Potential of a modified solvent-extraction flow-injection analysis

Potential of a modified solvent-extraction flow-injection analysis

Talanta, Vol. 36, No. 6, pp. 691-693, 1989 Printed in Great Britain. All rights rescrverl ~39.91~189 $3.00+ 0.00 Copyright 8 1989PerSamon press pk P...

285KB Sizes 0 Downloads 67 Views

Talanta, Vol. 36, No. 6, pp. 691-693, 1989 Printed in Great Britain. All rights rescrverl

~39.91~189 $3.00+ 0.00 Copyright 8 1989PerSamon press pk

POTENTIAL OF A MODIFIED SOLVED-EXTORTION FLOW-INJECTION ANALYSIS JUN’ICHI TOEI

Scientific Instrument Division, Tosoh Co. Ltd. 2743-l Hayakawa, Ayase-shi, Kanagawa 252, Japan ~Re~eived 16 March 1987. Revised 12 February 1988. Accepted 7 December 1988) Summary-A modified flow-injection solvent extraction procedure involving a phase-separator column has been developed. The main feature is the selective absorption of the injected aqueous phase by the separator column. The procedure is in effect a variety of liquid-liquid partition or affinity chromatography.

Since Karlberg and Tbalander’ described an extraction system, based on the principle of flow-injection

analysis (FIA) with a T-type phase separator, numerous applications of FIA solvent extraction have been developed and reported,r4 and phase separators based on a Teflon membrane are popular for the purpose since Nord et al.’ reported their use. Phase separators for FIA solvent extraction systems must be able to remove completely all traces of the unwanted phase and simultaneously isolate a large fraction of the phase of interest. However, the construction of the phase separator is critical with respect to the peak broadening which may arise in the FIA solvent extraction procedure, and to keeping the unwanted phase out of the detector flow-cell. Recently Sahlestriim and Karlberg’ reported an tmsegmented FIA extraction procedure. Though it had the advantage that the phase separation was complete, the extraction efliciency was not as high as usual. Another disadvantage was that the separation ratio was greatly influenced by the flow-rates of the two phases, making optimization of the system difficult. This paper describes a new type of unsegmented FL4 extraction system with a water-abso~tion column filled with a resin that selectively absorbs the aqueous phase.8 The aqueous phase is injected into a stream of organic solvent and extraction of the analyte takes place during flow through the column, followed by the selective absorption of the aqueous phase. Only the organic phase flows into the detector. The chief advantage of the system is its simplicity, because no reaction tube or complex phase separator is needed. The main disadvantage is that a higher column pressure is sometimes observed.*

Reagents and materials All chemicals were of analytical reagent grade and used without further purification. The chloroform used was HPLC grade and used without drying. The materials used for selective absorption of water were Sumicagel base gel 691

SP-520 (Sumitimokagaku, Tokyo, Japan; “gel-type”) and the inner fibres in a disposable diaper for babies (“Merries”, Kao, Tokyo, Japan; “cotton-type”). Apparatus The FIA system comprised a CCPM multifunction pump (metal-free, , Tosoh. Tokvo. Janan). a Rheedvae 7125 iniecvalve. with a lOO-or {O pl loop, a- glass c&&n (2-8 mm bore, 10 mm long) and a UV-8000 spectrophotometric detector (Tosoh). The absorbance of the organic phase was monitored at 254 nm and recorded on a CP-8000 data station (Tosoh).

ion

Packing of water-absorbents Both types of absorbent were Packed by hand. The net weight of the gel-type was 280 mg for the p~~-~p~ation column (8 mm bore, 10 mm long) and that of the cottontype was 300 mg. The gel-type material became sticky aa it absorbed water; the cotton-type also became glutinous but to a lesser extent. Before use the separator column was conditioned by pumping a 4:l v/v chloroform/methanol mixture through it for 15 min.

RESULTS AND DISCUSSION Manifolcis The extraction of caffeine from demineralized water into chloroform was selected as the model for evaluation of the system. In the usual FIA solvent extraction systems a steady-state signal level is reached with large injection volumes, and further volume increase results only in increased cycle time.’ All peak heights, Z&,,,, can be related to this steadystate level, H,, to give the quantity D’, defined as D’= ~~/~_;’ D’ is thus analogous’ to the dispersion coefficient, I), defined9 for a single-phase system. When the injection volume was increased in the new system, the peak height slowly increased but did not reach a steady-state level even when the injection volume exceeded 2 ml. This means that the dispersion of the injected sample in the cohunn is high, in contrast to the usual FIA solvent extraction systems. If the peak height for caffeine (50 ppm, injection volume 2 ml) is taken as the steady state, D’ is about 29 when the injection volume is 10 ~1.

692

SHORT

COMMUNICATIONS

Comparison of absorbents The choice of absorbent is very important with regard to physical stability. When the gel-type was used, the column pressure gradually increased with the number of injections. After 20-30 injections the column became unusable because of the high pressure required as the material in the column became glutinous through absorption of water. The pressure needed increased even when only the carrier was pumped through the column, because of absorption of the water in the chloroform. The cotton-type material performed better in this respect, with a slower increase in the pressure needed. This is presumably due to more rigid structure of the absorbent. The cotton-type absorbent, pretreated with methanol
I

0.064 obrorb4ncr

-

Optimization of column dimensions Different internal diameters (2, 4, 6, 8 mm) of the glass column were investigated, with a constant length (10 mm) of cotton-type absorbent. With the smaller diameters, effective and reproducible packing could not be achieved without mechanical aid, and higher pumping pressures were needed. A bore of 8 mm was therefore selected. Combination of solvents In an attempt to avoid increase in the pumping pressure needed, the effect of methanol (as a solvent miscible with both phases) was investigated. When 4: 1 v/v chloroform-methanol mixture was used as carrier, the column pressure needed did not increase with number of injections. Presumably the methanol resulted in mild degeneration of the cotton-type absorbent before the injections were begun. However, noisier signals were obtained when samples were injected. This is attributed to water dissolved in the methanolic carrier. It was found that if the column was first conditioned with methanolic chloroform and then chloroform was used as the carrier, no increase in pumping pressure was needed and the signal noise was reduced.

Extraction eflciency Typical recorder traces for caffeine are shown in Fig. 1. No blank peak or baseline noise was observed, but the peaks showed slight tailing. This is attributed to the additional dispersion of samples in the column. To estimate the extraction efficiency, injections were made of aqueous and organic phases with the same concentration of anaiyte. The rest&s are given in Table 1. The peak height for the aqueous phase was slightly higher than that for the organic phase. Considering that the dispersion of the aqueous sample is expected to be smaller than that of the organic phase because of the segmentation, the extraction etliciency must approach 100%.

2 mill

Fig. I. Peaks obtained by injection of 0, 100,200,400 ppm standard solutions of caffeine: flow-rate 1 ml/mitt, injection volume 10 pt.

Choice of injection volume In other FIA solvent extraction systems large injection vohunes are sometimes preferred because of the higher extraction efficiency then obtained.‘O In the new system, however, reproducible and stabie peaks were observed only when the injection volume was below 15 ~1. This was because absorption of the water was not perfect, and sometimes tiny water droplets entered the detector ffow-cell when the injection volume exceeded 15 pl. Therefore, 10-15 ~1 is selected as the injection volume. In its present form the system is restricted to an injection volume smaher than 20 ~1.

Water-loading of the column The column would be expected to become progressively loaded with water along its length, until breakthrough occurred. In the experiments done this effect was not observed, because the injection volume was very small in comparison with the waterabsorption capacity, and the column was renewed we11before the b~akthrough point was approached.

Table 1. Peak height for caffeine in an aqueous and organic phase (caffeine 1000 ppm, injection volume 10 ~1, aqueous phase demineraliaed water, organic phase chloroform, flow-rate 1 ml/min, 2.5 cm corresponds to 0.016 absorbance) Mean peak height, R.S.D. Phase cm (n = 3) Aqueous 19.67 1.4 organic 18.73 0.7

SHORT COMMUNlCATlONS

693

Conclusion

3. L. Nord and B. Karlberg, Anal. Chim. Acru, 1980, 118,

The water-absorption column acts as a phaseseparator as well as removing the segmentation. The apparatus is very simple and the extraction efficiency very high. With a disposable column the system would be applicable with solvent-extraction FIA.

4 T. M. Rocci, D. C. Shelly and I. M. Warner, And. ’ Chem., 1982, 54, 2056. 5. T. Imasaka, T. Harada and N. Isbibasbi, And Chim.

REFERENCES 1.

B. Karlberg and S. Thelander, And. Chim. Acru, 1978, % 1.

2. J.

Kawase., A. Nakata and M. Yamada, And. Chem.,

1979, 51, 1640.

285.

Actu.1981,129,195.

6. 0. Klingboffer, J. Rdiicka and H. Hansen, Tufuntu, 1980, 27, 169. Y. SablestrBm and B. Karlberg, Anal. Chim. ACM, 1986, 179,315. 8. J. Toei, Japan Patent pending. 9. H. Hansen, J. RdZiEka, F. Krug and E. A. G. Zagatto, And. Chim. ACM, 1983, 148,111. 10. L. Fossy and F. Cantwell, And. Chem., 1982,54, 1693. I.