Natural soil colloids To retard simazine and 2,4-D leaching in soil

Natural or synthetic sorbents for pesticides can be used to reduce contamination of soils and natural waters. The sorption of simazine and 2,4-D on montmorillonite minerals has been studied and their potential use to retard pesticide leaching in soil evaluated. Simazine and 2,4-D did not sorb on high-layer charge montmorillonite, whereas sorption on the lower layer charge montmorillonite SWy varied depending on the saturating cation. Simazine sorption increased in the order Ca(2+)SWy < K(+)SWy < Fe(3+)SWy. Simazine molecules sorb on hydrophobic microsites of the montmorillonite. Once protonated, further sorption through cation exchange takes place in the interlamellar space of the montmorillonite, as corroborated by X-ray diffraction and FT-IR studies. 2,4-D does not sorb on K(+)SWy or Ca(2+)SWy, but does sorb on Fe(3+)SWy, because the acidic character of this sorbent allows the molecular form of 2, 4-D to sorb by hydrogen bonding and/or by hydrophobic interactions. Leaching experiments in hand-packed soil columns indicate that simazine and 2,4-D application as a complex with FeSWy renders later breakthrough and lower maximum concentration peaks, and the total herbicide leached is lower than when applied as the pure analytical grade compound. These results suggest the possible use of natural soil colloids as sorbents for herbicides such as simazine and 2,4-D to retard pesticide leaching in soil, thus reducing their ground water contamination potential.     

Electron capture gas-liquid chromatographic method for the simultaneous analysis of 2,4-D, dicamba, and mecoprop residues in soil, wheat, and barley.

A method has been developed for the simultaneous analysis of 2,4-D (2,4-dichlorophenoxy-acetic acid), dicamba (2-methoxy-3,6-dichloro-benzoic acid), and mecoprop (MCPP; 2-[(4-chloro-o-tolyl) oxy] propionic acid) residues in soil, wheat, and barley. Soil and crop samples are extracted with acidic acetone and methanol, respectively. The extracts in diethyl ether are esterified with diazomethane and cleaned up by passing through a Florisil column. Extracts are analyzed by gas-liquid chromatography, using an electron capture detector to determine 2,4-D and dicamba residues. Mecoprop in the extract is not detected at low levels of concentration. However, bromination of the extract increases the response of the electron capture detector to mecoprop. The method is sensitive to about 0.05 ppm 2,4-D and dicamba and 0.5 ppm mecoprop. Recoveries of these 3 herbicides added to soil, wheat, and barley samples at 0.05, 0.1, 0.5, and 1.0 ppm levels were between 65 and 93%. The method was used for the simultaneous analysis of 2,4-D, dicamba, and mecoprop residues in wheat, barley, and soil samples obtained from fields sprayed with the herbicide formulation Kil-Mor.     

Behaviour of 2,4-D Herbicide in Coastal Area of Oka River,Russia

Laboratory and field experiments werecarried out with 2,4-D herbicide(2,4-Dichlorophenoxyacetic acid) to evaluate itstransformation and migration in the coastal waterprotection zones of the Oka river, Russia. In thefirst laboratory experiment, the transformation of2,4-D was studied in various soil samples from coastalslopes (1–0°) of 480 m length soil-geochemicalcatena on the right side of the Oka river incomparison with watershed and floodplain soils. Thetransformation of 2,4-D was the lowest in soil sampleswith minimal pH values and was independent of eitherslope values or vicinity to the Oka river channel.Using indirect estimates, the surface runoff potentialwas calculated for this herbicide. In the second fieldexperiment, the vertical migration and transformationof 2,4-D was carried out in soddy sand soil (EutricArenosol) placed in the left side of the Oka river(0-100 cm) under `soft' (40 mm 2 hr<sup>-1</sup>) and `hard'(40 mm 15 min^-1) irrigation regimes. Furthermore, thetransformation of this herbicide was studied in 0–20and 40–50 cm soil layers under various temperature andmoisture regimes. After 1 day of irrigation, the mainherbicide quantity was found in the 0–30 cm layerunder both irrigation regimes. The transformation ofthe herbicide was faster in the surface, 0–20 cmlayer, than in the deeper, 40–50 cm layer.     

Direct aqueous injection liquid chromatography/electrospray ionization-mass spectrometry/mass spectrometry analysis of water for atrazine, simazine, and their chlorotriazine metabolites

A method is reported for the determination of atrazine, simazine, and their respective dealkylated chlorotriazine metabolites in ground, surface, and finished drinking water. Water samples are diluted 1:4 in an injection vial prior to analysis using liquid chromatography/electrospray ionization-mass spectrometry/mass spectrometry (LC/ESI-MS/MS). The lower limit of method validation is 0.10 microg/L (ppb) for 2-chloro-4-(ethylamino)-6-isopropylamino)-s-triazine (atrazine, G-30027), 2-chloro-4, 6-(diethylamino)-s-triazine (simazine, G-27692), 2-amino-4-chloro-6-(isopropylamino)-s-triazine (deethylatrazine, DEA, or G-30033), 2-amino-4-chloro-6-(ethylamino)-s-triazine (deisopropylatrazine, DIA, or G-28279), and 2,4-diamino-6-chloro-s-triazine (didealkylatrazine, DDA, or G-28273). The overall mean procedural recoveries (and % relative standard deviations) for atrazine, simazine, DEA, DIA, and DDA are 98 (4.4), 102 (3.6), 99 (4.8), 103 (4.0), and 109% (4.8%), respectively, in finished drinking water; 108 (2.7), 104 (5.4), 113 (4.5), 111 (5.2), and 105% (5.3%), respectively, in groundwater; and 96 (6.9), 103 (4.2), 102 (4.4), 102 (5.2), and 102% (8.2%), respectively, in surface water. The method validation was conducted under U.S. EPA FIFRA Good Laboratory Practice Guidelines 40 CFR 160.     

Determination of N-nitrosodimethylamine levels in some Canadian 2,4-D amine formulations

Levels of N-nitrosodimethylamine (NDMA) were determined in 112 samples of 2,4-dichlorophenoxyacetic acid, (2,4-D), formulated as the dimethylamine salt, collected over a 2 year period from products on the Canadian market. A sample aliquot is partitioned with dichloromethane, and the co-extracted dimethylamine is removed by cleanup on a silica gel column. The eluates containing NDMA are concentrated, an internal standard of N-nitrosodipropylamine is added, and nitrosamine levels are determined using a gas chromatograph interfaced with a thermal energy analyzer. Recoveries of NDMA and N-nitrosodiethylamine spiked into samples were 103 +/- 16 and 96.3 +/- 9.8%, respectively. Of the 112 samples analyzed, 92 were below 1 part per million (ppm) relative to the amount of 2,4-D in the samples, 16 were between 1 and 5 ppm, and 4 were greater than 5 ppm. The gas chromatographic column used is compared to a conventional packing material for volatile nitrosamine analysis. Formation of NDMA during cleanup and analysis was shown not to occur.     

Denitrification in soil: Effects of herbicides

The effects of 20 herbicides on denitrification of nitrate in three soils were studied by determining the effects of 10 and 50μgg^?1 soil of each herbicide on the amounts of nitrate lost and the amounts of nitrite, N₂O and N₂ produced when soil samples were incubated anaerobically after treatment with nitrate. The herbicides used were butylate, EPTC, chlorpropham, propham, diuron, linuron, monuron, siduron, alachlor, trifluralin, 2,4-D amine, 2,4-D ester, atrazine, cyanazine, metribuzin, simazine, dalapon, chloramben, dicamba and dinoseb.None of the herbicides studied significantly affected denitrification of nitrate when applied at the rate of 10 μg g^?1 soil, but dinoseb increased the ratio of N₂ to N₂O in the gaseous products of denitrification when applied at this rate. Butylate, EPTC, diuron, simazine and dalapon had no significant effect on denitrification when applied at the rate of 50μgg^?1 soil, whereas metribuzin and dinoseb enhanced denitrification when applied at this rate. The influence of the other herbicides on denitrification when applied at the rate of 50μgg^?1soil depended on the soil, but all enhanced or inhibited denitrification in at least one soil.     

Uniformly (14)C-ring-labeled 2,4-dichlorophenoxyacetic acid: a metabolism study in bluegill sunfish, Lepomis macrochirus.

The fate of 2,4-dichlorophenoxyacetic acid (2,4-D), a mixture of [phenyl(U)-(14)C]-2,4-D and unlabeled 2,4-D, in bluegill sunfish was investigated after exposure to approximately 11 ppm under static conditions for 4 days. Total radioactive residues (TRR) in whole fish increased from 0.41 ppm on day 1 to 0.60 ppm on day 3. TRR levels in fillet (edible) and viscera (nonedible) of treated fish on day 4 were 0.41 and 1.9 ppm, respectively. Most residues in both matrices were acetonitrile soluble; small amounts were hexane soluble or unextractable with solvents. Acid and base hydrolyses with ethyl acetate partitioning were used to release the fillet unextractable residues. The identification of 2,4-D and 2,4-dichlorophenol (2,4-DCP) in the fillet was conclusively confirmed by GC-MS analysis. On the basis of the experimental data from this study, a metabolic pathway for 2,4-D in bluegill sunfish in which the 2,4-D is metabolized to 2,4-DCP and conjugates of 2,4-D and 2,4-DCP is proposed.     

Kinetic modeling of 2,4-dichlorophenoxyacetic acid (2,4-D) degradation in soil slurry by anodic fenton treatment

Anodic Fenton treatment (AFT) has been shown to be a promising technology in pesticide wastewater treatment. However, no research has been conducted on the AFT application to contaminated soils. In this study, the 2,4-D degradation kinetics of AFT in a silt loam soil slurry were investigated for the first time, and the effects of various experimental conditions including initial 2,4-D concentration, Fenton reagent delivery rate, amount of humic acid (HA) addition, and pH were examined. The 2,4-D degradation in soil slurry by AFT was found to follow a two-stage kinetic model. During the early stage of AFT (the first 4-5 min), the 2,4-D concentration profile followed a pseudo-first-order kinetic model. In the later stage (typically after 5 or 6 min), the AFT kinetic model provided a better fit. This result is most likely due to the existence of (*)OH scavengers and 2,4-D sorption on soil. The Fe(2+) delivery rate was shown to be a more significant factor in degradation rate than the H(2)O(2) delivery rate when the Fe(2+)/H(2)O(2) ratios were in the range of 1:2 to 1:10. The presence of HA in soil lowered the AFT rate, most probably due to the competition with 2,4-D for consumption of (*)OH and increased sorption of 2,4-D on soil. The optimal pH for 2,4-D degradation in soil slurry by AFT was observed to be in the range of pH 2-3.     

Identification of in situ 2,4-dichlorophenoxyacetic acid-degrading soil microorganisms using DNA-stable isotope probing

Stable isotope probing (SIP) was used to investigate the microorganisms responsible for degradation of the herbicide, 2,4-dichlorophenoxyacetic acid (2,4-D) in soil samples. Soils were unamended or amended with either unlabeled 2,4-D or UL(ring) ¹³C-2,4-D. Degradation of 2,4-D was complete after 17 days, whereas little removal (11±3%) was observed in the sterile controls. Terminal restriction fragment length polymorphism (TRFLP) on soil DNA after 17 days indicated a consistent increase in the relative abundance of one fragment (217 bp in Hae III digests) in soils spiked with 2,4-D (both unlabeled and labeled samples) compared to the unamended soils. DNA extracts from labeled and unlabeled 2,4-D amended soils were subject to ultracentrifugation, fractionation of centrifuged samples, followed by TRFLP on each fraction. TRFLP profiles from ultracentrifugation fractions illustrated that the same fragment experienced an increase in buoyant density (BD) in samples spiked with ¹³C-labeled 2,4-D. This increase in DNA BD indicates the organisms represented by this fragment were responsible for uptake and degradation of the herbicide. 16S rRNA sequencing of the heavy, ¹³C-enriched fraction suggests the organisms belong to the β subdivision of Proteobacteria. Herein, SIP facilitated the identification of unique organisms degrading 2,4-D in soil without the need for isolation and provided more direct evidence for a functional role of these organisms than would have been possible with the molecular-based methods alone.     

Gas chromatographic method for analysis of 2,4-D in wheat: interlaboratory study

A procedure is described for the determination of 2,4-D (2,4-dichlorophenoxyacetic acid) in dried green plant material. Samples are first extracted with dilute sodium hydroxide, and then after acidification and solvent extraction, the residues are methylated using boron trifluoride-methanol reagent. The methyl ester of 2,4-D is cleaned up on a Florisil column and quantitated using a gas chromatograph equipped with an electron capture detector. Six laboratories made quadruplicate determinations on control, dried green wheat check samples, on 4 similar samples fortified at the 0.50 ppm level, and on 4 samples fortified at the 1.00 ppm level with 2,4-D. Based on the data from 5 laboratories, the plant fortifications of 0.50 and 1.00 ppm yielded average interlaboratory recoveries of 2,4-D of 83.3 and 88.2%, respectively. The procedure also has potential for the determination of 2,4-D in wheat straw and wheat grain.     

Blind inlets: conservation practices to reduce herbicide losses from closed depressional areas

Purpose

In a 6-year study, we investigated the effectiveness of blind inlets as a conservation practice in reducing pesticide losses compared to tile risers from two closed farmed depressional areas (potholes) in the US Midwest under a 4-year cropping rotation.

Materials and methods

In two adjacent potholes within the same farm and having similar soils, a conventional tile riser and blind inlet were installed. Each draining practice could be operated independent of each other in order to drain and monitor each depression with either practice. Sampling events (runoff events) were collected from the potholes from 2008 to 2013 using autosamplers. The samples were analyzed for atrazine, metolachlor, 2,4-D, glyphosate, and deethylatrazine.

Results and discussion

The results of this study demonstrated that the blind inlet reduced analyzed pesticide losses; however, the level of reduction was compound dependent: atrazine (57 %), 2,4-D (58 %), metolachlor (53 %), and glyphosate (11 %).

Conclusions

Results from this study corroborate previous research findings that blind inlets are an effective conservation practice to reduce discharge and pollutants, including pesticides from farmed pothole surface runoff in the US Midwest.

     

Soil microbial communities response to herbicide 2,4-dichlorophenoxyacetic acid butyl ester

2,4-Dichlorophenoxyacetic acid butyl ester (2,4-D butyl ester) is extensively applied for weed control in cultivation fields in China, but its effect on soil microbial community remains obscure. This study investigated the microbial response to 2,4-D butyl ester application at different concentrations (CK, 10, 100 and 1000 μg g^?1) in the soils with two fertility levels, using soil dilution plate method and phospholipid fatty acid (PLFA) analysis. Culturable microorganisms were affected by the herbicide in both soils, particularly at the higher concentration. After treating soil with 100 μg g^?1 herbicide, culturable bacteria and actinomycetes were significantly higher, compared to other treatments. Treatment of soil with 1000 μg g^?1 2,4-D butyl ester caused a decline in culturable microbial counts, with the exception of fungal numbers, which increased over the incubation time. PLFA profiles showed that fatty acids for Gram-negative (GN) bacteria, Gram-positive (GP) bacteria, total bacteria and total fungi, as well as total PLFAs, varied with herbicide concentration for both soil samples. As herbicide concentration increased, the GN/GP ratio decreased dramatically in the two soils. The higher stress level was in the treatments with high concentrations of herbicide (1000 μg g^?1) for both soils. Principal component analysis of PLFAs showed that the addition of 2,4-D butyl ester significantly shifted the microbial community structure in the two soils. These results showed that the herbicide 2,4-D butyl ester might have substantial effects on microbial population and microbial community structure in agricultural soils. In particular, the effects of 2,4-D butyl ester were greater in soil with low organic matter and fertility level than in soil with high organic matter and fertility level.     

Quantitative nuclear magnetic resonance (QNMR) spectroscopy for assessing the purity of technical grade agrochemicals: 2,4-dichlorophenoxyacetic acid (2,4-D) and sodium 2,2-dichloropropionate (Dalapon sodium)

Comparison of quantitative NMR spectroscopy (QNMR) with chromatographic methods such as gas chromatography (GC) or high-pressure liquid chromatography (HPLC) for the determination of the purity of and impurities in technical grade agrochemicals, 2,4-dichlorophenoxyacetic acid (2,4-D), 1, and Dalapon sodium (sodium 2,2-dichloropropionate), 10, has revealed that QNMR is more precise and accurate than the chromatographic methods. Quantitative impurity profiling of technical grade 1 is rapid and accurate using 600 MHz (1)H NMR. Extra dispersion at the relatively high frequency allowed full assignment of the NMR spectrum of 1 and its related organic impurities in technical samples. The percentage purity of 1 was measured by the difference QNMR method, which involves summing the amounts of impurities and subtracting from 100%. Results are superior in consistency to those obtained by chromatographic methods. The percentage purity of Dalapon sodium, 10, in technical grade batches is readily obtained by (1)H QNMR, using either the difference method or the internal standard method, using dimethyl sulfone (DMSO2) internally as a reference material, that is chemically unrelated to the analyte. The latter method also allows the simultaneous identification and quantification of impurities, many of which are either not accessible to or detectable by the chromatographic methods. Uncertainty budgets for the QNMR method are presented and demonstrate that the major contributors to uncertainty lie in the weighing of the chemicals and in purity of the standard reference material prior to the QNMR experiment.     

Environmental analysis by on-line immunoextraction and reversed-phase liquid chromatography: optimization of the immunoextraction/RPLC interface

The use of antibodies in HPLC columns for on-line immunoextraction combined with reversed-phase liquid chromatography (RPLC) is of growing interest in environmental and agricultural analysis. This technique is typically performed by using a small RPLC precolumn to capture and concentrate analytes as they elute from the immunoextraction column; however, there is little information on the conditions required for optimizing this interface. This study examined the behavior of this interface by using 2,4-dichlorophenoxyacetic acid (2,4-D) and related herbicides as model analytes. It was found that analyte dissociation from immunoextraction columns followed first-order decay and that the elution of these analytes through the immunoextraction/RPLC interface gave an exponentially modified Gaussian profile. Computer simulations were used to see how analyte elution through the interface changed with different dissociation and retention conditions. Several guidelines were developed from this work that could be used for developing and optimizing on-line immunoextraction/RPLC systems for other chemicals of environmental or agricultural interest.     

Enantiomer resolution and assay of propionic acid-derived herbicides in formulations by using chiral liquid chromatography and achiral gas chromatography

Enantiomers of 6 propionic acid-derived herbicides in the form of their esters were resolved using liquid chromatography with a chiral column. Free acids are converted to methyl esters by means of a BF3-catalyzed reaction. Chromatographic resolutions for 6 of 8 herbicides investigated were in the range of 2 to 4. The method was applied for the simultaneous determination of mecoprop and 2,4-D content and individual mecoprop enantiomers in 2 formulations containing racemic and R-mecoprop in mixture with 2,4-D. Precision and accuracy of content determination was comparable to standard methods, and enantiomer contents were in good agreement with declared values. The enantiomers of dichlorprop and mecoprop were also resolved as diastereomeric menthyl esters by achiral high resolution gas chromatography (HRGC). HRGC data on enantiomer composition were in good agreement with those from the LC method and other data.     

Dissipation of 2,4-D glyphosate and paraquat in river water

Dissipation of herbicides in river water was determined by adding different concentrations of 2,4-D, glyphosate and paraquat to samples of river water. A small variation of dissipation of radioactivity for¹⁴C-2,4-D in higher and lower concentrations and in different samples of river water was found. But about half the radioactivity disappeared from water samples of original glyphosate concentration at 100 mg L^?1 and, in the case of 100 μg L^?1, only 11 to 22% remained in the samples of river water after 56 d incubation except the sample from Hsin-Tien River. More than 80% of paraquat remained in water samples. Determination of octanol-water partition coefficient (Kow) showed a large difference in amounts of 2,4-D partitioned in water phase at different pH values, 97.4% at the higher pH of ionic state and 5.2% at the lower pH of molecular state, implying that pH value of water might affect the bioaccumulation process of 2,4-D. The result showed that 95.0% of glyphosate present in water phase in ionic form (higher pH) and 82.3% in molecular form (lower pH), indicating that glyphosate might have no affect on the biomagnification, since most of glyphosate could be excreted with water by organisms.     

Degradation of [carboxyl-14C] 2,4-D and [ring-U-14C] 2,4-D in 114 agricultural soils as affected by soil organic carbon content

This study compared the degradation of [carboxyl-¹⁴C] 2,4-dichlorophenoxyacetic acid (2,4-D) (C2,4-D) and [ring-U-¹⁴C] 2,4-D (R2,4-D) in 114 agricultural soils (0–15 cm) as affected by 2,4-D sorption and soil properties (organic carbon content, pH, clay content, carbonate content, cation exchange capacity, total microbial activity). The sample area was confined to Alberta, Canada, located 49–60° north longitude and 110–120° west latitude and soils were grouped by soil organic carbon content (SOC) (0–0.99%, 1–1.99%, 2–2.99%, 3–3.99% and >4% SOC). Degradation rates of C2,4-D and R2,4-D followed first-order kinetics in all soils. Although total microbial activity increased with increasing SOC, degradation rates and total degradation of C2,4-D and R2,4-D decreased with increasing SOC because of increased sorption of 2,4-D by soil and reduced bioavailability of 2,4-D and its metabolites. Rates of R2,4-D degradation were more limited by sorption than rates of C2,4-D degradation, possibly because of greater sorption and formation of bound residues of 2,4-D metabolites relative to the 2,4-D parent molecule. Based on the sorption and degradation parameters quantified, there were two distinct groups of soils, those with less than 1% SOC and those with greater than 1% SOC. Specifically, soils with less than 1% SOC had, on average, 2.4 times smaller soil organic carbon sorption coefficients and 1.4 times smaller 2,4-D half-lives than soils with more than 1% SOC. In regional scale model simulations of pesticide leaching to groundwater, covering many soils, input parameters for each pesticide include a single soil organic carbon sorption coefficient and single half-life value. Our results imply, however, that the approach to these regional scale assessments could be improved by adjusting the values of these two input parameters according to SOC. Specifically, this study indicates that for 2,4-D and Alberta soils containing less than 1% SOC, the 2,4-D pesticide parameters obtained from generic databases should be divided by 2.5 (soil organic carbon sorption coefficient) and 1.5 (half-life value).     

Inheritance and physiological basis for 2,4-D resistance in prickly lettuce (Lactuca serriola L.)

Experiments were conducted to determine the inheritance and physiological basis for resistance to the synthetic auxinic herbicide (2,4-dichlorophenoxy)acetic acid (2,4-D) in a prickly lettuce biotype. Inheritance of 2,4-D resistance in prickly lettuce is governed by a single codominant gene. Absorption and translocation were conducted using (14)C-2,4-D applied to 2,4-D-resistant and -susceptible biotypes. At 96 h after treatment (HAT), the resistant biotype absorbed less applied 2,4-D and retained more 2,4-D in the treated portion of the leaf compared to the susceptible biotype. The resistant biotype translocated less applied 2,4-D to leaves above the treated leaf and crown at 96 HAT compared to the susceptible biotype. No difference in the rate of metabolism of 2,4-D was observed between the two biotypes. Resistance to 2,4-D appears to originate from a reduced growth deregulatory and overstimulation response compared to the susceptible biotype, resulting in lower translocation of 2,4-D in the resistant prickly lettuce biotype.     

Multiresidue analysis of cotton defoliant, herbicide, and insecticide residues in water by solid-phase extraction and GC-NPD, GC-MS, and HPLC-diode array detection

A multiresidue procedure was developed for analysis of cotton pesticide and harvest-aid chemicals in water using solid-phase extraction and analysis by GC-NPD, GC-MS, and HPLC-DAD. Target compounds included the defoliants tribufos, dimethipin, thidiazuron; the herbicide diuron; and the insecticide methyl parathion. Three solid-phase extraction (SPE) media, octadecylsilyl (ODS), graphitized carbon black (GCB), and a divinylbenzene-N-vinyl pyrollidine copolymer (DVBVP), were evaluated. On GCB and ODS, recoveries varied depending on compound type. Recoveries were quantitative for all compounds on DVBVP, ranging from 87 to 115% in spiked deionized water and surface runoff. The method detection limit was less than 0.1 microg L(-)(1). SPE with DVBVP was applied to post-defoliation samples of surface runoff and tile drainage from a cotton research plot and surface runoff from a commercial field. The research plot was defoliated with a tank mixture of dimethipin and thidiazuron, and the commercial field, with tribufos. Dimethipin was detected (1.9-9.6 microg L(-)(1)) in all research plot samples. In the commercial field samples, tribufos concentration ranged from 0.1 to 135 microg L(-)(1). An exponentially decreasing concentration trend was observed with each successive storm event.     

Rapid gas chromatographic method using nitrogen-phosphorus detection for N-nitrosodimethylamine in 2,4-D and MCPA herbicide formulations

A simple and rapid analytical method has been developed for the determination of N-nitrosodimethylamine (NDMA) in amine salts of phenoxy herbicide formulations of 2,4-D and MCPA, plus mixtures of these with mecoprop and dicamba amine salts. Sample preparation consists of direct extraction using pre-packed disposable extraction tubes eluted with dichloromethane followed by cleanup on a disposable silica gel mini-column using ethyl acetate as eluting solvent. Samples are injected on-column for gas chromatography with a Megabore fused silica column; the NDMA is measured by a thermionic specific detector (TSD) that is selective for nitrogen-phosphorus (NP). A detection limit of 0.1 microgram/mL was easily attainable without any concentration step because the solvent volume is minimal. TSD and thermal energy analyzer (TEA) results have been compared and confirmed by gas chromatography/mass spectrometry. Recovery studies were performed as well as a reproducibility study on one of the 2,4-D formulations.