HEDP (2) > NTMP (3).

439 EDTMP (4) and DTPMP (5) proved to be stable against UV radiation of 6 kWh/m³, even in

21
 ACCEPTED MANUSCRIPT
440 the presence of FeII. Lesueur et al. (2005) had shown a significantly stronger oxidation of the

441 aminophosphonates NTMP, EDTMP and DTPMP to o-PO43– with higher UV performance

442 (112 kWh/m³) and lower pH values (3 and 5–6) than was the case at pH 10. The results of

443 Experiment 2 show that the stable C-PO(OH)2 groups of the phosphonates may also be the

444 limiting factor with respect to the conversion rate of the phosphonates to o-PO43– by UV/FeII

PT
445 as in Experiment 1.

RI
446 All phosphonates were converted to o-PO43– by treatment with UV/H2O2 in the neutral and

447 alkaline pH range between 30 and 50% (Fig. 3c). With increasing pH, the aminophosphonates

SC
448 also showed a slightly increasing oxidation to o-PO43–, which is due either to an improved

449 efficiency of the UV/H2O2 process at higher pH values (Benjamin and Lawler, 2013) or

U
AN
450 possibly to a better degradability of phosphonate species with a larger negative charge. It is

451 noticeable that EDTMP and DTPMP could be degraded better by UV/H2O2 over the entire
M

452 tested pH range than by UV/FeII. This means that EDTMP and DTPMP do not necessarily

453 have to be present as complexes in order to be significantly converted to o-PO43– by hydroxyl

D

454 radicals. The nitrogen-free phosphonates PBTC and HEDP showed almost no pH dependence
TE

455 with respect to their conversion to o-PO43– by UV/H2O2 over the entire investigated range.

456 Interestingly, a small peak with respect to the oxidation to o-PO43– could be observed for both
EP

457 nitrogen-free phosphonates around pH 5. At pH 5, the doubly negatively charged species is

C

458 predominant in both phosphonates (pKa values of PBTC: 3.74, 4.23, 5.14, 6.80, 9.05 (Liu et
AC

459 al., 2000), pKa values of HEDP: <1, 2.54, 6.97, 11.41 (Carroll and Irani, 1967)). It is possible

460 that this reacts preferably to o-PO43–.

461 While PBTC was the only phosphonate reacting more weakly to o-PO43– by UV/FeII/H2O2

462 with increasing pH, the polyphosphonates showed a tendency towards more pronounced o-

463 PO43– formation, especially with increasing pH (Fig. 3d). For the phosphonates HEDP and

464 EDTMP, the maximum c(o-PO43–WMF)/c(TP0)-ratio was already established above a neutral

22
 ACCEPTED MANUSCRIPT
465 pH value, whereas for the phosphonates NTMP and DTPMP even with the highest pH values

466 (pH 8.0–9.5) the conversion to o-PO43– still increased further. PBTC was by far the most

467 easily degradable phosphonate (70–80% conversion to o-PO43–) in the acidic pH range, while

468 the conversion rate to o-PO43– of all other phosphonates at pH 3.4–3.6 was, at most, 35%. At

469 this pH, the oxidation of the phosphonates to o-PO43– correlated with their phosphonate group

PT
470 number (in parentheses): PBTC (1) >> HEDP (2) > NTMP (3) ≈ EDTMP (4) > DTPMP (5).

In the neutral pH range, the extent of the conversion of the phosphonates to o-PO43– was as

RI
471

472 follows: HEDP (2) ≈ PBTC (1) > EDTMP (4) ≈ NTMP (3) > DTPMP (5). At pH >6, HEDP

SC
473 had the highest degradability to o-PO43–, while here the maximum conversion of all other

474 phosphonates to o-PO43– was between 50 and 60%.

U
The decreasing degradation of PBTC to o-PO43– with increasing pH can be explained by the
AN
475

476 precipitation of iron hydroxide. This precipitation occurs rather for nitrogen-free
M

477 phosphonates, since these form less stable complexes with metals compared to

478 aminophosphonates (Knepper, 2003). Thus, at higher pH values, free PBTC is present, which
D

479 is significantly less degradable by UV/H2O2 compared to the FeII-PBTC complex (Fig. 3c). In
TE

480 the case of polyphosphonates, degradation by UV/FeII/H2O2 predominantly occurred in the

481 alkaline range. In this pH milieu, polyphosphonates preferably complex the dosed FeII, so that
EP

482 it can be assumed that the improved degradation with alkaline pH values is essentially due to
C

483 the fact that the phosphonates are more easily attacked in a complexed form by hydroxyl
AC

484 radicals (at least for HEDP and NTMP). Furthermore, this experiment clearly showed that

485 phosphonates are converted less to o-PO43– by means of UV/FeII and UV/FeII/H2O2 the larger

486 their number of phosphonate groups is. As already explained in Section 3.1, the C-PO(OH)2

487 groups in the phosphonates are considered to be the most stable component in the molecule,

488 so that a high number of phosphonate groups also limits the degradation rate of the

489 phosphonates. In conclusion, the treatment with UV/FeII/H2O2 (Photo-Fenton) in water

23
 ACCEPTED MANUSCRIPT
490 proved to be a good means of oxidizing all phosphonates. The next experiment (Exp. 3)

491 should show whether this good oxidisability is also present in wastewater where UV light

492 absorbing turbidity may be present or oxidation of other readily degradable organic

493 components as well as free-radical scavengers can occur.

3.3 UV/FeII, UV/H2O2, UV/FeII/H2O2 in cooling tower effluent (Experiment 3)

PT
494

495 Fig. 4 summarizes the results of an experiment in which cooling tower effluent spiked with

RI
496 3 mg/L PBTC was treated with UV, UV/FeII, UV/H2O2 and UV/FeII/H2O2 at pH 3.5. The

SC
497 results of the oxidation extent of organically bound phosphorus to o-PO43– (o-PO43–WMF) after

498 the one-hour reaction phase, as well as the phosphorus fractions immediately following

U
499 membrane filtration (‟MF”) at pH 3.5, and the same for the sample neutralized immediately
AN
500 after the reaction phase and also membrane-filtered, are shown.
M

0.50
0.45 Total P0
0.40
0.35
o-PO43–-P
D

0.30
mg/L

0.25
0.20
TE

0.15
0.10
0.05
0.00
Total P0 and o-PO43–-P o-PO43–-PWMF o-PO43–-P (MF) Total P (MF) o-PO43–-P (MF) Total P (MF)
(raw sample) (homogeneous) (pH 3.5) (pH 3.5) (pH 3.5) (pH 7.0) (pH 7.0)
EP

Raw sample UV UV/FeII UV/H2O2 UV/FeII/H2O2

501
502 Fig. 4: Formation and concentration of o-PO43– and TP in cooling tower effluent as a result of a
C

503 one-hour UV radiation (3 kWh/m³) at pH 3.5 (adjusted with H2SO4) with 4.6 mg/L FeII and
AC

504 30 mg/L H2O2 of a membrane-filtered (MF) sample before (pH 3.5) and after

505 neutralization using NaOH (pH 7.0). No significant pH drop could be observed during the

506 contact time.

507 The oxidation extent of organically bound phosphorus to o-PO43– was similar to the results

508 found in Experiment 2 with PBTC dissolved in pure water (UV/FeII ≥ UV/FeII/H2O2 >

509 UV/H2O2 > UV). An oxidation of PBTC to o-PO43– already occurred during treatment with

24
 ACCEPTED MANUSCRIPT
510 UV light without chemical dosage, whereas no oxidation of PBTC to o-PO43– had been

511 observed by sole UV radiation in the pure water test, even at a higher UV energy consumption

512 (6 kWh/m³). The UV oxidation of PBTC can be explained by metals present in the cooling

513 tower effluent (e.g., Fe and Al, Section 2.1.2). Thus, a photolysis catalyzed by metals could

514 also occur without FeII dosage, although in this case the o-PO43– formation was only half as

PT
515 high as with FeII dosage.

RI
516 In the sample treated with UV/FeII/H2O2 could a reduced concentration of o-PO43– and TP

517 after membrane filtration at pH 3.5 be seen. In this batch, an oxidation of ferrous iron to ferric

SC
518 iron was evident by H2O2. In contrast to FeII, FeIII precipitates at pH 3.5 (Barrera-Díaz et al.,

519 2003), resulting in adsorbent for o-PO43– and organically bound phosphorus. Furthermore,

U
AN
520 non-soluble FePO4 could form.

521 The neutralization immediately after the one-hour reaction phase resulted in an almost
M

522 complete removal of the o-PO43– formed in the samples treated with UV/FeII and

523 UV/FeII/H2O2 and a total of 80–85% TP decrease. FeII precipitates above pH 8.7 at the dosed
D

524 concentration of 4.6 mg/L FeII (Barrera-Díaz et al., 2003). Thus, in both samples, FeIII (mainly
TE

525 the adsorbent iron hydroxide and non-soluble FePO4) had to be formed. In the sample treated
EP

526 with UV/FeII/H2O2, this was done predominantly by the oxidation of FeII to FeIII by H2O2, as

527 already described. In the sample treated with UV/FeII, this oxidation must have taken place
C

528 during the neutralization by O2 dissolved in the sample.

AC

529 The samples without addition of FeII (UV, UV/H2O2) retained the concentration of dissolved

530 o-PO43– largely after neutralization. The TP could only be slightly removed by neutralization

531 since there was no iron as precipitant.

25
 ACCEPTED MANUSCRIPT
532 3.4 FeII/H2O2 in phosphonate production wastewater (Experiment 4)

533 During the reaction phase (c(FeII)/c(H2O2)=0.08 [g/g]) at pH 2.1–2.5, only about 15% of the

534 organically bound phosphorus of phosphonate production wastewater were oxidized to o-

535 PO43– (50 out of 350 mg/L P) (Fig. 5 left). When the wastewater was treated with Fenton

536 reagent in the course of the neutralization variant 1 (direct neutralization following the

PT
537 reaction phase), an almost 100% TP (supernatant) and o-PO43–-PWMF (supernatant) decrease

RI
538 occurred at an H2O2 concentration of 10.2 g/L (corresponding to the stoichiometrically

539 required oxidant concentration, thus 100% H2O2) and 0.82 g/L FeII. With the neutralization

SC
540 variant 2 (reaction, then sedimentation (acidic), then neutralization of the supernatant and

U
541 sedimentation (neutral)), an almost 100% TP decrease was already achieved at 5.1 g/L H2O2

(50%) and 0.41 g/L FeII.

AN
542

FeII (g/L)
M

0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0
350 o-PO43–-PWMF (homogeneous) o-PO43–-PWMF (supernatant) Total P (supernatant)
300
250
mg/L

200
150
100
TE

50
0
0 25 50 75 100 125 0 25 50 75 100 125 0 25 50 75 100 125 150
cH₂₂O₂₂ · 100 % · COD0–1 · 2.125–1 (%)
Supernatant N1 (neutral) Supernatant N2a (acidic) Supernatant N2b (neutral)
543
EP

544 Fig. 5: Formation and concentration of o-PO43– and TP after treatment of phosphonate

production wastewater with Fenton reagent at a constant c(FeII)/c(H2O2) ratio of 0.08

C

545

[g/g] (proportional increase of FeII and H2O2) and pH 2.1–2.5 (pH 2.5 adjusted with
AC

546

547 H2SO4; the pH could decrease during the reaction phase to a minimum of 2.1) according

548 to the neutralization variants 1 and 2. Raw sample: 350 mg/L TP, 4.8 g/L COD0.

549 The low formation extent of o-PO43– was in accordance with the o-PO43– formation found in

550 Experiment 1 (pure water) when no UV radiation was applied. The low o-PO43– formation

551 indicates that complexing agents may have remained in the sample after the reaction phase,

26
 ACCEPTED MANUSCRIPT
552 probably predominantly in the form of phosphonates and their degradation products capable

553 of complex formation. Since the reaction rate of the reduction of FeIII to FeII (Equation 2) is

554 slower than the reaction rate of the oxidation of FeII to FeIII (Equation 1), over time an

555 accumulation of FeIII and thus the precipitate Fe(OH)3 in the samples with Fenton reagent

556 occurs. Although, as mentioned, there were probably still complexing agents in the sample

PT
557 after the reaction phase, at pH 2.1–2.5, iron hydroxide sludge could precipitate due to the

RI
558 decreased effectiveness of these complexing agents because of the low pH. Furthermore, o-

559 PO43–, phosphonates, as well as their degradation products have the best adsorption affinity

SC
560 for iron hydroxides at low pH values (Nowack and Stone, 1999). When the sample was

561 neutralized directly after the reaction phase (neutralization variant 1), the residual complexing

562
U
agents could have interfered with the precipitation of iron hydroxide, and thus, with the
AN
563 formation of adsorbent for the TP, because the increase in pH led to a more pronounced
M

564 effectiveness of the residual complexing agents. The complete absence of a TP elimination in

565 a small dosage concentration range (below 30 %) and mere part removal at values between 50
D

566 and 100 % (stoichiometric ratio) when performing the neutralization variant 1 indicated that
TE

567 clearly. Thus, the neutralization variant 2 performed better than neutralization variant 1 since

568 most of the complexing agents were separated from the sample when a sludge separation
EP

569 occurred at an acidic pH.

C

570 The similarly high TP concentrations in both of the supernatants (N2a and N2b) resulting
AC

571 from the neutralization variant 2 show that the phosphorus compounds were removed mainly

572 in the first sedimentation stage (‟supernatant N2a (acidic)”), thus at acidic pH. Phosphorus

573 compounds which had not been removed in the first sedimentation stage were separated by

574 iron hydroxide flocks at the latest produced during neutralization. The slightly higher o-PO43–-

575 PWMF concentration in the supernatant of neutralization variant 1 (‟supernatant N1 (neutral)”)

576 compared to the concentration prior to the sedimentation (‟o-PO43–-PWMF (homogeneous)”)

27
 ACCEPTED MANUSCRIPT
577 can be explained by residual H2O2 in the sample, which could further oxidize organic

578 phosphorus compounds during the sedimentation phase.

579 4 Conclusions

580 The degradability of the phosphonates PBTC, HEDP, NTMP, EDTMP and DTPMP by means

PT
581 of metal-catalyzed photolysis (UV/FeII), Fenton method (FeII/H2O2) and Photo-Fenton method

582 (UV/FeII/H2O2) was investigated.

RI
583 This work has shown that polyphosphonates can be converted to phosphate by a maximum of

SC
584 20% by means of Fenton reagent at pH 2.5–3.5 in a pure water matrix even at a very high

585 excess of H2O2 (also high Fe excess, thus, influence of complex formation is marginal).

586
U
Therefore, in order to oxidize phosphonates to a higher extent, the oxidation process has to be
AN
587 supported by UV radiation (whether in the UV/FeII or in the UV/FeII/H2O2 process).

Interestingly, the degradation of polyphosphonates in the UV/FeII/H2O2 process increased

M

588

589 with increasing pH. This is in contrast to the general assumption that the Fenton process
D

590 works best at low pH values. Further tests with wastewater would therefore have to
TE

591 investigate whether, in the special case of the desired phosphonate oxidation, higher reaction

592 pH values than usual lead to better results on the TP removal.

EP

CATALYZED PHOTOLYSIS: UV-FSR: UV free surface reactor

catalyst: H 2 SO 4/ NT: neutralization tank
FeSO4 NaOH ST: sludge separation tank
C

UV RT: reaction tank

H 2SO 4 /NaOH
phosphonate
concen-
ST reduced
AC

trate UV-FSR NT wastewater

FENTON:
Fenton reagent: sludge
FeSO4 /H 2 O 2 Ca(OH) 2/NaOH

H 2SO 4
phosphonate
raw ST ST reduced
wastewater RT NT wastewater

sludge sludge
(acidic) (neutral)
593

28
 ACCEPTED MANUSCRIPT
594 Fig. 6: Proposal for the implementation of a continuously operated wastewater treatment plant

595 according to the UV/FeII and Fenton principle for industrial wastewater containing

596 phosphonates.

597 An experiment with clear, Ca2+-rich, organically little loaded and PBTC-containing cooling

598 water could confirm the results from the pure water experiments with PBTC. Accordingly, the

PT
599 organically bound phosphorus was successfully degraded to phosphate in this wastewater by

600 the UV/FeII and the UV/FeII/H2O2 method. Furthermore, the TP removal performance

RI
601 between the UV/FeII and the UV/FeII/H2O2 method was hardly different. It was found that the

SC
602 dosed iron precipitates even when no H2O2 is dosed in the UV/FeII process through pH

603 neutralization carried out following the reaction phase at pH 3.5 if, during the reaction phase,

604
U
the phosphonate is decomposed completely or decomposed into reaction products which are
AN
605 no longer capable of complexation. Thus, neutralization can take place immediately following

606 the reaction phase (neutralization variant 1) (Fig. 6). The o-PO43– formed during the reaction
M

607 phase is thereby discharged from the wastewater via filtration or sedimentation. The UV/FeII
D

608 process is very selective and is recommended only if the organically bound phosphorus
TE

609 fraction in the wastewater is mainly composed of compounds susceptible to metal-catalyzed

610 photolysis such as phosphonates. Furthermore, only a very small sludge amount is to be
EP

611 expected due to the fact that only very low amounts of iron are sufficient for successful

612 oxidation of the phosphonate. Accordingly, the UV/FeII method is particularly suitable for
C

613 concentrates with nitrogen-free phosphonates, a low turbidity and COD load and no
AC

614 competing complexing agents.

615 The low oxidation extent of polyphosphonates to phosphate by means of sole Fenton reagent

616 (no UV radiation) in pure water could also be found for a phosphonate-containing, organically

617 highly polluted wastewater matrix (phosphonate production wastewater). However, it was

618 discovered that complete conversion of the phosphonate to phosphate was not absolutely

29
 ACCEPTED MANUSCRIPT
619 necessary for removal of TP in this kind of wastewater. Since phosphonates adsorb on the

620 Fenton sludge, they are also removed via the sludge separation (sedimentation or filtration).

621 The experiments showed that the following procedure should be applied for a chemical-

622 saving application of the Fenton method when applied to organically polluted wastewater

623 without support of UV radiation (neutralization variant 2, Fig. 6): reaction → sludge

PT
624 separation (acidic) → neutralization of the supernatant → sludge separation (neutral; in the

case of full precipitation of added FeII in the acidic sludge separation no sludge separation

RI
625

626 could be necessary here). If, on the other hand, neutralization takes place immediately

SC
627 following the reaction phase, phosphonates or other complexing agents remaining in the

628 sample after the reaction phase can interfere with the precipitation of iron hydroxide due to

629
U
their higher effectiveness as complexing agents at higher pH. Thereby, less adsorbent is
AN
630 present for the removal of phosphonates and their degradation products, which, moreover,
M

631 exhibit a weaker adsorption affinity for iron hydroxide in the neutral and alkaline pH range

632 than in the acidic pH range (Nowack and Stone, 1999).

D

633 The UV lamp used in our experiments was too weak for very turbid, heavily organically
TE

634 polluted wastewaters, which is why further investigations with stronger UV lamps have to be

635 carried out to find out whether there is a more pronounced phosphonate degradation by the
EP

636 Photo-Fenton method compared to the Fenton method in these kind of wastewaters.
C

637 Acknowledgements
AC

638 The authors are grateful for the financial support by the Willy-Hager-Stiftung, Stuttgart. We

639 would also like to thank the employees of Zschimmer & Schwarz Mohsdorf GmbH & Co. KG

640 for providing phosphonate samples.

30
 ACCEPTED MANUSCRIPT
641 References

642 ACS, 2016. Database of https://scifinder.cas.org. American Chemical Society.

643 Barbusiński, K., Filipek, K., 2001. Use of Fenton's reagent for removal of pesticides from

644 industrial wastewater. Polish Journal of Environmental Studies 10(4), 207–212.

PT
645 Barrera-Díaz, C., Ureña-Nuñez, F., Campos, E., Palomar-Pardavé, M., Romero-Romo, M.,

RI
646 2003. A combined electrochemical-irradiation treatment of highly colored and polluted

647 industrial wastewater. Radiation Physics and Chemistry 67(5), 657–663.

SC
648 Benjamin, M.M., Lawler, D.F, 2013. Water Quality Engineering – Physical/Chemical

649 Treatment Processes. John Wiley and Sons, Inc., USA, ISBN 978-1-118-16965-0.

U
AN
650 Boels, L., Tervahauta, T., Witkamp, G.J., 2010. Adsorptive removal of

651 nitrilotris(methylenephosphonic acid) antiscalant from membrane concentrates by iron-

M

652 coated waste filtration sand. Journal of Hazardous Materials 182(1–3), 855–862.
D

653 Bordas, F., Bourg, A.C.M., 1998. Effect of complexing agents (EDTA and ATMP) on the
TE

654 remobilization of heavy metals from a polluted river sediment. Aquatic Geochemistry

655 4(2), 201–214.

EP

656 Carroll, R.L., Irani, R.R., 1967. On the acidity of substituted methylenediphosphonates and

657 their interaction with alkali metal ions. Inorganic Chemistry 6(11), 1994–1998.
C
AC

658 Carter, R.P., Crutchfield, M.M., Irani, R.R., 1967. Nitrilotri(methylenephosphonic acid) N-

659 oxide and nitrilotriacetic acid N-oxide: Acidity and complexing of calcium and magnesium

660 ions. Inorganic Chemistry 6(5), 943–946.

661 Croft, S., Gilbert, B.C., Smith, J.R.L., Stell, J.K., Sanderson, W.R., 1992. Mechanisms of

662 peroxide stabilization. An investigation of some reactions of hydrogen peroxide in the

31
 ACCEPTED MANUSCRIPT
663 presence of aminophosphonic acids. Journal of the Chemical Society, Perkin Transactions

664 2, 2, 153–160.

665 Davenport, B., DeBoo, A., Dubois, F., Kishi, A., 2000. CEH report: chelating agents. SRI

666 Consulting, Menlo Park, CA, USA. Cited by Nowack (2003).

PT
667 EPA, 2013. Phosphonates in detergents. EPA detergent phosphonates dossier, European

668 Phosphonate Association.

RI
669 Faust, B.C., Hoigné, J., 1990. Photolysis of Fe3+-hydroxy complexes as sources of OH

SC
670 radicals in clouds, fog and rain. Atmospheric Environment 24(1), 79–89.

671 Fischer, K., 1992. Sorption of chelating agents (HEDP and NTA) onto mineral phases and

672
U
sediments in aquatic model systems, Part II: Sorption onto sediments and wastewater
AN
673 sludges. Chemosphere 24(1), 51–62.
M

674 Fischer, K., 1993. Distribution and elimination of HEDP in aquatic test systems. Water

675 Research 27(3), 485–493.

D

676 Fürhacker, M., Lesueur, C., Pfeffer, M., Popp, M., Mentler, A., 2005. Phosphonates – AMPA
TE

677 (aminomethylphosphonic acid), estimation of origin, environmental concentrations and

678 photolysis degradation (Phosphonate – AMPA (Aminomethylphosphonsäure),

EP

679 Herkunftsabschätzung, Umweltkonzentrationen und Photolyseabbau). Final report, Vienna.

C

680 Gartiser, S., Urich, E., 2002. Use of environmentally compatible chemicals in cooling water
AC

681 conditioning (Einsatz umweltverträglicher Chemikalien in der Kühlwasserkonditionie-

682 rung). Umweltforschungsplan des Bundesministeriums für Umwelt, Naturschutz und

683 Reaktorsicherheit, Forschungsbericht 200 24 233, UBA-FB 000298.

32
 ACCEPTED MANUSCRIPT
684 Gledhill, W.E., Feijtel, T.C.J., 1992. Environmental properties and safety assessment of

685 organic phosphonates used for detergent and water treatment applications. The Handbook

686 of Environmental Chemistry 3(F), Anthropogenic Compounds – Detergents, Springer-

687 Verlag, ISBN 3-540-53797-X.

688 Grohmann, A., Horstmann, B., 1989. The use of phosphonates under environmental aspects

PT
689 (Der Einsatz von Phosphonaten unter umwelttechnischen Gesichtspunkten). Research

RI
690 report 102 063 22 UFA-FB 89-018, Umweltbundesamt Berlin, Germany.

691 Horstmann, B., Grohmann, A., 1984. The influence of phosphonates on the phosphate

SC
692 elimination 1. Part: 1-hydroxyethane-1,1-diphosphonic acid (HEDP) (Der Einfluß von

U
693 Phosphonaten auf die Phosphatelimination 1. Teil: 1-Hydroxyethan-1,1-diphosphonsäure
AN
694 (HEDP)). Zeitschrift für Wasser- und Abwasser-Forschung 17, 177–181.

695 Horstmann, B., Grohmann, A., 1988. Investigations about the biological degradation of
M

696 phosphonates (Untersuchungen zur biologischen Abbaubarkeit von Phosphonaten). Vom

D

697 Wasser 70, 163–178.

TE

698 ISO 6878:2004. Water quality – Determination of phosphorus – Ammonium molybdate

699 spectrometric method. Beuth Verlag GmbH, Berlin, Germany.

EP

700 Klinger, J., Sacher, F., Brauch, H.J., Maier, D., 1997. Determination of organic phosphonates
C

701 in aqueous samples using liquid chromatography/particle-beam mass spectrometry. Acta

AC

702 hydrochimica et hydrobiologica 25(2), 79–86.

703 Klinger, J., Sacher, F., Brauch, H.J., Maier, D., Worch, E., 1998. Behaviour of phosphonic

704 acids during drinking water treatment. Vom Wasser 91, 15–27.

705 Knepper, T.P., 2003. Synthetic chelating agents and compounds exhibiting complexing

706 properties in the aquatic environment. Trends in Analytical Chemistry 22(10), 708–724.

33
 ACCEPTED MANUSCRIPT
707 Lee, Y., von Gunten, U., 2010. Oxidative transformation of micropollutants during municipal

708 wastewater treatment: Comparison of kinetic aspects of selective (chlorine, chlorine

709 dioxide, ferrateVI, and ozone) and non-selective oxidants (hydroxyl radical). Water

710 Research 44(2), 555–566.

711 Legrini, O., Oliveros, E., Braun, A.M., 1993. Photochemical processes for water treatment.

PT
712 Chemical Reviews 93(2), 671–698.

RI
713 Lesueur, C., Pfeffer, M., Fuerhacker, M., 2005. Photodegradation of phosphonates in water.

714 Chemosphere 59(5), 685–691.

SC
715 Liu, Y., Gao, L., Yu, L., Guo, J., 2000. Adsorption of PBTCA on alumina surfaces and its

U
716 influence on the fractal characteristics of sediments. Journal of Colloid and Interface
AN
717 Science 227(1), 164–170.

718 Matthijs, E., de Oude, N.T., Bolte, M., Lemaire, J., 1989. Photodegradation of ferric ethylene-
M

719 diaminetetra(methylenephosphonic acid) (EDTMP) in aqueous solution. Water Research

D

720 23(7), 845–851.

TE

721 Metzner, G., 1990. Behavior of organic complexing agents based on phosphonic acid in

722 sewage treatment plants (Verhalten von organischen Komplexbildnern auf

EP

723 Phosphonsäurebasis in Kläranlagen). Umweltverträglichkeit von Wasch- und

C

724 Reinigungsmitteln, Münchener Beiträge zur Abwasser-, Fischerei- und Flußbiologie 44,
AC

725 323-336, R. Oldenbourg Verlag GmbH, München, ISBN 3-486-26198-3.

726 Müller, G., Steber, J., Waldhoff, H., 1984. On the influence of hydroxyethanediphosphonic

727 acid on the phosphate elimination with FeCl3 and the remobilization of heavy metals:

728 results from laboratory and field tests (Zum Einfluß von Hydroxyethandiphosphonsäure

729 auf die Phosphatelimination mit FeCl3 und die Remobilisierung von Schwermetallen:

730 Ergebnisse von Labor- und Feldversuchen). Vom Wasser 63, 63–78.

34
 ACCEPTED MANUSCRIPT
731 Nowack, B., 1997. Determination of phosphonates in natural waters by ion-pair high-

732 performance liquid chromatography. Journal of Chromatography A 773(1–2), 139–146.

733 Nowack, B., 1998. The behavior of phosphonates in wastewater treatment plants of

734 Switzerland. Water Research 32(4), 1271–1279.

PT
735 Nowack, B., Stone, A.T., 1999. Adsorption of phosphonates onto goethite-water interface.

736 Journal of Colloid and Interface Science 214(1), 20–30.

RI
737 Nowack, B., Stone, A.T., 2000. Degradation of nitrilotris(methylenephosphonic acid) and

SC
738 related (amino)phosphonate chelating agents in the presence of manganese and molecular

739 oxygen. Environmental Science & Technology 34(22), 4759–4765.

740
U
Nowack, B., 2002. Aminopolyphosphonate removal during wastewater treatment. Water Re-
AN
741 search 36(18), 4636–4642.
M

742 Nowack, B., 2003. Environmental chemistry of phosphonates. Water Research 37(11), 2533–

743 2546.
D

744 Pirkanniemi, K., Sillanpää, M., Sorokin, A., 2003. Degradative hydrogen peroxide oxidation
TE

745 of chelates catalysed by metallophthalocyanines. The Science of the Total Environment

746 307(1–3), 11–18.

EP

747 Schmidt, C.K., Raue, B., Brauch, H.J., Sacher, F., 2014. Trace-level analysis of phosphonates
C

748 in environmental waters by ion chromatography and inductively coupled plasma mass
AC

749 spectrometry. International Journal of Environmental Analytical Chemistry 94(4), 385–

750 398.

751 Schowanek, D., Verstraete, W., 1990. Phosphonate utilization by bacterial cultures and

752 enrichments from environmental samples. Applied and Environmental Microbiology

753 56(4), 895–903.

35
 ACCEPTED MANUSCRIPT
754 Studnik, H., Liebsch, S., Forlani, G., Wieczorek, D., Kafarski, P., Lipok, J., 2015. Amino

755 polyphosphonates – chemical features and practical uses, environmental durability and

756 biodegradation. New Biotechnology 32(1), 1–6.

757 Talinli, I., Anderson, G.K., 1992. Interference of hydrogen peroxide on the standard COD

758 test. Water Research 26(1), 107–110.

PT
759 Ya Sychev, A., Isak, V.G., 1995. Iron compounds and the mechanisms of the homogeneous

RI
760 catalysis of the activation of O2 and H2O2 and of the oxidation of organic substrates.

761 Russian Chemical Reviews 64(12), 1105–1129.

SC
762 Zhou, Z., Qiao, W., Lin, Y., Shen, X., Hu, D., Zhang, J., Jiang, L.M., Wang, L., 2014.

U
763 Phosphonate removal from discharged circulating cooling water using iron–carbon micro-
AN
764 electrolysis. Water Science and Technology 70(3), 524–532.
M
D
TE
C EP
AC

36
Dr. Eduard Rott ACCEPTED MANUSCRIPT UNIVERSITÄT
INSTITUTE FOR SANITARY ENGINEERING, WATER STUTTGART
QUALITY AND SOLID WASTE MANAGEMENT

Institute for Sanitary Engineering, Water Quality and Solid Waste Management D-70569 Stuttgart (Büsnau)
● Bandtäle 2 ● 70569 Stuttgart
Bandtäle 2

Editorial Board Phone +49 711 685 – 60497

Water Research Fax +49 711 685 – 63729

PT
To whom it may concern
E-Mail:
[email protected]

Date

RI
th
February 7 2017

SC
Subject: Paper highlights

U
AN
• Phosphonates could be degraded well using UV/FeII and UV/Fenton.
M

• The stability of phosphonates depends on their number of phosphonate groups.

D

• UV/FeII treatment worked well on clear, less organically polluted wastewater.

TE

• For a better Fenton performance, the acidic supernatant should be neutralized.

C EP
AC