determination of dodecylbenzenesulphonates and ethoxylated alkylphenols in liquid pesticide...

Journal of Chromatography, 435 (1988) 13-82 Elsevier Science Publishers B.V., Amsterdam - Printed in The Netherlands

CHROM. 20 035

DETERMINATION OF DODECYLBENZENESULPHONATES AND ETH- OXYLATED ALKYLPHENOLS IN LIQUID PESTICIDE FORMULATIONS BY HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY

R. H. SCHREUDER* and A. MARTIJN

Plant Protection Service, P.O. Box 9102, 6700 HC Wageningen (The Netherlands)

(Received June 22nd, 1987)

SUMMARY

A method is presented for determining mixtures of calcium dodecylbenzenesulphonate (CaDBS) and ethoxylated alkylphenols in pesticidal emulsifiable concentrates. The pesticide formulations are adsorbed on a pre-concentration column. The active ingredients, the solvents and the emulsifiers are eluted succesively by using solvents with increasing polarities. The various homologues of CaDBS and the ethoxylated alkylphenols are separated by reversed-phase high performance liquid chromatography using methanol-water containing tetramethylammonium bromide. The oligomers of the ethoxylated phenols are separated on an aminopropyl-modified column using a solvent gradient (propan-Zol-water and hexane-tetrahydrofuran). Quantitative results are given for synthetic samples. The degree of ethoxylation of the ethoxylated alkylphenols has been determined and the CaDBS components char- acterized.

INTRODUCTION

Blends of dodecylbenzenesulphonates and ethoxylated alkylphenols are exten- sively used as emulsifiers in liquid pesticidal formulations. The qualitative and quantitative analysis of these compounds using a system of titrimetric, spectrometric and chromatographic techniques has been reported l. The procedure, however, is com- plicated and time-consuming, lacks specificity and is subject to interferences. This was a reason to look for a better method, which preferably ,should use only one analytical technique. Because of the ionic nature of the dodecylbenzenesulphonates and the comparatively low volatility of the ethoxylated alkylphenols, an high-performance liquid chromatographic (HPLC) approach seemed to offer the best solution to the problem. It is known2 that dodecylbenzenesulphonates can be determined, quantitatively and qualitatively, by reversed-phase HPLC. Identification of ethoxylated alkylphenols with respect to the alkylphenol moiety is also possible with reversed-phase HPLC1q3, whereas a good impression of the frequency distribution of the oligomers and a reasonable estimate of the degree of ethoxylation can be obtained by using an alkylamine-modified column1,4.

0021-9673/88/$03.50 0 1988 Elsevier Science Publishers B.V

74 R. H. SCHREUDER, A. MARTIJN

The separation of emulsifiers from active ingredients and solvents can be achieved by classical column chromatography. The modern preconcentration columns however are likely to require considerably less solvent and to shorten the analysis time.

This paper describes procedures for determining the components of commonly used mixtures of calcium dodecylbenzenesulphonates and ethoxylated alkylphenols in emulsifiable concentrates. Results of the quantitative and qualitative analysis of samples, using a reversed-phase HPLC column, are presented. The degree of ethoxylation of the ethoxylated component is established with the aid of analkylamine- modified silica column.

EXPERIMENTAL

Apparatus The HPLC system consisted of two high-pressure pumps (Model 9208), a sol-

vent programmer (Model 9224) an autosampler (Model 9209) with a 20-~1 injection volume, all from Kipp Analytica (Delft, The Netherlands), a Pye Unicam PU 4020 variable-wavelength UV detector (Philips, Eindhoven, The Netherlands), set at 225 nm and an Hewlett-Packard 3390, electronic integrator. Columns, stainless steel, 250 mm x 4.6 mm I.D., packed with either LiChrosorb 10 RP-8 or Hypersil APS (aminopropyl-modified silica) were obtained from Chrompack (Middelburg, The Nether- lands). Pre-concentration cartridges filled with silica gel and aminoalkyl-modified silica gel (Sep-Pak, Art. Nos. 51900 and 1083, Waters Associates) were used to isolate the surfactants. The cartridges fitted onto 20-ml Luer-LOK syringes.

Reagents Dichloromethane, ethyl acetate, methanol, propan-2-01 and xylene were of

analytical quality; hexane and tetrahydrofuran (THF) were of HPLC quality. De- mineralized water was purified further using a Mini-Q filtration system (Millipore). Tetramethylammonium bromide (TMAB) and silica gel (Art. Nos. 8127 and 7724) were obtained from Merck. Linear and branched-chain calcium dodecylbenzenesulphonates (CaDBS) were obtained from Tensia (Zaandam, The Netherlands); ethoxylated nonylphenol (NP-EO) with an ethoxylation degree of 8.5 was obtained from ICI (U.K.). Triton X-100 with an ethoxylation degree of 9.5 (Merck, Art. No. 11869) was used as a source for ethoxylated octylphenol (OP-EO). Aromatic process oil (boiling range, 310-4OO”C; aromatic content, 50%), heating oil and Shells011 K were obtained locally. Pesticides were technical grade and were obtained from the usual suppliers.

Standard solutions Standard solutions were prepared as follows. CaDBS, linear type (100 mg),

and NP-EO or OP-EO (100 mg) were dissolved in 50 ml methanol-water (75:25) with 0.005 M TMAB. An aliquot of 5 ml was pipetted into a 50-ml volumetric flask and diluted to volume in methanol-water (75:25) with 0.005 M TMAB.

NP-EO or OP-EO (330 mg) was dissolved in 50 ml tetrahydrofuran. An aliquot of 5 ml was pipetted into a 50-ml volumetric flask and diluted to volume in hexane-THF (70:30).

HPLC OF DODECYLBENZENESULPHONATES AND ALKYLPHENOLS 15

Samples The samples were prepared by dissolving the emulsifiers and the technical grade

pesticides in xylene to obtain solutions containing 20 g/l CaDBS (linear), 20 g/l NP-EO or OP-EO and 200 g/l pesticide, unless otherwise stated. NP-EO-containing samples were alachlor, bupirimate, chlorfenvinphos, 2-methyl-4,6-dinitrophenol (DNOC) (100 g/l in heating oil-aromatic process oil, l:l), parathion (methanol), permethrin (100 g/l in Shellsol K) and triallate. The OP-EO-containing samples were chlorpropham and lindane (70 g/l). Blank samples consisted of xylene solutions containing 20 g/l CaDBS (linear) and 20 g/l NP-EO or OP-EO.

Determination of surfactants To isolate the surfactants, 1 ml of the sample was diluted to 50 ml in

hexanedichloromethane (1: 1). A 20-ml syringe was connected to the silica gel cartridge wetted with hexane-dichloromethane (1: l), and 5 ml of the solution were pipetted into the syringe. The solution was pressed slowly through the cartridge. After the syringe and the cartridge had been flushed with two l-ml portions of hexanedichloromethane (1: l), the solvent and the pesticides were removed from the cartridge by elution successively with 5 ml hexane-dichloromethane (l:l), 5 ml dichloromethane and 5 ml dichloromethane-methanol(99: 1). In a separate series of determinations the polar pesticides alachlor, bupirimate, permethrin and triallate were also eluted with consecutively 5 ml hexanedichloromethane (l:l), 5 ml dichloromethane and 5 ml dichloromethane-methanol (98:2). Then the surfactants were eluted with 5 ml methanol. This fraction was collected in a 25-ml round-bottomed flask and the solvent was removed with the aid of a rotating evaporator. The residue was taken up in 10.0 ml of methanol-water (75:25), with 0.005 M TMAB. To separate the CaDBS homologues and OP-EO, 20-~1 aliquots were injected onto the Cs reversed-phase column and eluted with methanol-water (75:25), containing 0.005 M TMAB. The flow-rate was set at 1.0 ml/min and the absorption was measured at 225 nm. When NP-EO was present, a linear gradient was applied with the following eluents: A = methanol-water (65:35) with 0.01 M TMAB; B = methanol-water (85:15) with 0.01 M TMAB. The percentage of B was increased from 40 to 70 from time 0 to 20 min. Peak areas were determined by electronic integration and compared with those of standard solutions obtained under the same chromatograhic conditions. The areas of the four main peaks of linear CaDBS were summed and compared with the sum of the corresponding peaks of the standard. All other minor peaks of CaDBS were neglected.

Isolation of single oligomers from ethoxylated alkylphenols Silica gel columns were prepared by filling glass columns (200 mm x 13 mm

I.D.), fitted with a sintered glass disk, with a slurry of silica gel in ethyl acetate. Pressure was applied carefully to condense the adsorbent layer. Additional amounts of slurry were added until a layer of 15 cm silica gel remained. Care was taken that the column did not dry out. A small plug of cotton wool was placed on top of the silica gel layer and the column was washed with 50 ml of ethyl acetate. A lo-ml volume of a solution containing 1 g of NP-EO or OP-EO dissolved in ethyl acetate was then transferred to the top of the column and the oligomers were eluted with lOO-ml portions of ethyl acetate, ethyl acetate-methanol (99.5:0.5) and ethyl


acetate-methanol (99:l). Three fractions of 100 ml were collected. From the last fraction the solvent was removed with the aid of a rotating evaporator and the residue was taken up in hexane-THF, (70:30). The oligomer distribution was determined by the WPLC method below. In both cases the chromatogram showed three peaks, a main peak flanked by two smaller ones. Field-desorption mass spectrometry gave molecular masses of 616 and 602 respectively for the main components, corresponding to nine ethylene oxide units per molecule for both NP-EO and OP-EO. The retention times of the main peaks were used to identify the corresponding peaks in the chromatograms of the samples and standards.

Determination of oligomer distribution of ethoxylated alkylphenols The first steps of the clean-up were the same as the ones used for the deter-

mination of the total surfactant content. After the active ingredients and the solvents had been removed, thus after the elution step with dichloromethane-methanol(99: l), the outlet of the silica gel cartridge was connected to an alkylamine-modified silica gel cartridge by a piece of PTFE tubing (40 mm x 1.6 mm I.D. x 4.1 mm O.D.), and 5 ml dichloromethane-methanol(80:20) were applied by means of a syringe. The effluent was collected in a 25ml round-bottomed flask, the solvent was removed with a rotating evaporator and the residue was dissolved in 3.0 ml hexane-THF (70:30). Of this solution, 20-~1 aliquots were injected.

Single oligomers of NP-EO and OP-EO were separated on the alkylamine- modified silica column with a linear gradient of eluent D (propan-2-ol-water, 90:10)

b

Fig. 1. Chromatograms of ethoxylated phenols: (a) NP-EO, degree of ethoxylation 8.5; (b) OP-EO, degree of ethoxylation 9.5; peak 9 = nonamec. For conditions see text.

HPLC OF DODECYLBENZENESULPHONATES AND ALKYLPHENOLS 1-l

in eluent C (hexane-THF, 70:30). The ratio of D to C was increased from 0.05 at time 0 to 0.5 in 60 min. The flow-rate was 1.0 ml/min, the absorption was measured at 225 nm (Fig. 1). Peak areas were determined by electronic integration and compared with those of the standard solution obtained under the same conditions. Start- ing with the peak of the oligomer with the known number of ethylene oxide units, the contributions of the other peaks were calculated taking into account the respective relative molecular masses and assuming equal molar absorptivity. Then the contri- bution of each peak to the degree of ethoxylation was determined.

RESULTS AND DISCUSSION

Selection of the HPLC system During preliminary investigations it proved to be impossible to separate with

one single chromatographic system all the CaDBS homologues and the numerous NP-EO and OP-EO oligomers. The ion-pair HPLC method used for the CaDBS homologues could not separate the oligomers of the ethoxylated phenols. It was, however, also selective with respect to the alkyl chains attached to the phenols, so that the different kinds of ethoxylated alkylphenols could be distinguished. Obviously all the oligomers of one ethoxylated alkylphenol had the same retention time. It was, therefore, necessary to turn to another method to obtain separation of the individual oligomers.

In a previous study it had been possible to separate the oligomers of NP-EO and OP-EO' . This method could be used with some modifications. Measures had to be taken to remove CaDBS because under the conditions used it would have remained on the analytical column and caused severe deterioration of the column performance. A guard column alone proved to be insufficient to retain CaDBS. After a few injections, CaDBS appeared in the eluate of the guard column. Therefore, an additional clean-up step was used, which. consisted of separating the ethoxylated alkylphenols from CaDBS by means of a alkylamine-modified silica gel cartridge. On this cartridge 99% of the CaDBS remained, whereas the ethoxylated phenols were completely eluted.

The ion-pair HPLC system used TMAB as a counter ion. As it might be ex- pected that the TMAB concentration would influence the capacity factors, k’, of the single homologues, a series of experiments were set up to establish the optimum concentration. The results are shown in Table I. Apart from the concentration of TMAB, the conditions were as given in the Experimental section. Linear CaDBS gives four well separated main peaks together with some minor peaks, whereas branched-chain CaDBS shows a cluster of badly separated peaks (Fig. 2). The capacity factors for NP-EO, OP-EO and BP-E0 (ethoxylated triisobutylphenol) were determined at the same time. They remained constant with changing TMAB concentrations. The following values were found: NP-EO, 4.94; OP-EO, 3.41; BP-EO, 13.4 and 16.4. Earlier studies’ had also shown that the degree of ethoxylation did not affect the retention times. If one compares the capacity factors for NP-EO and OP-EO with those from Table I, it is evident that the highest TMAB concentration that could be used was about 0.005 M. At higher concentrations the tetradecylbenzenesulphonate peak would start to overlap the OP-EO peak. On the other hand, a somewhat better separation was obtained at higher TMAB concentrations. So, for mixtures containing


TABLE I

CAPACITY FACTORS OF ALKYLBENZENESULPHONATES WITH INCREASING TMAB CON- TENT IN THE MOBILE PHASE

Compound ThkAB concentration (M)

,0.0025 0.005 0.01 0.02 0.04

Decylbenzenesulphonic acid 0.24 0.55 0.99 1.33 3.35 Undecylbenzenesulphonic acid 0.42 0.83 1.40 1.87 3.78 Dodecylbenzenesulphonic acid 0.68 1.23 1.96 2.61 4.35 Tridecylbenzenesulphonic acid 1.03 1.77 2.73 3.63 5.13 Tetradecylbenzenesulphonic acid 1.52 2.53 3.61 4.97 6.19

NP-EO the TMAB concentration was increased to 0.01 M and a solvent gradient was applied to cause the NP-EO to elute more quickly, which resulted in a shortening of the analysis (Fig. 3).

Sample composition For several reasons we preferred to work with made-up samples instead of

commercial ones. First the composition of commercial samples is usually not known exactly. Secondly, even if the emulsifier content is given, the exact surfactant content is usually uncertain because commercial emulsifiers used in pesticide formulation practice often contain unknown amounts of water and other solvents. Thirdly, by using the same emulsifiers for all the samples and by dosing the same amount the results could be better compared. The active ingredients were selected to cover a whole range of polarities. This was necessary in order to see how well the surfactants could be separated from the other components of the formulation. Other selection

2

a

3 .?(I: 2 4

5

b

A; Fig. 2. Chromatograms of CaDBS: (a) linear; peaks: 1 = decyl-; 2 = undecyl-; 3 = dodecyl-; 4 = tridecyl-; 5 = tetradecylbenzenesulphonate; (b) branched chain. For conditions see text.

HPLC OF DODECYLBENZENESULPHONATES AND ALKYLPHENOLS 79

criteria were the availability and the stability of the respective pesticides. Further, the samples were composed in such a way that they resembled pesticide formulations encountered in practice. This was also the reason why linear CaDBS was chosen as the anionic component of the emulsifier, and NP-EO or OP-EO as the non-ionic component. Blank samples contained the emulsifier mixture and the solvent, but no pesticide. The emulsifiers used were commercial preparations of technical quality, known to contain one single type of surfactant. Because of the lack of suitable reference materials, no attempts were made to determine the actual surfactant content.

Standardization and quantitative aspects As no pure reference material was available, the technical product was used

as a standard. In the calculation of the results, the sum of the areas of the four main peaks of the sample was compared with the sum of the areas of the corresponding peaks of the standard. The minor peaks were neglected in order to avoid too large errors caused by inaccurate area determinations. The procedure described above can- not be applied in all cases because the relative abundances of the alkyl chain lengths may differ, depending on the source of the dodecylbenzenesulphonate. When the standard differs from the CaDBS in the sample, the calculation must be based on the contributions of the individual peaks. One then has to know the relative abundances of the various alkylbenzenesulphonates of the standard. A reasonably good approximation can be obtained by applying a normalization procedure to the four main peaks. Thereby, it is implicitly assumed that the molar absorptivity is the same for all the alkyl homologues. Table II gives the average results of ten injections. To identify the individual compnents and to carry out a check by an independent

7

b

L , min , 0 5 10 15 20

Fig. 3. Chromatograms of mixtures of CaDBS and ethoxylated alkylphenols: (a) CaDBS + NP-EO with gradient elution, peak 6 = NP-EO; (b) CaDBS + OP-EO with isocratic conditions, peak 7 = OP-EO. For conditions see text; peaks l-5 as in Fig. 2.


TABLE II

COMPOSITION (MOL.%) OF TECHNICAL DODECYLBENZENESULPHONATES DETER- MINED BY HPLC AND FIELD DESORPTION MASS SPECTROMETRY (FD-MS)

Compound HPLC FD-MS

Decylbenzenesulphonic acid Undecylbenzenesulphonic acid Dodecylbenzenesulphonic acid Tridecylbenzenesulphonic acid Tetradecylbenzenesulphonic acid

10.0 10 40.1 38 31.3 32 18.6 18 - 2

method, the relative abundances were also determined by field-desorption mass spectrometry. To this end, CaDBS was converted into the free sulphonic acid by passing it over a cation-exchange column. The main peaks observed were those corresponding to the Ci0-Ci4 benzenesulphonic acids. The calculated relative abundances are also given in Table II. There is reasonably good agreement between the results obtained by mass spectrometry (MS) and those obtained by HPLC. No attempts were made to integrate the C14-benzenesulphonic acid peak in the case of the HPLC mea- surement, because it was too small for an accurate estimate of its area. The value for Ci4-benzenesulphonic acid found by MS should also be considered as highly inaccurate. It is only given to show that there was a minor fraction of tetradecylbenzenesulphonic acid in the sample.

The results of the determination of CaDBS and NP-EO or OP-EO in the synthetic samples are given in Table III. The values represent single determinations. In a few instances the preconcentration column insufficiently separated the active in- gredient and the surfactant, and interference was observed with one of the CaDBS peaks. In those cases the calculation of the total CaDBS content was not based on

TABLE III

RECOVERIES (%) OF CaDBS AND ETHOXYLATED ALKYLPHENOLS IN SYNTHETIC SAM- PLES*

Sample CaDBS OP-EO NP-EO

Alachlor

Bupirimate

Chlorfenvinphos Chlorprofam DNOC Lindane Parathion Permethhrin

Triallate

99.6**

109.8***

95.0*** 102.0***

93.6***

101.4 96.1

98.8

97.2

99.6*** 95.4***

106.0***

99.6***

_ 103.0 _ 103.1**+ _ 96.4 _ 100.8*** _ 105.9 99.3 -

_ 92.5 97.1 -

- 93.6 - 95.9 - 92.4*** _ 102.0 _ 112.2***

l Corrected for recovery of blank samples, see text. l * Three CaDBS peaks taken for the calculation. l ** Third elution solvent: dichloromethane-methanol (98:2).

determination of dodecylbenzenesulphonates and ethoxylated alkylphenols in liquid pesticide...

Documents