³¹P NMR Investigations on Roundup Degradation by AOP Procedures

Abstract

The reactions of (N-(PhosphonoMethyl)Glycine) PMG with H₂O₂ in homogenous systems were investigated using ³¹P NMR (Nuclear Magnetic Resonance). These reactions were carried out in two reaction modes: without UV radiation and under UV radiation. The reactions of PMG with H₂O₂ without UV radiation were carried out in two modes: the degradations of PMG (0.1 mmol) by means of 5–10 molar excess of hydrogen dioxide (PMG-H₂O₂ = 1:5 and 1:10) and the degradation of PMG (0.1 mmol) in homogenous Fenton reactions (PMG-H₂O₂-Fe²⁺ = 1:10:0.05 and 1:10:0.1). All reactions were carried out at ambient temperature, at pH 3.5, for 48 h. The reactions of PMG (in Roundup herbicide composition, 12 mmol) with H₂O₂ under UV radiation (254 nm) were carried out using 5 × molar excess of H₂O₂ (60 mmol), in the pH range of 2 ≤ pH ≤ 12, for 6 h. In this mode of PMG oxidation, the splitting of C-P was observed in the ratios dependent on the applied pH of the reaction mixture.

1. Introduction

Glyphosate (N-(PhosphonoMethyl)Glycine) (PMG) is a broad-spectrum systemic herbicide, invented by Franz and brought to market in 1974 by Monsanto under the trade name Roundup [1,2].

Glyphosate, absorbed mainly through foliage, inhibits a plant enzyme —5-enolpyruvylshikimate-3-phosphate (EPSP) synthetase (EC 2.5.1.19)—involved in the bio-synthesis of aromatic amino acids (AAA) substrates in a build-up of plant lignins [3,4,5].

A significant fraction of glyphosate sprayed on crops returns to the soil, and in spite of strong adsorption to soil solids [6], is able to contaminate surface water (runoff and erosion) [7,8,9].

Animal and epidemiological studies published in recent decades point to a potential glyphosate toxicity [10,11,12]. Further, the World Health Organization’s International Agency for Research on Cancer concluded that glyphosate is “probably carcinogenic to humans” [13] and Anifandis et al. [14,15] demonstrated that glyphosate/PMG induce DNA fragmentation.

Therefore, various treatment processes have been investigated to reduce the pesticide concentration in water and to minimize the potential health risks associated with exposure to these chemicals by the consumption of contaminated water [16,17,18,19,20,21,22,23,24,25,26]. Advanced oxidation processes (AOPs) are a key technology to solve pesticide contamination problems during both water and wastewater treatment (

Table 1. Representative applications of advanced oxidation process (AOP) technology in the chemical degradation of N-(PhosphonoMethyl)Glycine (PMG) and aminomethylphosphonic acid (Gly^P).

No	P-Herbicide	AOP Systems	Phosphorus Degradation Products	Analysis	Reference
1	PMG and GlyP	Birnessite (Mn4+ and Mn3+)	H3PO4	VIS (P-Mo Blue)	[20]
2	PMG	Fe(III)(C2O4)nm−/UV (365 nm)	H3PO4	VIS (P-Mo Blue)	[21]
3	PMG	TiO2/UV (312 nm)	H3PO4	HPLC-UV	[22]
4	PMG	O2/TiO2/SiO2/UV (365 nm)	H3PO4	TOC, HPLC	[23]
5	PMG and GlyP	O3 (pH = 6.5 and 10)	H3PO4	TOC, HPLC-FD	[24]
		Photolysis (292 nm)	H3PO4	TOC, HPLC-FD
		TiO2/O2 (292 nm)	H3PO4	TOC, HPLC-FD
6	PMG	H2O2/UV (254 nm)	H3PO4	HPLC-FD	[25]
7	PMG	Birnessite (Mn4+ and Mn3+)	H3PO4 + GlyP + C-Px	31P NMR	[26]

Birnessite (Na_0.3Ca_0.1K_0.1)(Mn⁴⁺, Mn³⁺)₂O₄ × 1.5 H₂O. Fe^(III)(C₂O₄)_n^m− − Fe(C₂O₄)₂⁻ and Fe(C₂O₄)₃³⁻. P_i colorimetrical determination using the phospho-molybdate blue reaction. TOC—total carbon. C-P^x—unidentified compound.

).

The various mechanisms of PMG degradation using AOP technology, according to the literature, were presented by Manassero [25]. These not always coherent results (Figure 1) led us to investigate these process by using ³¹P NMR monitoring of the PMG degradation processes. This ³¹P NMR technique has been applied for the analysis of PMG metabolites and degradation products in only a few earlier research works [26,27,28,29,30,31,32,33].

2. Materials and Methods

2.1. Materials

The phosphonic amino acids used in the studies (

Table 2. Names, abbreviations, and structures of aminophosphonic acids discussed in this work ^a.

Structure			Name	Trivial Name	Abbr.	Synth.
			1-aminoalkylphosphonic acid	phosphono amino acid	AAP
No	R	R1
1	H	H	aminomethylphosphonic acid	phosphono-glycine	GlyP	[35]
	Me	H	(N-methylamino)methyl-phosphonic acid	phosphono-(N-methyl)-glycine	Me-GlyP	[37]
	Me	Me	(N,N-dimethylamino)-methylphosphonic acid	phosphono-(N,N-dimethyl)-glycine	Me2-GlyP

^a Applied names are in accordance with the IUPAC rules, and abbreviations are in agreement with the general rules elaborated by [38,39].

) were synthesized as follows: aminomethylphosphonic acid (Gly^P) was synthesized according to the Soroka method [34,35]; (N-methylamino)methylphosphonic acid (Me-Gly^P) was obtained by the hydrophosphonylation of 1,3,5-trimethylhexahydro-1,3,5-triazine by means of diisopropylphosphite according to Maier [36]; (N,N-dimethylamino)-methyl-phosphonic acid (Me₂-Gly^P) was obtained by the modified Kabachnik-Fields condensation [37,38].

Phosphonomethyl glycine, methylphosphonic acid, 1,3,5-trimethylhexahydro-1,3,5-triazine, catalase and other applied reagents and solvents were purchased from Aldrich. Diisopropylphosphite was purchased from ACROS Organic^TM.

Herbicide Roundup Ultra 170 SL, containing glyphosate-isopropylammonium salt [227.2] (CAS: 38641-94-0; 170 g/L; 15.67%; 0.75 M), and surfactant (CAS not given; 8%) were purchased from Monsanto Europe S.A (Scotts Poland, Warsaw, Poland).

2.2. Synthesis of Aminophosphonic Standards

2.2.1. Synthesis of Aminomethylphosphopnic Acid (Gly^P)

Phosphorus chloride(III) (8.75 mL; 0.10 mol) was added dropwise to a well-stirred mixture of N-(hydroxymethyl)benzamide (synthesized by the hydroxymethylation of benzamide [34]) (15.1 g; 0.10 mol) and anhydrous acetic acid (20 mL), at ambient temperature. The mixture was then refluxed for 1 h, evaporated (25 °C at 10–20 mm Hg for 15 min, and 75 °C at 0.05 mm Hg) to an oily residue and dissolved in hydrochloric acid solution (5 M aq.; 100 mL). The mixture was heated under reflux for 8 h, cooled to room temperature, and the separated benzoic acid was filtered off. The filtrate was evaporated, the residue was dissolved in water (20 mL), then the solution of crude Gly^P was purified on a Dowex 50 W × 8 ion exchange column using water elution. The obtained Gly^P (8.1 g; 0.072 mol; 72%) was homogeneous at ³¹P NMR solutions. Elemental analysis data (determined %/(calculated %)) for CH₆NPO₃ [111.04] C = 10.70 (10.82); H = 5.47 (5.45); N = 12.50 (12.61).

2.2.2. Synthesis of N-Methylaminomethylphosphopnic Acid (Me-Gly^P)

To trimethylhexahydro-s-triazine (1.29 g; 0.01 mol diisopropylphosphite (5.0 g; 0.03 mol), was added and the mixture was heated with stirring to 100–110 °C for 4 h. The reaction mixture was evaporated (25 °C at 10–20 mm Hg for 15 min, and 75 °C at 0.05 mm Hg), then diluted with 5 M HCl (100 mL) and refluxed for 8 h. The hydrolyzate was evaporated to an oily residue, which was dissolved in water (20 mL) and extracted with ethyl ether (20 mL). The aqueous layer was passed through a Dowex 50 W × 8 ion exchange column using water elution. Fractions were collected (molybdate test) and evaporated. The crystalline product was washed with ethanol, filtered, and dried to give 2.0 g (53.3%) of Me-Gly^P, homogeneous ³¹P NMR solutions. Elemental analysis data (determined %/(calculated %)) for C₂H₈NPO₃ [125.06]: C = 19.12 (19.21); H = 6.55 (6.45); N = 11.10 (11.2).

2.2.3. Synthesis of N,N-Dimethylaminomethylphosphopnic Acid (Me₂-Gly^P)

The formaldehyde aqueous solution (37%; d = 1.11 g/mL; 2.9 g; 0.05 mol) was gradually added to a stirred mixture of equimolar quantities of dimethylamine (2 M solution of Me₂NH in methanol; 25 mL; 0.05 mol) and diethyl phosphite (7.8 g; 0.05 mol), keeping the temperature below 85 °C. The reaction mixture was stirred for 30 min, and evaporated (25 °C at 10–20 mm Hg for 15 min, and 75 °C at 0.05 mm Hg) to an oily residue. The residue was dissolved in 5 M HCl (100 mL) and the solution was refluxed for 8 h. The hydrolyzate was evaporated to dryness (60 °C; 10–20 mm), and the residue was passed through a Dowex 50 W × 8 ion exchange column using water elution. The collected fractions (phosphomolybdate assay) were evaporated to dryness giving white crystals of Me₂Gly^P (4.2 g; 60.0%), homogeneous in³¹P NMR solutions. Elemental analysis data (determined %/(calculated %)) for C₃H₁₀NPO₃ [139.09]: C = 25.78 (25.91); H = 7.32 (7.25); N = 9.98 (10.07).

2.2.4. Solutions

• Solution of 0.01 M Fe(II): a sample of FeSO₄ × 7H₂O [278] (28 mg) was dissolved in water (10 mL).
• Solution of 0.02 M Fe(II): a sample of FeSO₄ × 7H₂O [278] (56 mg) was dissolved in water (10 mL).
• Catalase solution: a sample of 10 mg of catalase was dissolved in 50 mL of distilled or deionized water.
• Solution of 2 M H₂SO₄ (in 20% D₂O): samples of 2.5 M H₂SO₄ (20 mL) were diluted to 25 ml with D₂O (5 mL).

2.3. Instruments

³¹P NMR spectra were recorded on a Bruker AC 200 spectrometer operating at 81.01 MHz and on a Bruker Avance III 600 spectrometer operating at 242.9 MHz. ¹H NMR spectra were recorded on a Bruker Avance III 600 spectrometer operating at 600 Hz. Positive chemical shift values of ³¹P were reported for compounds absorbed at lower fields than H₃PO₄.

The pH measurements were performed using a CX-505 multifunction laboratory meter (Elmetron, Zabrze, Poland) equipped with a combination pH electrode EPP-1 (Elmetron, Zabrze, Poland). The chemical degradation of aqueous solutions of the Roundup herbicide formulation in PMG-H₂O₂ and PMG-H₂O₂-UV systems was carried out. Experiments were performed in the reactor shown in Figure 2.

2.4. Degradation of Glyphosate

2.4.1. Degradation of Phosphonomethyl Glycine (PMG)

Samples of PMG (0.1 mmol) were dissolved in appropriate volumes of FeSO₄ solution (0.5 mL of 0.01 M/0.02 M solution of FeSO₄), followed by the addition of dihydrogen dioxide (the exact amounts of the solutions are given in

Table 3. Preparation of the reaction mixtures for PMG-H₂O₂ and PMG-H₂O₂-Fe²⁺ systems.

Exp.	PMG 0.1 mmol	H2O	10 M H2O2		FeSO4
			0.5 mmol	1 mmol	0.01 M	0.02 M
1	17 mg	0.55 mL	0.05 mL	−	−	−
2	17 mg	0.5 mL	−	0.1 mL	−	−
3	17 mg	−	−	0.1 mL	0.5 mL (0.005 mmol)	−
4	17 mg	−	−	0.1 mL	−	0.5 mL (0.01 mmol)

) and kept at room temperature for a desired reaction time. Then, the reaction mixtures were centrifuged, acidified with 2 M H₂SO₄ to pH = 3.5 if necessary, and the volumes of 0.4 mL were taken and mixed with D₂O (0.05 mL) and 0.1 M EDTA (0.05 mL), and analyzed by means of ³¹P NMR.

2.4.2. Degradation of Roundup Herbicide Formulation by Means of AOP Technology

Reaction mixtures were prepared in accordance with

Table 4. Preparation of the reaction mixtures for AOP degradations of PMG in Roundup herbicide.

No	Roundup (0.75 M)		H2O	H2O2 (30%; 10 M)		H2SO4 (2 M)	NaOH (5 M)	pH
	mL	PMG mmoL	mL	mL	mmoL	mL	mL
1	16.0	12.0	3 680	−	−	−	−	4.85
2	16.0	12.0	3 680	6.0	60.0	4.6	−	2.0
3	16.0	12.0	3 680	6.0	60.0	0.5	−	4.0
4	16.0	12.0	3 680	6.0	60.0	−	4.0	6.0
5	16.0	12.0	3 680	6.0	60.0	−	4.8	8.0
6	16.0	12.0	3 680	6.0	60.0	−	6.0	10.0
7	16.0	12.0	3 680	6.0	60.0	−	40.0	12.0

. Thus, samples of PMG contained in Roundup Ultra 170 SL Herbicide solution (0.75 M; 16 mL; 12 mmol of PMG), were diluted in water (3680 mL), dihydrogen dioxide samples (60 mmol) were added, and the reaction mixtures were adjusted to the desired pH value by means of acidification with 2 M H₂SO₄ or alkalized by means of 5 M KOH. The oxidative degradations of herbicide were performed for the desired time, during which the reaction progress was monitored by the ³¹P NMR analysis. Thus, the appropriate samples (4 mL) were treated with catalase (0.1 mL), kept for 30 min at room temperature, and evaporated to an oily residue. These were dissolved in 2 M H₂SO₄ (in 20% D₂O) (0.5 mL) and analyzed using ³¹P NMR.

3. Results and Discussion

3.1. Protonation Equilibria of Reagents

Representative values of pK_a are given in

Table 5. Representative works on pKa determination of PMG.

No.	pK				Method	Reference
	pK1	pK2	pK3	pK4
1	2.0	2.6	5.6	10.6	pH metric titration	[40]
2		2.32	5.86	10.86	pH metric titration	[41]
3	<1	2.0	5.5	10.5	1H and 31P NMR	[42]
4	0.3	2.3	5.6	10.6	1H and 31P NMR	[43]
5		2.09	5.52	10.28	pH metric titration	[44]
6		logβ 9.66 (1.58)	logβ 14.86 (5.20)	log β 16.44 (9.66)	pH metric titration	[45]

. On this basis, protonation equilibria of phosphonomethyl glycine (PMG) are presented in Figure 3.

Despite the large amount of research data, there is conflicting information in the literature concerning the first protonation step of glyphosate, namely whether the first dissociable proton in the HL²⁻ formations is attached to the nitrogen atom of the amino group or to the oxide atom of the phosphonate function. As a matter of fact, only two recent reports consider that the first protonation step occurs on the one of the oxygen atoms in the phosphonate group (Figure 4, magenta arrows) [45,46].

On the basis of this analysis of the speciation graph (Figure 4), we assumed that in aqueous solutions PMG exists in the following forms: at pH = 0—mainly as H₄L⁺ form; at pH = 1.5—mainly as H₃L form; at pH = 3.5–4—mainly as H₂L⁻; at pH = 8—mainly as HL²⁻; and at pH ≥ 12—mainly as L³⁻ form.

The protonation equilibrium of dihydrogen dioxide is shown in Figure 5; dihydrogen dioxide speciation is also presented in Figure 6. In the literature, the pKa values for H₂O₂ are as follows: pK_a1 = −3.1 [48] and pK_a2 = 11.6 [49]. This means that in concentrated H₂SO₄ (e.g., 2 M), dihydrogen dioxide can exist in the H₃L⁺ form, at pH = 0–8 it exists in the molecular form H₂L, at pH = 14 it is dissociated in ca. 50% to HL⁻, and for 2 M KOH (pH > 14) it is almost fully ionized.

3.2. Reaction of PMG and H₂O₂

It is generally known that the oxidation potential of H₂O₂ greatly increases during UV irradiation (Mierzwa et al., 2018, and references cited therein) [50] as well as in the presence of metal ions (Figure 7) [50,51,52,53,54,55].

Therefore, we assumed that the reaction between PMG and dihydrogen peroxide consumed either the molecular form of H₂O₂ in the absence of irradiation, or hydroxide and peroxide radicals during UV irradiation or in the presence of Fenton reagents. The results of PMG degradation in both modes are illustrated in Figure 8.

The reaction of PMG with H₂O₂, both with H₂O₂ and H₂O₂-Fe³⁺, did not occur at the applied pH = 3.5, at which PMG exists mainly in the H₂L⁻ protonated on nitrogen form (Figure 3 and Figure 4) and hydrogen peroxide mainly in H₂L forms (Figure 5 and Figure 6). Therefore, we assumed that the protonation of the amino function in PMG efficiently reduces the interaction of PMG and H₂O₂ (no trace of P-C bond splitting was observed in a 48-h period) (Figure 8). However, during the irradiation of aqueous solutions of PMG (in the form of the herbicide Roundup) and H₂O₂ (1:5), for a pH range of 2 ≤ pH ≤ 12, the splitting of the P-C bond of PMG was observed, to an extent dependent on the pH of the applied solution (Figure 9 and Figure 10). Is worth noting that the irradiation of aqueous PMG solution without H₂O₂ during a 48-h period did not exhibit any sign of PMG decomposition (100% of PMG).

The residual PMG quantities were calculated from the corresponding ³¹P NMR spectra using Equation (1):

$$\mathrm{PMG}=\frac{{\mathrm{S}}_{\left(\mathrm{PMG}\right)}}{{\mathrm{S}}_{\left(\mathrm{PMG}\right)}+{\mathrm{S}}_{\left(\mathrm{R}-\mathrm{P}\right)}+{\mathrm{S}}_{\left(\mathrm{Pi}\right)}}\times 100\%,$$

(1)

where S_(PMG), S_(R−P), and S_(Pi) are the ³¹P signals corresponding to PMG, phosphonic acids, and inorganic phosphate, respectively.

The ³¹P NMR spectra of the degradation mixtures of PMG-H₂O₂-(UV) (UV; 25 °C) recorded for reactions carried out at pH = 2, 8, 10, and 12 are presented in Figure 11. For the identification of the reaction products of PMG-H₂O₂-(UV) mixtures, we recorded the ³¹P NMR spectra of PMG potential degradation products. The chemical shifts (δ(³¹P)) of these compounds are listed in

Table 6. ³¹P Chemical shifts (δ(ppm)) of PMG and its potential degradation products in acidic and basic solutions.

2 M HCl
Comp.	PMG	GlyP	Me-GlyP	Me2-GlyP	Me-PO3H2	H3PO4	H3PO3
δ (31P) (ppm)	10.6	13.9	11.4	9.4	30.7	−0.47	5.15
2 M KOH
Comp.	PMG	GlyP	Me-GlyP	Me2-GlyP	Me-PO3H2	H3PO4	H3PO3
δ (31P) (ppm)	16.3	19.3	16.0	15.0	20.5	5.4	3.2

³¹P δ(ppm): in 2 M HCl solutions (protonated forms of P-acids); in 2 M KOH solutions (deionized forms of P-acids).

.

The results of ³¹P NMR investigations on PMG degradation with H₂O₂ are shown graphically in Figure 12.

4. Conclusions

The data presented suggest the PMG inertness toward H₂O₂ in the modes without UV irradiation, both with (PMG-H₂O₂-Fe²⁺) as well as without Fe²⁺ catalyst (PMG-H₂O₂). The data considering the reaction modes of PMG with H₂O₂ under UV irradiation (PMG- H₂O₂-(UV)) exhibit the slow degradation of PMG at 2 ≤ pH ≤ 10, which becomes faster at pH = 12. The analysis of the ³¹P NMR spectra of PMG-H₂O₂ reaction mixtures obtained for reactions carried out at 2 ≤ pH ≤ 10 indicate the presence of initial PMG and H₃PO₄, and the mixture of PMG, Gly^P, and H₃PO₄/H_xPO₄^3−x for reactions carried out at pH = 12. The results suggest the slow formation of an intermediate PMG × H₂O₂ phase in the first stage of degradation which decomposes very fast (no intermediates were observed in the ³¹P NMR spectra) by the scission of the P-C bond of PMG and the subsequent release of phosphoric acid/phosphate ion.

Recapitulating, in the experiments carried out without UV radiation we observed:

• full stability of PMG in reaction with H₂O₂ (48 h);
• full stability of PMG in reaction with H₂O₂/Fe²⁺ (48 h).

In the experiments carried out with UV radiation (PMG-H₂O₂-(UV)), the P-C rapture type of PMG degradation was observed, the extent of which was dependent on the applied pH of the reaction mixtures. As a result, for the reactions run at 2 ≤ pH ≤ 10, the partial formation of phosphoric acid/phosphate ions (PMG → PMG + H₃PO₄/H_xPO₄^3−x) was observed, whereas for reactions run at pH = 12, mixtures of PMG, Gly^P, and PO₄³⁻ were found.

We did not observe:

• any formation of nitrone-type derivatives (see [56,57,58,59,60]);
• the formation of Me-Gly^P (Sar^P) or Me-P(O)(OH)₂.