Journal of Water Process Engineering 17 (2017) 188–196
Contents lists available at ScienceDirect
Journal of Water Process Engineering journal homepage: www.elsevier.com/locate/jwpe
Removal of phosphorus from phosphonate-loaded industrial wastewaters via precipitation/ﬂocculation
Eduard Rotta, , Ralf Minkea, Heidrun Steinmetzb a b
Institute for Sanitary Engineering, Water Quality and Solid Waste Management, University of Stuttgart, Bandtäle 2, 70569 Stuttgart, Germany Chair of Resource Eﬃcient Wastewater Technology, University of Kaiserslautern, Paul-Ehrlich-Str. 14, 67663 Kaiserslautern, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords: Phosphonates Flocculation Precipitation Wastewater treatment Industrial wastewater
Phosphonates are to be removed from industrial wastewater before they reach municipal wastewater treatment plants or surface waters. Industrial wastewaters contaminated with the phosphonates 2-phosphonobutane-1,2,4tricarboxylic acid (PBTC), 1-hydroxyethylidene-(1,1-diphosphonic acid) (HEDP), nitrilotrimethylphosphonic acid (NTMP), ethylenediamine tetra(methylene phosphonic acid) (EDTMP) and diethylenetriamine penta (methylene phosphonic acid) (DTPMP) can be subdivided into: (1) mostly clear concentrates with a high water hardness, and (2) organically polluted wastewaters, for instance from the paper and textile industries. Our own ﬂocculation experiments with FeIII and AlIII salts showed that, at a pH of 7.5, the adsorption aﬃnity of polyphosphonates onto iron hydroxides and aluminum hydroxides signiﬁcantly decreases with an increasing number of C-P bonds (HEDP > NTMP > EDTMP > DTPMP). In comparison to pure water spiked with phosphonates, the total P removal from concentrates occurred at similar and even lower ﬂocculant dosage concentrations (ß = 4–8) (ß is the molar ratio of dosed metal concentration to total P concentration in the raw sample). In organically polluted industrial wastewaters, the formation of ﬂocks, and therefore the total P decrease, only occurred after exceeding a certain ﬂocculant concentration, which varied strongly depending on the wastewater type (ß = 0.6–86). Below that concentration, the ﬂocculant underwent complexation, and no elimination could be observed. For most of the wastewaters, the required ß values for at least 80% total P decrease were very similar for both FeIII and AlIII. Furthermore, Ca(OH)2 as ﬂocculant turned out to be an eﬀective tool for total P removal and simultaneous softening of calcareous concentrates.
1. Introduction 1.1. Motivation Phosphonates such as PBTC, HEDP, NTMP, EDTMP, and DTPMP (Fig. 1) are used in a wide range of applications due to their high complex stability, corrosion inhibiting, dispersing and substoichiometric eﬀectiveness as “thresholders” [1,2]. They are applied in paper and textile industry and are an essential ingredient of household and industrial cleaning products as well as of cosmetic products. Furthermore, they are utilized as antiscalants in membrane technology, as hardness stabilizers in cooling systems, in the metal industry, in oil production, as medicine, and for cement modiﬁcation. In 1998, the worldwide use of phosphonates was 56,000 t [3,4]. A total consumption of 94,000 t in 2012 illustrates the signiﬁcant increase in their use . Phosphonates are subject to natural elimination mechanisms [7–9], which suggest a long-term release of bioavailable phosphate (o-PO43−) ⁎
Corresponding author. E-mail address: [email protected]
http://dx.doi.org/10.1016/j.jwpe.2017.04.008 Received 18 January 2017; Received in revised form 3 April 2017; Accepted 19 April 2017 2214-7144/ © 2017 Elsevier Ltd. All rights reserved.
in water. Thus, a contribution of phosphonates to eutrophication cannot be excluded . Grohmann and Horstmann  clearly proved in their ﬁeld trials that HEDP serves as a nutrient resource in slowlyﬂowing water bodies due to its photolysis. More critical degradation products of some phosphonates in Fig. 1 are glyphosate and the metabolite AMPA [12–17]. Because phosphonates are complexing agents, remobilized metals can reach toxic concentrations for aquatic life forms or enter drinking water via bank ﬁltration [18,19]. In wastewater treatment plants, phosphonates are removed solely by the adsorption on activated sludge . Biological degradation neither occurs under aerobic  nor anaerobic [22–24] conditions during wastewater treatment. Yet, still very little is known about the fate and concentrations of phosphonates in the environment due to their complex analysis [10,25]. Phosphonates are present in municipal wastewater as well as in various industrial wastewaters. Industrial wastewaters – either untreated or treated – are partly discharged to central wastewater treatment plants (indirect discharge). Since phosphonates can complex
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
Fig. 1. Structural formulas of important phosphonic acids (based on Ref. ).
This adsorption tendency can be used in wastewater treatment plants targeting a selective phosphonate elimination.
the precipitant, the phosphate precipitation in wastewater treatment plants can be disrupted [26,27]. Furthermore, due to the complexation, precipitant dosages are needed in large surplus. This can result in interferences with the nitriﬁcation due to acidiﬁcation and a constant lack of bioavailable phosphate . Other phosphonate-rich industrial wastewaters are discharged directly into water bodies (direct discharge), such as cooling wastewaters and concentrates from membrane plants. This can result in locally increased phosphonate concentrations in the receiving waters. Phosphonates must, therefore, be eliminated locally in advance to their indirect or direct discharge. For industrial companies, the selective phosphonate removal should be performed in accordance with the branch stream wastewater concept. Accordingly, the stream with the highest phosphonate concentration should be treated before its mixture with the total wastewater stream. Industrial wastewaters contaminated with phosphonates can be subdivided into two groups. The ﬁrst group includes mostly clear, organically less-polluted concentrates with a high water hardness and characteristically high anion concentrations. The second group consists of strongly organically polluted wastewaters, for instance from rinsing processes or from the paper and textile industries.
1.2.2. Precipitation/ﬂocculation Precipitation and ﬂocculation are well-known processes for the removal of phosphate. Eq. (1) gives an example for the precipitation of phosphate that is transformed into a hardly soluble compound by the chemical bond with a precipitant (Me: metal, for example, Fe and Al). Phosphonates form complexes with typical ﬂocculants such as FeIII and AlIII. These ﬂocculants hydrolyze to hardly soluble metal hydroxide ﬂocks (Eqs. (2) and (3)) which provide an adsorption surface for inorganic and organic substances such as phosphonates [16,26,27,31,32]. Adsorbed onto the ﬂocks, these substances can be removed from the wastewater by sedimentation or ﬁltration. The ß value describes the molar ratio between the dosed ﬂocculant and the phosphorus in the raw sample (Eq. (4)). The complexation of the ﬂocculant by phosphonates leads to the expectation of overstoichiometric ß values in wastewater treatment. In the following, the terms ‘ﬂocculant’ and ‘precipitant’ will be used as synonyms.
1.2. Elimination processes 1.2.1. General mechanisms Phosphonates can be eliminated by diﬀerent physicochemical and biological processes in natural systems (Fig. 2). They can be oxidized by O2 as Mn2+ complexes . Furthermore, they can be degraded photolytically in the presence of metals , transformed to o-PO43− by ozone (O3) [15–17] or undergo a very slow hydrolysis . Only a very small proportion of microorganisms is capable of metabolizing phosphonates by the direct degradation of the very stable CeP bond [9,29]. Additionally, phosphonates – despite their high water solubility – exhibit a high adsorption tendency toward mineral surfaces due to their polar and negatively charged phosphonate groups [22,23,30].
Me3+ + PO43− → MePO4 ↓
FeCl3 + 3 H2O → 3 H+ + 3 Cl− + Fe(OH)3 ↓
Al2(SO4)3 + 6 H2O → 6 H+ + 3 SO42− + 2 Al(OH)3 ↓
Nowack and Stone [33–35] conducted numerous experiments regarding the adsorption of phosphonates on the iron oxide hydroxide goethite. At pH 7, NTMP adsorbed on goethite (3 mg/L NTMP, 0.42 g/L goethite) entirely within just a few minutes. The highest adsorption extents of 100% were found in the pH range lower than 7 (lowest pH measured was 5.6). With the increase of the pH the adsorption extent decreased gradually until above pH 12 where adsorption could no longer be observed. However, between pH 7 & 12 an adsorption increase of phosphonates occurred in the presence of CaII. This adsorption enhancement was ascribed to ternary surface-phosphonate-Ca complexes . Furthermore, phosphonates compete with oPO43– for free adsorption sites on iron hydroxide . Zenobi and Rueda  conducted similar experiments with the aluminum oxide hydroxide boehmite. They also found the highest elimination extents of NTMP to be in the pH range lower than 7. Horstmann and Grohmann [26,27] discovered a signiﬁcant impact by aminophosphonates on the formation of ﬂocks in the ﬂocculation process. Due to the dispersing eﬀect of aminophosphonates, a signiﬁcantly high proportion of the total P in the ﬁltrate of a gravel ﬁlter was still colloidal or dissolved, so the separating barrier could be penetrated. In the investigations by Klinger et al.  50 μg/L PBTC enriched in buﬀered (pH 7), demineralized water could be reduced to 90% with ß values above 60. For the 90% decrease of 50 μg/L EDTMP a ß value of 30 was required. The dosage of NaOH (caustic soda) or Ca(OH)2 (lime milk) leads to
Fig. 2. Presence, degradation and adsorption paths of phosphonates in natural water bodies (Me: metal).
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
the wastewater of a coal power plant cooling tower (CT: PBTC as hardness stabilizer, pH 7.4, ca. 20 mg/L COD, 0.50 mg/L total P, 0.03 mg/L o-PO43–-P, 40 dGH) were investigated. Highly organically polluted wastewaters were covered by the sampling of the phosphonate-containing wastewater branch stream of a paper production facility (PA: supernatant of paper machine wastewater, DTPMP, pH 7.3, 1.0 g/L COD, 1.1 mg/L total P, 0.1 mg/L o-PO43−-P) and wastewater from a phosphonate production facility (production of polyphosphonates, mainly HEDP, NTMP, EDTMP and DTPMP) mainly accruing during rinsing processes (PR: pH 9.0, 4.8 g/L COD, 400 mg/L total P, 10 mg/L o-PO43−-P). The samples were stored at 4 °C without prior treatment and used for analysis and experiments within a week. In all wastewater samples, the total P mainly consisted of the dissolved organic phosphorus fraction—which is typical for phosphonates. The wastewater samples of the cooling tower (3 mg/L PBTC; 0.35 mg/L PBTC-P plus 0.15 mg/L total P of raw wastewater resulting in 0.5 mg/L total P) and the paper production (3 mg/L DTPMP; 0.8 mg/L DTPMP-P plus 0.3 mg/L total P of raw wastewater resulting in 1.1 mg/L total P) were spiked with phosphonates in advance of the experimental procedure, since due to seasonal inﬂuences, less phosphonate than usual was applied in the production process while these samples were taken.
an increase of the pH. This pH increase causes the precipitation of metals and alkaline earth metals in the sample, therefore the precipitation products can serve as an adsorbent for the wastewater ingredients that need to be removed. Metzner and Nägerl  are the only authors who describe a successful elimination of phosphorus from a phosphonate and polyacrylate containing water conditioning agent (Aktiphos 640) with Ca(OH)2. Almost all investigations regarding the above-mentioned elimination mechanisms include experiments with a pure water matrix [31,32,34]. Furthermore, a wide spectrum of phosphonates is usually not investigated [31,32,37]. So far, there is no experience or established method for the selective removal of phosphonates from phosphonaterich wastewaters. However, the results from pure water experiments suggest a good elimination potential of phosphonates from wastewater by ﬂocculation/precipitation, especially from concentrates. According to current knowledge, it is not possible to estimate which ß values are necessary in the treatment of phosphonate-containing industrial wastewaters, the extent to which the elimination process depends on the pH value and is impaired by the phosphonate or other complexing agents in the wastewater or in which form the technical implementation should occur. Furthermore, according to current knowledge, AlIII allows the elimination of phosphonates with signiﬁcantly lower β values than FeIII. Whether this is also applicable to wastewater also needs further clariﬁcation. This study, therefore, describes experiments of the ﬂocculation/precipitation method (FeIII, AlIII, Ca(OH)2, NaOH) regarding its elimination potential for phosphonates from industrial wastewaters by particularly taking into account a wide range of diﬀerent wastewater matrices and diﬀerent phosphonates.
2.3. Reagents and chemicals Pure water was produced with an ion exchanger (Seradest SD 2000) and a downstream ﬁltering unit (Seralpur PRO 90 CN). 2Phosphonobutane-1,2,4-tricarboxylic acid (PBTC) as technical solution (CUBLEN P 50 [50%]), ethylenediamine tetra(methylene phosphonic acid) (EDTMP·1.4H2O) and diethylenetriamine penta(methylene phosphonic acid) (DTPMP·6H2O) both as solid substances were provided by Zschimmer and Schwarz. 1-Hydroxyethylidene-(1,1-diphosphonic acid) (HEDP·H2O) and nitrilotrimethylphosphonic acid (NTMP) were SigmaAldrich products. H2SO4 (95–97%, p.a.), Ca(OH)2 (p.a.) and FeCl3·6H2O (p.a.) were Merck products. NaOH (Ph. Eur.) was purchased from VWR Chemicals and Al2(SO4)3·16H2O (p.a.) was purchased from Carl Roth.
2. Materials and methods 2.1. Experimental concept Publications about the ﬂocculation of phosphonates usually describe experiments with only one phosphonate and therefore do not provide any general information about the comparability of all relevant phosphonates for diﬀerent ﬂocculants. An experiment with a pure water matrix (Exp. 1) therefore covered all relevant phosphonates (PBTC, HEDP, NTMP, EDTMP, DTPMP) with regard to their removal by two diﬀerent ﬂocculants (FeIII and AlIII). In Experiments 2 and 3, the inﬂuence of wastewater matrix on phosphorus elimination was investigated. For this purpose, the necessary ﬂocculant concentrations (FeIII and AlIII) were determined for four diﬀerent industrial wastewaters and the inﬂuence of the pH on the elimination process was checked. Membrane concentrates are a typical example of wastewaters with high phosphonate concentrations (antiscalants). Due to their high water hardness, these wastewaters are usually very unstable, so that longterm blockages occur in waste disposal pipes. Therefore, in Exp. 4, next to the removal of phosphorus also the simultaneous softening by NaOH and Ca(OH)2 was investigated for membrane concentrate. Experiments 2 and 4 were designed from the perspective of a water process practitioner, which is why the sole ﬂocculant dosage and no pH adjustment was applied with regards to chemical saving in plants of large scale.
2.4. Experimental procedure 2.4.1. General procedure Flocculant concentrations were adjusted using FeCl3·6H2O or Al2(SO4)3·16H2O dissolved in pure water. Phosphonates were spiked with 3 g/L bidistilled water stock solutions. All experiments were conducted in 100 mL bottles with 100 mL samples on magnetic stirrers with a stirring rate of 100 rpm. Phosphonates adsorb on mineral surfaces within a very short time , so that for the experiments of this work a contact time of approx. 15 min was considered suﬃcient. All sedimentation phases lasted 15–25 h. No diﬀerence of results could be noticed within this range. At the end of the experiments, the supernatants were withdrawn with an Eppendorf pipette in order to analyze parameters such as total P, o-PO43−-P (only in membrane concentrate samples), COD concentrations (only in paper and phosphonate production wastewaters) and general water hardness (only in membrane concentrate samples). In the following, ‘n = 1’ stands for single batches and ‘n = 2’ for duplicate batches. Here, the results are presented as mean values with standard deviations. Detailed descriptions of each experiment are given in Sections 2.4.2–2.4.4.
2.2. Industrial wastewater samples The selection of wastewater samples had to cover a representative range of phosphonate applications. In Section 1.1 a brief summary of possible types of wastewaters is given. Phosphonate-rich wastewaters can be subdivided into a group of hard, less organically polluted concentrates and a group of highly organically polluted wastewaters. In order to investigate the behavior of phosphonates in concentrates, membrane concentrate of a drinking water nanoﬁltration plant (MC: antiscalant DTPMP, pH 7.9, ca. 15 mg/L COD, 1.1–1.5 mg/L total P, 0.5–0.6 mg/L o-PO43–-P, 110–120 dGH [general water hardness]) and
2.4.2. FeIII and AlIII applied to pure water (Exp. 1) According to Fettig et al. , for Exp. 1 with a pure water matrix, two procedures were performed. In the ‘adsorption experiment’, the phosphonate was added to freshly precipitated metal hydroxides, whereas in the ‘ﬂocculation experiment’, the ﬂocculants were precipitated in the presence of the phosphonate. By distinguishing Exp. 1 into these two procedures, the contribution of sorption processes to the overall removal mechanism could be studied . 190
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
Fig. 3. The amount of phosphonates in supernatants from the adsorption experiment and the ﬂocculation experiment at pH 7.0–7.5 in pure water as a function of the molar ratios of added ﬂocculant to initial phosphonate concentration and added ﬂocculant to initial phosphorus concentration of the phosphonate (c0 = 3 mg/L phosphonate).
• Adsorption experiment: In several pure water samples the desired
12.5 with H2SO4 or NaOH. Each sample was stirred for 15 min followed by a subsequent sedimentation phase. The supernatant was analyzed for the COD, total P and o-PO43−.
concentrations of FeIII and AlIII were adjusted (n = 1). Then, the pH was adjusted to 7.5 using NaOH. Stirring followed for 5 min, during which time ﬂock formation occurred. Next, the phosphonate was dosed resulting in a 3 mg/L concentration. Now, each sample was stirred for 15 min followed by a subsequent sedimentation phase. During the 15 min of stirring the pH was continuously monitored. It remained in the range of 7.0–7.5. After the sedimentation phase, the supernatant was analyzed for its total P concentration. Flocculation experiment: In several pure water samples the desired concentrations of FeIII and AlIII were adjusted (n = 1). Additionally, the phosphonate was dosed resulting in a 3 mg/L concentration. Then, the pH was adjusted to 7.5 using NaOH. Next, each sample was stirred for 15 min followed by a subsequent sedimentation phase. During these 15 min of stirring, the pH was continuously monitored. As in the adsorption experiment, it remained in the range of 7.0–7.5. Because the pH drift in all samples was not very high and the pH values from the adsorption and ﬂocculation experiments were in the same range, the inﬂuence of the pH drift is considered to be low. After the sedimentation phase, the supernatant was analyzed for its total P concentration.
2.5. Analytics At the time the experiments were conducted there was not yet an established phosphonate analysis method feasible for application on wastewater with very high Ca concentrations and organic loads. Thus, the surrogate parameter total P was analyzed instead according to ISO 6878 (molybdenum blue method) applying a one-hour peroxodisulphate/sulfuric acid digestion. In the experiments with pure water, phosphonates were the only phosphorus-containing compounds contained in the solutions (very pure phosphonate samples, no phosphate impurity). This is why the concentration (c) of the phosphonate could be calculated from the measured total P concentration by simple conversion using the molar mass (M) of the phosphonate and the total P (e.g. c[PBTC] = M[PBTC]/M[P]·c[P]). The determination of o-PO43− was carried out according to ISO 6878  as well. Turbid and colored samples were additionally treated with compensation solution. The spectral absorbance at 880 nm wavelength of these solutions was subtracted from the extinction of the actual dye solution that had been distorted by turbidity and color. All glass materials that came in contact with the sample were rinsed with hydrochloric acid and pure water in advance. The extinctions were measured using the UV/VIS spectrophotometer JASCO V-550. The COD was measured using the Hach Lange cuvette rapid tests LCK 414 (5–60 mg/L O2), LCK 314 (15–150 mg/L O2), LCK 614 (50–300 mg/L O2), and LCK 514 (100–2000 mg/L O2). The general water hardness (dGH) was determined using the Hach Lange cuvette rapid test LCK 327 (1–20 dGH). The pH value was determined with the WTW pH electrode SenTix 81 in combination with the instrument WTW pH91.
III 2.4.3. FeIII , Al , NaOH and Ca(OH)2 applied to wastewater without pH correction (Exp. 2 and 4) In several raw wastewater samples, the desired concentrations of FeIII and AlIII (Exp. 2) or NaOH and Ca(OH)2 (Exp. 4) were adjusted (concentrates: n = 2, organically polluted wastewaters: n = 1). Each sample was stirred for 15 min. During that period, the pH was continuously monitored. As soon as the pH was stable (mostly within 1 min), the pH value was noted. A sedimentation phase followed. The supernatants were analyzed for total P, COD, o-PO43−, or water hardness.
2.4.4. FeIII and AlIII applied to wastewater at ﬁxed pH (Exp. 3) In several raw wastewater samples, those concentrations of FeIII and AlIII were adjusted to levels that had caused at least 80% total P decrease in Exp. 2 (n = 1). Furthermore, the paper production wastewater was treated with an FeIII and AlIII concentration at which a clear inhibition of the ﬂock formation had occurred in Exp. 2 (n = 1). After the dosage of FeIII and AlIII , the pH values were ﬁxed at between 1.5 and
3. Results and discussion 3.1. FeIII and AlIII applied to pure water (Exp. 1) Fig. 3 shows the results of the adsorption experiment and the ﬂocculation experiment with pure water. The same measurement 191
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
(299.05 g/mol, 3) > EDTMP (436.12 g/mol, 4) > DTPMP (573.20 g/ mol, 5). Nowack and Stone [33,35] and Klinger et al.  had similar ﬁndings in their experiments. However, PBTC cannot be arranged consistently into this pattern. Although this phosphonate contains only one phosphonic acid group, it has a signiﬁcantly larger molar mass (270.13 g/mol) than HEDP (206.03 g/mol), which contains two phosphonic acid groups. Furthermore, PBTC has a very diﬀerent chemical structure compared to the polyphosphonates since, as the only phosphonate, it contains three carboxylate groups. The trend deviation becomes especially apparent when looking at the ß value necessary for the PBTC removal, which is signiﬁcantly higher than the ones for the polyphosphonates. Due to the relatively high molar mass of PBTC, the phosphorus content in the molecule is only 11.5%. For the other phosphonates studied, however, the phosphorus content is between 27.0 and 31.1%. The more phosphonic acid groups a phosphonate contains, the lower the β value compared to the required ﬂocculantphosphonate ratio for a signiﬁcant removal. Thus, for example, ﬁve phosphorus atoms are eliminated simultaneously when only one DTPMP molecule gets adsorbed (accordingly, β = c(FeIII)/c(DTPMP)/ 5). In comparison, for HEDP, β is c(FeIII)/c(HEDP)/2.
values are compared once in relation to the molar metal phosphonate ratio (top) and once in relation to the ß value (bottom). In both experiments, all investigated phosphonates underwent an at least 80% decrease through adsorption in FeIII and AlIII solutions. However, very high ß values were needed, depending on the experiment and the investigated phosphonate. The elimination behavior of phosphonates in the adsorption experiment diﬀered from the behavior in the ﬂocculation experiment signiﬁcantly. In the adsorption experiment, the polyphosphonate concentration decreased quasi-linearly up to a dosage of ßFe = 10 (ßFe = 30 for PBTC) and ﬂattened at dosages above ßFe = 10 (ßFe = 30 for PBTC). In the case of aluminum, a similar development occurred with a stronger concentration decrease at ßAl < 10 (ßAl < 20 for PBTC). In the ﬂocculation experiment, a lack of phosphonate removal in the low dosage concentration range appeared with both ﬂocculants. Thus, depending on the phosphonate, a FeIII surplus between 10 and 40 [c(FeIII)/c(phosphonate)] had to be dosed in order to cause the formation of ﬂocks. In the case of AlIII, a 5–20-fold surplus [c(AlIII)/c(phosphonate)] was required. Interestingly, for both ﬂocculants only a very small concentration beyond that was suﬃcient for total phosphonate removal. Klinger et al.  found a similar elimination behavior in their experiments (initial phosphonate concentrations of 50 μg/L) with much higher FeIII surpluses necessary (PBTC: 50-fold, EDTMP: 80-fold). They ascribed the lack of elimination in the low dosage concentration range to complexation of the ﬂocculant. These signiﬁcantly higher-required dosage concentrations, compared to our experiments, indicate that phosphonate removal using the ﬂocculation method is inferior when phosphonates are present in small concentrations such as in the inﬂuent of wastewater treatment plants. This points out that a pretreatment of higher concentrated wastewaters from indirect dischargers is necessary. For all polyphosphonates, aluminum turned out to be the ﬂocculant with the higher eﬀectiveness in both experiments. In the adsorption experiment, an 80% polyphosphonate decrease was achieved at ßFe = 15 (ßFe = 50 for PBTC), while with aluminum a signiﬁcantly smaller ßAl value of 7 (ßFe = 25 for PBTC) was suﬃcient. Furthermore, in the ﬂocculation experiment for a similar polyphosphonate decrease, the ß values were ßFe = 6–10 (ßFe = 15 for PBTC) and ßAl = 4 (ßFe = 15 for PBTC). A greater speciﬁc surface of the aluminum hydroxide precipitate compared to the iron hydroxide precipitate could be the explanation for this observation . In the ﬂocculation experiment, the phosphonates were removed almost completely with signiﬁcantly lower ﬂocculant concentrations when compared to the adsorption experiment, despite the complexation in the low dosage concentration range. These observations had also been made by Fettig et al. . Next to the pure adsorption, the authors assumed an additional mechanism during the phosphonate elimination in the ﬂocculation process. Taking into account reactions of phosphonates with metal ions or positively charged metal-hydroxo complexes, which occur simultaneously with the precipitation of metal hydroxides, the authors assumed that reaction products were either partially incorporated into the ﬂocks or contributed to a better phosphonate attachment onto the ﬂock surface. Furthermore, an additional eﬀect can play an important role. Phosphonates are known for their ability to incorporate into crystalline structures and thus to weaken their growth. However, in the adsorption experiment the phosphonate was added after the ﬂocks had already formed, so such incorporation into the ﬂocks was more diﬃcult due to steric factors. Thus, phosphonates could mainly attach to the external surface of the ﬂocks. The results of the adsorption experiment and the ﬂocculation experiment show, for both ﬂocculants, a dependency of the adsorption aﬃnity on the molar mass or the number of phosphonic acid groups of the respective phosphonate (a high adsorption aﬃnity corresponds to a low ﬂocculant-phosphonate ratio for an eﬃcient phosphonate removal). Thus, the adsorption aﬃnity behaves according to the following pattern: HEDP (206.03 g/mol, 2 phosphonic acid groups) > NTMP
3.2. FeIII and AlIII applied to wastewater without pH correction (Exp. 2) Fig. 4 summarizes the results of an experiment with four types of wastewaters. In this experiment, diﬀerent concentrations of the ﬂocculants FeIII and AlIII were dosed to the samples. Subsequently, the supernatant was analyzed without prior pH adjustment. In all wastewaters, a total P decrease of 90% or more could be achieved. However, very diﬀerent ß values ranging from 0.6 to 86 were required depending on the wastewater. A signiﬁcantly better total P removal with AlIII compared against FeIII was not observed for all wastewaters as observed in Exp. 1 and other studies [16,31]. Furthermore, a very diﬀerent elimination behavior of total P could be noticed between concentrates and organically polluted wastewaters. In both concentrates, total P and o-PO43−-P were removed to a similar extent by both ﬂocculants. For an 80% total P decrease in membrane concentrate, a ß value of 4 was required for FeIII and AlIII, while an 80% total P decrease from cooling tower wastewater containing at least 3 mg/L PBTC (0.5 mg/L total P) was reached at ß values between 6 and 7 with both metal salts. Even with increasing ß values (> 10), the total P concentration did not drop below 0.1 mg/L (MC) and 0.05 mg/L (CT), resulting in a maximum decrease level of 90% for both concentrates. After the dosage of the ﬂocculants, the pH in the concentrates (MC and CT) only decreased slightly (membrane concentrate [MC]: from 7.9 to 7.1; cooling tower wastewater [CT]: from 7.4 to 6.3) and therefore remained in a favorable range for the formation of metal hydroxides. Possible matrix eﬀects (e.g. disturbance of the crystal growth of the aluminum ﬂocks or improvement of the crystal growth of the iron ﬂocks by organic or inorganic wastewater constituents) which cause the β values to be similar for both ﬂocculants cannot be explained without deeper investigations. From the known antiscalant dosage concentration in the membrane plant (ca. 0.6 mg/L DTPMP in the feed) and the known yield of 80%, the DTPMP concentration in the membrane concentrate could be calculated to be approximately 3 mg/L. A comparison with the ﬂocculation experiment from Experiment 1 is therefore possible. Here, for an 80% decrease of 3 mg/L DTPMP with AlIII, a ß value of 4 was also needed, while with FeIII a higher ß value of 8–10 was required. The cooling tower wastewater had been spiked with 3 mg/L PBTC prior to Exp. 2. Thus, it is reasonable to compare the results from Exp. 2 with those from Exp. 1 regarding the cooling tower wastewater as well. For an 80% decrease of 3 mg/L PBTC in the ﬂocculation experiment (Exp. 1), much higher ß values between 10 and 15 than in Exp. 2 with CT were required independent of the ﬂocculant. Thus, for both concentrates, the total P elimination was achieved with similar or even smaller ﬂocculant concentrations compared to the experiments 192
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
Fig. 4. Concentrations of total P and COD in the supernatants of diﬀerent wastewaters after ﬂocculation/precipitation as well as pH during the stirring phase of 15 min depending on the ﬂocculant concentration (raw samples: MC: 1.1 mg/L P0; CT: 0.5 mg/L P0; PA: 1.1 mg/L P0, 1.0 g/L COD; PR: 400 mg/L P0, 4.8 g/L COD).
than this, a signiﬁcantly stronger pH decrease could be observed with FeIII in comparison to AlIII. While with FeIII the pH steadily diminished down to values of 2.5 in PA (highest ß value tested: 230) and 2.9 in PR (highest ß value tested: 1.4), even with the highest AlIII concentrations, regardless of the wastewater, the pH leveled at minimum values between 4.1 and 4.5. In that pH range, AlIII does not react totally to Al(OH)3 (Ksp = 2.0·10−32), while for that purpose for FeIII pH values in the pH range 2.0–2.5 (Ksp = 1.6·10−39) are required . Thus, when as a result of a high ﬂocculant concentration a pH near the solubility limit of the metal hydroxide was reached with higher dosage concentrations, the dosed metal merely was enriched as a dissolved component in the sample.
with pure water. A matrix loaded with high concentrations of alkaline earth metals, as it is the case for concentrates, can contribute to the absence of inhibiting ﬂocculant-phosphonate complex formations [16,32]. Furthermore, due to ternary complexes the adsorption aﬃnity of phosphonates is augmented . In both organically polluted wastewaters, the formation of ﬂocks, and thus a signiﬁcant COD and total P removal, only occurred above a certain ﬂocculant concentration (PA: ßFe = 86, ßAl = 54; PR: ßFe = ßAl = 0.6–0.7). Below that concentration, the dosed ﬂocculant was enriched homogeneously in the sample, most likely due to complex formation. Above this concentration, the COD was eliminated almost constantly with increasing ﬂocculant concentrations up to a maximum of 55% by FeIII and AlIII. In contrast, the total P could be removed almost entirely from both wastewaters above this ﬂocculant concentration. The extremely high ß values necessary for the removal of total P in paper production wastewater are prominent. The required high ﬂocculant concentration cannot only be ascribed to the complex formation of DTPMP with the ﬂocculant. Paper machine wastewaters can contain diﬀerent substances such as glues, dyes, cellulose, tannins, and lignins. Sundin  for instance, observed that lignins, phenolic macromolecules, could also only be removed above a certain AlIII dosage concentration. Lignins, amongst others, have functional groups such as catechol (1,2-dihydroxybenzene) which are capable of complex formations with polyvalent metal cations . Hence, in each individual case, it is mandatory to examine in previous experiments whether the ß value can be used as the decisive factor for ﬂocculant dosage or not. In the organically polluted PA and PR samples, for AlIII with concentrations slightly higher than those to cause ﬂock formation, even a gradual decrease of the total P removal could be observed. In order to understand this observation, the strong pH decrease due to the ﬂocculant dosage has to be considered. For both ﬂocculants, a formation of ﬂocks occurred at pH values below 5.5–6.0 with PA and below 5.0 with PR. With dosage concentrations resulting in lower pH values
3.3. FeIII and AlIII applied to wastewater at ﬁxed pH (Exp. 3) In Experiment 3, the supernatant was analyzed after the ﬂocculant dosage, pH adjustment, and sedimentation. To each wastewater (the same raw samples as in Exp. 2), a ﬂocculant concentration that had caused a total P decrease of > 80% in Exp. 2 was added (MC: ßFe = ßAl = 5; CT: ßFe = ßAl = 5; PA: ßAl = 83, ßFe = 101; PR: ßAl = 0.74, ßFe = 0.92). For paper production wastewater, an additional test series was conducted with a concentration that had caused no formation of ﬂocks in Exp. 2 (ßAl = 21, ßFe = 31). Fig. 5 shows the measured total P and COD concentrations. Furthermore, in Fig. 5 those pH values are highlighted that had occurred in Exp. 2 without pH correction at the corresponding ﬂocculant concentration. The results show that for both concentrates (MC and CT), FeIII had a good elimination eﬀect at pH values < 7. Below pH 5.5, total P and o-PO43− could be found in relatively high concentrations after the treatment with AlIII. For both ﬂocculant agents, the pH values between 5.5 and 7.0 as well as those exceeding 10 led to the lowest P concentrations. Between pH 7.0 and 9.5, the total P concentration remained at a high level. In order to explain this behavior, the following has to be considered: in samples containing phosphonates, iron, and aluminum, the following 193
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
Fig. 5. Concentrations of total P and COD in the supernatant of diﬀerent wastewaters after treatment with FeIII and AlIII in comparison to the pH ﬁxed with H2SO4 and NaOH (raw samples: MC: 1.1 mg/L P0; CT: 0.5 mg/L P0; PA: 1.1 mg/L P0, 1.0 g/L COD; PR: 400 mg/L P0, 4.8 g/L COD). * At the given ß value this pH occurs when no pH adjustment is made.
interacting processes are of great signiﬁcance (Fig. 6): (1) FeIII signiﬁcantly begins to precipitate above pH values of ca. 1.7 (Ksp[Fe (OH)3] = 1.6·10−39); AlIII begins to precipitate above pH values of ca. 4.0 (Ksp[Al(OH)3] = 2.0·10−32); Al(OH)3 gets transferred to soluble [Al (OH)4]−above pH 8.6 (K = 1.1·1033) . (2) The metal hydroxides serve as an adsorbent for phosphonates; the adsorption aﬃnity decreases with increasing pH values, with an increasing number of CeP bonds, in the presence of other anions such as o-PO43− and with a decreasing water hardness [33–36]. (3) Phosphonates/complexing agents can form complexes with the metal ions (preferably in the neutral/alkaline pH range) and disrupt the precipitation of metal hydroxides (Exp. 1). (4) Furthermore, in samples with high alkaline earth metal and salt concentrations at high pH values, more precipitation products such as Ca(OH)2 (Ksp = 1.3·10−6 ) and CaCO3 occur, which can serve as adsorbents for phosphonates [37,41]. In the pH range below 5, AlIII did not precipitate entirely, so total P and o-PO43− were poorly removed due to missing adsorbent. In contrast, there was still suﬃcient iron hydroxide in this pH range, resulting in a good phosphorus removal. With pH values between 7.0 and 9.5, in cooling tower wastewater for both ﬂocculants, a decreased total P elimination could be observed. This is in compliance with the decreasing adsorption aﬃnity of phosphonates to iron-containing and aluminum-containing solid substances with increasing pH values [33,36]. Furthermore, an interfering eﬀect by metal-phosphonate complexes in the pH range 7.0–9.5 might play an important role. In this pH range, PBTC, which had a concentration of at least 3 mg/L in the cooling tower wastewater, predominantly exists as the HL4− species that is capable of complex formation. However, in both concentrates (MC and CT), the ﬂocculant was added in such a high stoichiometric abundance compared to the phosphonate [c(ﬂocculant)/c(phosphonate) > 7], that metal hydroxides could hypothetically precipitate throughout the whole pH range. The high rate of total P reduction at pH values above 10 can be ascribed to the additional precipitation of more precipitation products containing alkaline earth metals (mainly CaCO3) and, therefore, the resulting increase of adsorbent in the sample.
Fig. 6. Processes inﬂuencing the adsorption behavior of phosphonates during ﬂocculation depending on the pH (based on [16,33–36,41]).
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
concentrate in which diﬀerent concentrations of NaOH or Ca(OH)2 were applied to the sample. Even low concentrations of Ca(OH)2 and NaOH in the mmol/L range led to the precipitation of white ﬂocks and the removal of total P and o-PO43−-P. With Ca(OH)2, higher removal rates of total P and o-PO43−-P could be achieved throughout the whole dosage range. Already with a Ca(OH)2 concentration of 4 mmol/L (pH 8.0) a total P decrease of > 85% was achieved, resulting in residual total P concentrations of 0.1–0.2 mg/L, whereas NaOH had to be dosed at 10 mmol/L (pH 8.3) in order to achieve a residual total P concentration of approximately 0.4 mg/L (75% removal). With higher NaOH dosage concentrations, the residual total P concentration had an insigniﬁcant decrease. Using Ca(OH)2, even with the highest dosage concentrations, merely a weak pH increase of up to 8.2 could be observed. In Fig. 7 (top) a gray dashed line describes the pH that would occur in water without any buﬀering capacity due to the same OHe concentrations applied in the experiment. The large diﬀerence of the hypothetical value compared with the measured value indicates an excellent buﬀering capacity of the wastewater. Since CO32− is formed at pH values above 8 (Kb = 1.8·10−4 ), most likely CaCO3 was the predominant precipitate. The very high water hardness of membrane concentrates causes considerable problems for operators of membrane plants. The initial water hardness of 112 dGH in the membrane concentrate sample investigated here decreased gradually with increasing NaOH and Ca(OH)2 dosage concentrations.
The behavior of phosphorus elimination in the acidic pH range from organically polluted phosphonate production wastewater was similar to that seen for the concentrates. Thus, in organically polluted wastewaters, a poorer elimination by AlIII compared to FeIII could be noticed as well. While with both concentrates a ﬂock formation could be observed in the neutral and alkaline range, absolutely no precipitation occurred in phosphonate production wastewater above pH 7. The absence of precipitation in the organically polluted wastewater eventually led to a lack of the total P and COD removal. This can be explained by the fact that the eﬀectiveness of complexing agents increases with higher pH values. Complexing agents were supposedly present in such high concentrations that a formation of ﬂocks and the adsorption of organic phosphorus compounds could not occur at higher pH values at all. The use of two ﬂocculant concentrations in the experiment with paper production wastewater allows a diﬀerentiated view. While, without pH adjustment, ß values of 21 for AlIII and 31 for FeIII did not cause any total P decrease (Exp. 2), yet, with an additional acidiﬁcation, sludge was formed at these ﬂocculant concentrations. This sludge served as an adsorbent for organophosphorus compounds. Thus, by acidiﬁcation, the added metal could be withdrawn from its complex formation. However, with too much acidiﬁcation, a reduction of the sludge formation occurred. With ﬂocculant concentrations that had caused an elimination of total P even without pH correction (ßAl = 83, ßFe = 101), a low pH value was reached solely because of the acidic eﬀect of the ﬂocculant. When the pH at this ﬂocculant concentration was increased, no decrease in the total P removal could be observed for both ﬂocculants. At this ﬂocculant concentration, the metal was present in such a high abundance that even with the increase of the complexing agent’s eﬀectiveness, the precipitation of the metal hydroxide could not be disrupted. Nevertheless, an excessive surplus of ﬂocculant results in a high sludge production.
4. Conclusions Phosphonates can be removed from any type of wastewater using FeIII and AlIII, but the required ß values can diﬀer strongly, especially in the case of organically polluted wastewaters (ß = 0.6–86). In contrast, for typically hard wastewaters such as membrane concentrate and cooling eﬄuent, the ß values required for an 80% total P removal were quite similar (ß = 4–7). Furthermore, a very important factor is the water hardness, which is why hard concentrates and soft organically polluted wastewaters diﬀer vastly in their total P elimination behavior. Additionally, the adsorption aﬃnity of phosphonates decreases with the number of CeP bonds of the phosphonate. The total P removal from the wastewater can thus be simpliﬁed by an appropriate choice of phosphonate used in production. The decreasing removal eﬃciency with increasing AlIII concentrations in organically polluted wastewaters illustrates a big disadvantage of the ﬂocculant AlIII with regard to its use in wastewater treatment plants targeting the removal of phosphonates. The very small usable concentration range (below: no ﬂock formation, above: decreasing P elimination) restricts ﬂexibility to ﬂuctuating wastewater compositions during the day-to-day plant operation. Thus, with strongly varying wastewater compositions, operators of ﬂocculation plants risk causing overdosages of AlIII. The deterioration of the total P elimination is synonymous with the deterioration of the phosphonate elimination. Consequently, in a subsequent neutralization step, the phosphonate and other complexing agents still dissolved in the sample can disrupt the desired precipitation of the enriched AlIII. Due to much lower pH values necessary to cause the dissolution of iron hydroxide, the risk of reaching overdoses in the daily plant operation with the ﬂocculant FeIII is much smaller. The fact that in three out of four types of wastewaters the ß values required for a successful total P removal were similar for both ﬂocculants (FeIII and AlIII) shows that FeIII is the approriate ﬂocculant for industrial wastewaters. The results suggest that, in treatment plants targeting a selective elimination of phosphonates from industrial wastewater, solely an acidiﬁcation and no alkalization should be considered in order to prevent the ﬂocculant from complexation. This can result in pH values that do not meet the legal limits, so that a pH neutralization becomes necessary. This pH neutralization should only be applied to the supernatant or ﬁltrate that has been separated from the solid substances at
3.4. NaOH and Ca(OH)2 applied to membrane concentrate (Exp. 4) Fig. 7 summarizes the results of an experiment with membrane
Fig. 7. Concentration of total P and water hardness in the supernatant of membrane concentrate after the treatment with NaOH and Ca(OH)2 as well as pH during the stirring phase of 15 min depending on the OHe concentration (raw sample: 1.52 mg/L P0, 0.58 mg/L o-PO43−-P, pH 7.8, 112 dGH).
Journal of Water Process Engineering 17 (2017) 188–196
E. Rott et al.
the low ﬂocculation pH value. Furthermore, regardless of the sludge separation technique (sedimentation or ﬁltration), the dispersing eﬀectiveness of phosphonates should always be considered [26,27]. Because the total P removal and the reduction of water hardness occurred simultaneously using Ca(OH)2, this method proved to be a good alternative for the treatment of membrane concentrates. For organically polluted wastewaters this method is not recommended since the pH values required are too high.
8) (2004).  F. Bordas, A.C.M. Bourg, Eﬀect of complexing agents (EDTA and ATMP) on the remobilization of heavy metals from a polluted river sediment, Aquat. Geochem. 4 (2) (1998) 201–214, http://dx.doi.org/10.1023/A:1009620513655.  B. Horstmann, A. Grohmann, Investigations about the biological degradation of phosphonates (Untersuchungen zur biologischen Abbaubarkeit von Phosphonaten), Vom Wasser 70 (1988) 163–178.  H. Bachus, Complexing agents in degradation tests and in reality (Komplexbildner in Abbautests und in der Realität), Textilveredlung, Melliand Textilber. 3 (2003) 202–204.  J. Steber, P. Wierich, Properties of hydroxyethane diphosphonate aﬀecting its environmental fate: degradability, sludge adsorption, mobility in soils, and bioconcentration, Chemosphere 15 (7) (1986) 929–945, http://dx.doi.org/10. 1016/0045-6535(86)90058-5.  J. Steber, P. Wierich, Properties of aminotris(methylenephosphonate) aﬀecting its environmental fate: degradability, sludge adsorption, mobility in soils, and bioconcentration, Chemosphere 16 (6) (1987) 1323–1337, http://dx.doi.org/10. 1016/0045-6535(87)90069-5.  B. Nowack, The behavior of phosphonates in wastewater treatment plants of Switzerland, Water Res. 32 (4) (1998) 1271–1279, http://dx.doi.org/10.1016/ s0043-1354(97)00338-2.  B. Nowack, Determination of phosphonates in natural waters by ion-pair highperformance liquid chromatography, J. Chromatogr. A 773 (1–2) (1997) 139–146, http://dx.doi.org/10.1016/s0021-9673(97)00202-1.  B. Horstmann, A. Grohmann, The inﬂuence of phosphonates on the phosphate elimination 1. Part: 1-hydroxyethane-1,1-diphosphonic acid (HEDP) (Der Einﬂuß von Phosphonaten auf die Phosphatelimination 1. Teil: 1-Hydroxyethan-1,1diphosphonsäure (HEDP)), Zeitschrift für Wasser- und Abwasser-Forschung 17 (1984) 177–181.  B. Horstmann, A. Grohmann, The inﬂuence of phosphonates on the phosphate elimination 2. Part: 2-phosphonobutane-1,2,4-tricarboxylic acid (PBTC) aminotri (methylenephosphonic acid) (ATMP) ethylenediaminetetra(methylenephosphonic acid) (EDTMP) (Der Einﬂuß von Phosphonaten auf die Phosphatelimination 2. Teil: 2-Phosphono-butan-1,2,4-tricarbonsäure (PBTC) Aminotri (methylenphosphonsäure) (ATMP) Ethylendiamintetra(methylenphosphonsäure) (EDTMP)), Zeitschrift für Wasser- und Abwasser-Forschung 19 (1986) 236–240.  H. Wenger-Oehn, S. Salzmann, J. Gaugg-Salzmann, Experiences with a biological treatment plant for wastewater from the beverage industry (Erfahrungen mit einer biologischen Reinigungsanlage für Abwasser aus der Getränkeindustrie), KA–Abwasser, Abfall 52 (11) (2005) 1252–1257.  H. Raschke, H.G. Rast, R. Kleinstück, H. Sicius, Utilization of 2-phosphonobutane1,2,4-tricarboxylic acid as source of phosphorus by environmental bacterial isolates, Chemosphere 29 (1) (1994) 81–88, http://dx.doi.org/10.1016/00456535(94)90092-2.  K. Fischer, Sorption of chelating agents (HEDP and NTA) onto mineral phases and sediments in aquatic model systems, part II: Sorption onto sediments and sewage sludges, Chemosphere 24 (1) (1992) 51–62, http://dx.doi.org/10.1016/00456535(92)90566-a.  S. Held, About the environmental behavior of complexing agents on the basis of phosphonic acids (Zum Umweltverhalten von Komplexbildnern auf Phosphonsäurebasis), Textilveredlung 24 (11) (1989) 394–398.  J. Fettig, J. Seydel, C. Steinert, Removal of phosphonates and polyphosphates in the coagulation process, Chemical Water and Wastewater Treatment VI, 279–288, Proceedings of the 9th Gothenburg Symposium, October 02–04, 2000, Istanbul, Turkey, 2000, http://dx.doi.org/10.1007/978-3-642-59791-6_26.  B. Nowack, A.T. Stone, Adsorption of phosphonates onto goethite-water interface, J. Colloid Interface Sci. 214 (1) (1999) 20–30, http://dx.doi.org/10.1006/jcis. 1999.6111.  B. Nowack, A.T. Stone, The inﬂuence of metal ions on the adsorption of phosphonates onto goethite, Environ. Sci. Technol. 33 (22) (1999) 3627–3633, http://dx.doi.org/10.1021/es9900860.  B. Nowack, A.T. Stone, Competitive adsorption of phosphate and phosphonates onto goethite, Water Res. 40 (11) (2006) 2201–2209, http://dx.doi.org/10.1016/j. watres.2006.03.018.  M.C. Zenobi, E.H. Rueda, Ternary surface complex: coadsorption of Cu2+, Zn2+, Cd2+ and nitrilotris(methylene phosphonic) acid onto boehmite, Química Nova 35 (3) (2012) 505–509.  G. Metzner, H.D. Nägerl, Environmental behavior of two water conditioning agents based on phosphonates and polyacrylates (Umweltverhalten zweier Wasserkonditionierungsmittel auf Phosphonat- und Polyacrylatbasis), Tenside Deterg. 19 (1) (1982) 23–29.  ISO 6878:2004, Water quality–determination of phosphorus−ammonium molybdate spectrometric method.  J. Sundin, Precipitation of Kraft Lignin Under Alkaline Conditions, Doctoral Thesis, Department of Pulp and Paper Chemistry and Technology, Royal Institute of Technology, Stockholm, 2000.  P. Kelter, M. Mosher, A. Scott, Chemistry the Practical Science, Houghton Miﬄin Company, U.S.A, 0-547-05393-2, 2009.  E. Frigge, E. Jackwerth, Preconcentration and determination of organophosphonic acids: application to natural waters, Anal. Chim. Acta 254 (1–2) (1991) 65–73, http://dx.doi.org/10.1016/0003-2670(91)90010-3.
Acknowledgements We are grateful for the ﬁnancial support by the Willy-HagerStiftung, Stuttgart. We would also like to thank the employees of Zschimmer & Schwarz Mohsdorf GmbH & Co. KG for providing phosphonate samples. References  W.E. Gledhill, T.C.J. Feijtel, Environmental properties and safety assessment of organic phosphonates used for detergent and water treatment applications, in: O. Hutzinger, N.T. de Oude (Eds.), The Handbook of Environmental Chemistry 3(F), Anthropogenic Compounds–Detergents, Springer-Verlag, Berlin, Heidelberg, 1992, pp. 261–285.  T.P. Knepper, Synthetic chelating agents and compounds exhibiting complexing properties in the aquatic environment, Trends Anal. Chem. 22 (10) (2003) 708–724, http://dx.doi.org/10.1016/s0165-9936(03)01008-2.  B. Davenport, A. DeBoo, F. Dubois, A. Kishi, CEH Report: Chelating Agents, SRI, Consulting, Menlo Park, CA, USA, 2000 (cited by Nowack 2003).  B. Nowack, Environmental chemistry of phosphonates, Water Res. 37 (11) (2003) 2533–2546, http://dx.doi.org/10.1016/s0043-1354(03)00079-4.  European Phosphonate Association (EPA), Phosphonates in Detergents, EPA Detergent Phosphonates Dossier, (2013).  American Chemical Society, database of https://sciﬁnder.cas.org, 2016 (Accessed 24 February 2016).  B. Nowack, A.T. Stone, Degradation of nitrilotris(methylenephosphonic acid) and related (amino)phosphonate chelating agents in the presence of manganese and molecular oxygen, Environ. Sci. Technol. 34 (22) (2000) 4759–4765, http://dx.doi. org/10.1021/es0000908.  E. Matthijs, N.T. de Oude, M. Bolte, J. Lemaire, Photodegradation of ferric ethylenediaminetetra(methylenephosphonic acid) (EDTMP) in aqueous solution, Water Res. 23 (7) (1989) 845–851, http://dx.doi.org/10.1016/0043-1354(89) 90008-0.  D. Schowanek, W. Verstraete, Phosphonate utilization by bacterial cultures and enrichments from environmental samples, Appl. Environ. Microbiol. 56 (4) (1990) 895–903.  H. Studnik, S. Liebsch, G. Forlani, D. Wieczorek, P. Kafarski, J. Lipok, Amino polyphosphonates–chemical features and practical uses, environmental durability and biodegradation, New Biotechnol. 32 (1) (2015) 1–6, http://dx.doi.org/10. 1016/j.nbt.2014.06.007.  A. Grohmann, B. Horstmann, The Use of Phosphonates Under Environmental Aspects (Der Einsatz Von Phosphonaten Unter Umwelttechnischen Gesichtspunkten), Research Report 102 063 22 UFA-FB 89-018, Umweltbundesamt, Berlin, Germany, 1989.  F. Mañas, L. Peralta, J. Raviolo, H. García Ovando, A. Weyers, L. Ugnia, M. Gonzalez Cid, I. Larripa, N. Gorla, Genotoxicity of AMPA, the environmental metabolite of glyphosate, assessed by the Comet assay and cytogenetic tests, Ecotoxicol. Environ. Saf. 72 (3) (2009) 834–837, http://dx.doi.org/10.1016/j. ecoenv.2008.09.019.  M. Kwiatkowska, B. Huras, B. Bukowska, The eﬀect of metabolites and impurities of glyphosate on human erythrocytes (in vitro), Pestic. Biochem. Physiol. 109 (2014) 34–43, http://dx.doi.org/10.1016/j.pestbp.2014.01.003.  C. Bolognesi, G. Carrasquilla, S. Volpi, K.R. Solomon, J.P. Marshall, Biomonitoring of genotoxic risk in agricultural workers from ﬁve Colombian regions: association to occupational exposure to glyphosate, J. Toxicol. Environ. Health A 72 (2009) 986–997, http://dx.doi.org/10.1080/15287390902929741.  J. Klinger, M. Lang, F. Sacher, H.J. Brauch, D. Maier, E. Worch, Formation of glyphosate and AMPA during ozonation of waters containing ethylenediaminetetra (methylenephosphonic acid), Ozone Sci. Eng. 20 (2) (1998) 99–110, http://dx.doi. org/10.1080/01919519808547279.  J. Klinger, F. Sacher, H.J. Brauch, D. Maier, E. Worch, Behavior of organic phosphonic acids during drinking water puriﬁcation (Verhalten organischer Phosphonsäuren bei der Trinkwasseraufbereitung), Vom Wasser 91 (1998) 15–27.  J. Klinger, F. Sacher, H.J. Brauch, Formation of glyphosate and AMPA during ozonation: a problem for waterworks? (Bildung von Glyphosat und AMPA bei der Ozonung: ein Problem für die Wasserwerke?), Vom Wasser 94 (2000) 149–159.  Human & Environmental Risk Assessment (HERA) on ingredients of European household cleaning products-Phosphonates (CAS 6419-19-8; 2809-21-4; 15827-60-