Preconcentration of Pesticides from Waters

V. Tatar and M. Popl

Department of Analytical Chemistry, Institute of Chemical Technology, CS-166 28 Prague 6, Czechoslovakia

Anreicherung yon Pesticiden aus W/issern

Zusammenfassung. Die Anreicherung yon Pestieiden aus verschiedenen Wfissern und durch Sorption auf einem orga- nischen Sorbent und anschlieBende Desorption mit einem organischen L6sungsmittel wird durchgeffihrt. Die ent- sprechenden Eigenschaften einiger Sorbentien (Separon SE, Tenax GC, Porapak Q, Separon SIC 18) wurden auf Grund von Parametern, die die Sorptionseffektivitfit beeinflussen, verglichen. Die Untersuchungen erwiesen die vorzfiglichen Eigenschaften von Separon S[ C 18 fiir den genannten Zweck.

Summary. A procedure for the preconcentration of pesticides from water is described using the sorption on an organic sorbent with following desorption by means of an organic solvent. The usefulness of a few sorbents (Separon SE, Tenax GC, Porapak Q, Separon SIC 18) was compared on the basis of several factors influencing the sorption. The experiments showed the excellent suitability of Separon SI C 18 for this purpose.

Introduction

Sorption on organic sorbents is increasingly used as the preconcentration step in the analysis of trace amounts of pesticides in waters. In addition to a sufficient sorption capacity, the behaviour of the sorbent during the sorption and desorption of an analyte is of importance: the sorbent should exhibit a steep desorption curve and no irreversible sorptions or Other anomalous phenomena should occur.

In our previous papers [1, 2] we have dealt with the preconcentration of s-triazines and phenoxycarboxylic acids by sorption on Separon SE 50/50 macroporous polymeric sorbent. The so-called sorbent hysteresis has been studied with these pesticides [3] and Separon SE has been compared with other sorbents used for this purpose (Porapak Q, Tenax GC) [4].

In this paper, a modified procedure is described for the preconcentration of pesticides from waters by sorption on organic sorbents; as compared with the conventional approach, this modification is less time consuming and more economical. Attention is paid to the choice of the sorbent and to the optimum operation conditions.

Offprint requests to: M. Popl

Experimental

Apparatus, Sorbents, Chemicals

The sorption column used (Fig. 1) was a glass column 30 mm long, 3.2 mm i.d. and 7.4 mm o.d., fitted on both ends with stainless-steel microsieves and Teflon sealing gaskets and accomodated in a metal jacket. The sample or desorbing agent was injected onto the column by means of a syringe; the liquid could also be delivered with a micropump.

Porapak Q (Waters), Tenax GC (Enka) and Separon SE 50/50 (Laboratorni pfistroje, Prague) polymeric sorbents and Separon SIC 18 sorbent (Laboratorni pfistroje, Prague) were used. The column was always full of sorbent, so that the bed length was constant whereas the weight of the packing was different according to the different volume weight of the various sorbents. Data of the sorbents used are given in Table 1.

A linear dispenser was used for forcing the water samples through the sorption column at a constant and defined flow rate with injection syringes, whereas an MC 300 double- acting plunger pump (Mikrotechna, Prague) was employed in experiments aimed at the determination of the sorption capacity.

The test substances in desorbate were determined on a Packard Model 428 gas chromatograph fitted with a flame

Fig. 1 Sorption column (this column is supplied as precolumn for HPLC by Laboratorni pfistroje, Prague)

Fresenius Z Anal Chem (1985) 322:419-422 �9 Springer-Verlag 1985

Table 1. Properties of the tested sorbents

Sorbent Symbol Chemical structure Weight of sorbent Particle size [mgl [mm]

Separon SE 50/50 1 A Styrene-ethylenedimethacrylate copolymer 90 0.063- 0.08 Separon SE 50/50 1 B 90 0.16 --0.25 Separon SE 50/50 l C 90 0.25 -0.315 Separon SE 50/50 1 D 90 0.315-0.40 Separon SE 50/50 l E 90 0.40 -0.50 Separon SE 50/50 1 F 90 0.50 --0.63 Porapak Q 2 83 0.18 -0.30 Tenax GC 3 33 0.18 -0.25 Separon SIC 18 4 100 0.063-0.08

Ethylvinylbenzene-divinyl-benzene copolymer Polymer on the basis of 2,6-diphenyl-n-phenylene oxide Silica modified with octadecyl groups

ionization detector and a WCOT type 25 m x 0.25 mm glass capillary column coated with SE-30 stationary phase.

A Pye Unicam 800 B spectrophotometer was employed for the determination of chlorpropham in water during the sorption capacity measurements.

Chlorpropham (isopropyl-N- 3-chlorophenyl carba- mate), whose solubility in water is 80 rag/l, served as the basic test substance. The preconcentration was also tested on phenoxycarboxylic acid pesticides (MCPA, 2,4-D), s-tri- azines (atrazine, simazine, propazine, ametryne, desmetryne, prometryne and terbutryne), and triallate as an additional carbamate pesticide. The model samples were prepared by adding methanolic stock solutions of the pesticides to dis- tilled water.

Procedure

The isolation and preconcentration of pesticides from water involves the following steps: a) forcing the sample through the sorption column by means of a syringe; b) removal of water from the sorbent by air purging (about ten volumes of a 20 ml syringe); c) desorption with acetone - this solvent is added using a syringe, and the first 0.5 ml of desorbate is collected for the determination; d) flushing the column with acetone and distilled water.

The sorption behaviour was examined as follows: 1. The recovery was determined in dependence on the

sample flow rate. 10 ml of water sample containing 4 ppm of chlorpropham was forced through the column; the syringe plunger motion was controlled by the linear dispenser so that the flow rate was varied over the region of 0.6 to 20 ml/min.

2. For Separon SE, the sorption efficiency was measured in dependence on the sorbent particle size at a constant flow rate of 3.5 ml/min using the same sample as above. Six batches of Separon SE of different size were tested (Table 1).

3. The capacity of the sorbents was determined. A water sample containing chlorpropham in a concentration of 80 ppm was delivered at a rate of 1.2 ml/min by using the double-acting plunger pump. 10 ml fractions were collected and chlorpropham was determined in them spectrophoto- metrically at 283 nm. The sorption capacity is expressed as the amount of retained chlorpropham, in mg per g of sorbent, at the moment the pesticide concentration in the effluent is 10% of that in the influent.

4. The desorption curves were measured for all the sorbents tested. 10ml of water containing 10gg of chlorpropham was delivered at a rate of 1.2 ml/min and

420

t n- 80]

0 . . . . . . . . . . . . . . 5 1 0 1'5 " m l / m i n 2'0 SAMPLE FLOW RATE

Fig. 2. Dependence of sorption recovery on the sample flow rate in the sorption column (10 ml of water containing 4 ppm of chlorpropham). Curves designated according to Table 1

solute was desorbed with acetone; 0.1 ml volumes were collected, diluted to 0.5 ml with this solvent, and submitted to analysis.

5. The recovery was measured in dependence on the sample volume. 10, 20, 50, and 100 ml volumes, containing invariably 40 gg of chlorpropham, were passed through the column at a rate of 1.2 ml/min.

6. The effect of the sorbent hysteresis was investigated. Sorption and desorption of 4 mg of chlorpropham from 50 ml of water was repeated three times, and a blank exper- iment with 10 ml of water followed; chlorpropham was determined in the desorbate from the blank experiment. The same experiment was carried out with 2,4-D and with propazine; 0.4 mg of the latter was sorbed.

Results and Discussion

The sorption recoveries in dependence on sample flow rate are shown in Fig. 2. Except for Separon SIC 18, the sorption of chlorpropham is complete only at flow rates not exceeding 1.2 ml/min, whereas at higher flow rates the recovery de- creases, particularly for Separon SE and Tenax. For Pora- pak the recovery decrease is comparable with that for the two former sorbents up to 3 ml/min, but at higher flow rates the recovery does not vary appreciably any more. For Separon SI C 18 the recovery is 100% for flow rates up to 10 ml/min, and even at 20 ml/min the recovery is better than 90%. Thus the preconcentration on Separon S I C 18 can be performed nearly ten times faster than on the other sorbents

% 100 "

>90 E

~ e o ~ 1A m

70

60

5 o

4o 1

30 . . . . g . . . . lb . . . . 1'5 ' ml/min ~ SAMPLE FLOW RATE

Fig. 3. Dependence of sorption recovery on the sample flow rate for different particle sizes of Separon SE (10 ml of water containing 4 ppm of chlorpropham). Curves designated according to Table 1

2

~0.3- /

[ / " ,,, 2,I

k, / / / ' ]

0 20 40 60 80 100 120 140 160 ml 1~0 SAMPLE V O L U M E

Fig. 5. Determination of sorbent capacity (sample: water containing 80 ppm chlorpropham) Curves designated according to Table t

100'

O U E

8(

70.

60 0.1 0.2 0.3ram2

SQUARED PARTICLE DIAMETER

Fig. 4. Dependence of sorption recovery on particle size of sorbent Separon SE (water containing 4 ppm of chlorpropham, flow rate 3.5 ml - rain 1)

tested, and the flow rate need not be held constant because the recovery is independent of the flow rate up to 10 ml/ min. This dissimilarity between Separon S I C 18 and the other sorbents is evidently caused by the different structure nature: while Separon SE, Tenax, and Porapak are porous polymeric sorbents where the sorption occurs also inside the sorbent particles, Separon S I C 18 is silica gel modified by C1 s groups where the sorption is limited to the surface layer. Increase in the flow rate has an effect particularly for the porous sorbents because, with the surface layers saturated, the sorbate molecules are not retained in the column for a time long enough for their permeation into the particle bulk.

For the porous sorbents, the particle size can also be expected to have an effect on the recovery. The dependence as in Fig.2 is plotted in Fig. 3 for three batches of Separon SE with different particle size (sorbents 1 A, 1 B, and 1 F). In spite of the same particle size, sorbent 1 A does not attain the quality of Separon S I C 18 from the recovery point of view. The recovery decreases with increasing particle size of sorbent. Figure 4 shows the dependence of sorption recovery on the squared average particle size of Separon SE (sorbents

Table 2. Capacity of tested sorbents

Sorbent Symbol Capacity [mg/g]

Separon SE 50/50 1A 106 Separon SE 50/50 1 B 13 Separon SE 50/50 1 C 4.5 Porapak Q 2 144 Tenax GC 3 1.5 Separon SIC 18 4 138

1 A - 1 F) at a constant flow rate. The dependence is linear over the region examined (regression coefficient 0.9976); this is consistent with the above assumption of diffusion of sorbate molecules into the particle bulk. According to the second Fick's law the mean path of molecules depends on their diffusion coefficient and time as

x= = 2 D t .

I f the mean path of the molecules travelled during their contact with a particle is shorter than as necessary for their permeation into the particle bulk while the surface layers are saturated, the recovery is incomplete.

Figure 5 shows the results of experiments the purpose of which was to determine the capacity of the sorbents under study. The values for the conditions as given above are listed in Table 2. A chlorpropham concentration four times lower, i.e., 20 ppm, had to be used for Tenax whose capacity is about two orders of magnitude lower than that of Porapak Q, Separon S I C 18, or Separon SE of 0 .06-0.08 mm particle size. Table 2 also shows the effect of particle size on the capacity of Separon SE at a constant sorbent weight and volume and at a constant flow rate.

The desorption curves of the sorbents are plotted in Fig. 6. The porous sorbents (1 A, 2, 3) give curves similar in shape; quantitative desorption is attained using 0.5 ml of acetone. The curve for Separon S I C 18 is considerably steeper, and 0.3ml of acetone is sufficient for the

421

Or~giHa~ Papers

% 100

280 o>- r162 60

40

20

o~ o~ 0:3 o54 6.s o~ ml VOLUME OF DESORBING AGENT

Fig. 6. Desorption curves (Sorbed solute: 10 gg of chlorpropham) Curves designated according to Table I

quantitative desorption, owing to which a higher degree of preconcentration can be achieved. The shape of the desorp- tion curve is unaffected by the acetone flow rate up to 3 ml/ min.

As to the effect of sample volume on the sorption re- covery, no effect is observed over the region of 10 to 100 ml unless the sorbent capacity is surpassed.

A phenomenon referred to as sorbent hysteresis has been described and studied previously [3, 4]. In the present work, however, where a simplified preconcentration procedure was used omitting the sorbent drying by heating the column at 110~ before the desorption, the hysteresis effect was never observed.

Conclusions

The results give evidence that Separon S I C 18 (or other sorbents of this kind) is well suited for the preconcentration of water pollutants. This sorbent exhibits a sufficient ca- pacity and a steep desorption curve, which permit a degree of preconcentration better than 300 to be achieved with 100 ml of sample. The fact that the sorption recovery is independent of the sample flow rate is also advantageous because the sample can then be conveniently forced through the column manually, without having to keep the flow rate constant. Of the polymeric sorbents, Porapak Q and Separon SE of small particle size also give good results, the sorption, however, must be conducted ten times more slowly (or with a lower recovery at a constant flow rate), and the degree of preconcentration is poorer. With Separon SE of a small particle size (0.06-0.08 mm), difficulties also arise from its volume changes occurring as one eluent is replaced by another (water - acetone); the resistance of the column increases and ultimately the column can get clogged. Tenax is of limited use because of its low capacity.

References

1. Popl M, Vozfifikovfi Z, Tatar V, Strnadovfi J (1983) J Chromatogr Sci 21 : 39

2. Tatar V, Vozfi/tkovfi Z, Popl M (1983) Sci Papers Prague Inst Chem Technol H 18:187

3. Popl M, Tatar V, Vozfifikov/t Z (1982) Fresenius Z Anal Chem 313:137

4. Tatar V, Vozfifikovfi Z, Popl M (1983) Hydroch6mia 83:187

Received March 2, 1985

422