Determination of ammonium ion in sea-water by capillary isotachophoresis

Talanra. Vol. 35, No. IO, pp. 799-802, 1988 Printed in Great Britain. All rights reserved

Copyright 0

0039-9140/88 S3.00 + 0.00 1988 Pergamon Press plc

DETERMINATION OF AMMONIUM ION IN SEA-WATER BY CAPILLARY ISOTACHOPHORESIS KEIICHI FUKUSHI Kobe University of Mercantile Marine, Fukae, Higashinada, Kobe 658, Japan KAZUO HIIRO Government

Industrial Research Institute, Osaka, Midorigaoka,

Ikeda, Osaka 563, Japan

(Receiued 15 March 1988. Accepted 2 May 1988) Summary-A new procedure for determination of ammonium ion in sea-water by means of capillary isotachophoresis and a gas-liquid separator with a tubular microporous polytetrafluoroethylene membrane for preliminary enrichment has been developed. Ammonia generated by adding sodium hydroxide solution to the sea-water samples is allowed to permeate through the membrane and then dissolve in sulphuric acid. A linear calibration graph has been obtained with artificial sea-water samples containing up to 3OOyg/l. ammonium ion. The method has been applied to the determination of ammonium ion in surface and bottom sea-water samples.

It is important to determine the concentration of ammonium ion in sea-water, because the ion is closely related to marine-life production.’ The indophenol blue method*-’ is widely applied for the purpose, but is complex and time-consuming and subject to several interferences. Recently, separation with a gas-permeable membrane has been applied to the determination of volatile substances in water samples, and also to the determination of ammonium ion in river and lakewaters,*-lo blood and urine,“~‘* and condensed steam and boiler feed-water,13 but has been little applied to the determination of ammonium ion in sea-water.r4,r5 There is no report of the analysis of sea-water by capillary isotachophoresis with use of a gas-permeable membrane for prior separation. We have already described the isotachophoretic determination of sulphide,r6 total carbon dioxide” and bromide’* in sea-water after preliminary enrichment by use of a gas-liquid separator with a tubular microporous polytetrafluoroethylene (PTFE) membrane, and now report extension of the technique to the determination of ammonium ion in sea-water.

EXPERIMENTAL Apparatus

A Shimadzu IP-2A isotachophoretic analyser was used, with a potential gradient detector. The main column was a fluorinated ethylene-propylene (FEP) copolymer tube, 15 cm long, 0.5 mm inner diameter; the precolumn was a PTFE tube, 30 cm long, 1.0 mm inner diameter. A Hamilton 1725-N microsyringe was used for the injection of samples. Two kinds of gas-liquid separator were used. One was the double-tube structure used in our previous studies;‘“‘8 it consisted of an inner microporous PTFE tube (20 cm long, 2.0 mm i.d., 2.8 mm o.d.) and an outer glass tube (3.5 mm i.d., 6.0 mm o.d.). The second was similar, but had an inner

microporous PTFE tube (100 cm long, 1.Omm id., I .8 mm o.d.) and an outer FEP tube (2.4 mm i.d., 3.2 mm o.d.). Both PTFE tubes (Japan Goretex Inc.) had maximum pore size of 2 pm, and 50% porosity. The flow system is shown in Fig. 1. The gas-liquid separators, wound round a cylindrical PVC former, were immersed in a TokyoRikakikai SB-35 water-bath. The sample solution (the pH of which was raised by addition of sodium hydroxide solution) was circulated through the outer tube; sulphuric acid was circulated through the inner PTFE tube by means of an Atto SJ-1220 peristaltic pump. The temperature, pH, salinity and dissolved oxygen (DO) content of sea-water were measured as already described.i6’* Sea-water samples were filtered through a 0.45~pm membrane and stored in polyethylene bottles inside a refrigerator as soon as possible after the sample was collected.‘* Reagents

All solutions were prepared from analytical reagent grade chemicals. 18-Crown-6 was obtained from the Aldrich Chemical Co. Standard solutions of ammonium ion were prepared by dissolving ammonium chloride in water from a Yamato-Kagaku WA-22 automatic still and a Nihon Millipore-Kogyo Milli QII system. The preparation of the artificial sea-water was based on a Japanese Standard.” Procedure

Sea-water samples were analysed by the following procedure as soon as possible after collection. Add 0.5 ml of 5M sodium hydroxide to 45 ml of sea-water sample to convert ammonium ion into ammonia. Centrifuge the suspension at 3000 rpm for 5 min to remove the magnesium hydroxide precipitate. With the water-bath temperature set at 60”, circulate 40 ml of the supematant solution through the outer FEP tube and 2.0 ml of 2.0 x 10m4Msulphuric acid through the inner PTFE tube, both at a flow-rate of 6 ml/min, for 20 min. Ammonia permeates through the wall of the inner PTFE tube and dissolves in the suIphuric acid. Pump out the sulnhuric acid and iniect 150 ~1 of it into the isotachophoretic analyser. Maintain the migration current at 200 PA for the first 16 min and then reduce it to 50 PA. As leading electrolyte use 5mM hydrochloric acid/2mM 18-crown-6/0.01% Triton X-100, and as terminating electrolyte 1OmM lithium chloride/O.Ol% Triton X-100. Prepare a calibration graph by applying the method to synthetic standards.

KENHI FUKUSHIand KAZUOHIIRO

800

Peristaltic pump Gas-liquid sepoxhx \

Water-bath $

r$cehrlarideSample r

230”

0 Flow-rate,

Fig. 1. Schematic diagram of flow system.

ml/ min

Fig. 3. Effect of the flow-rate of sample solution.

RESULTSAND

DISCUSSION

Electrolyte system

It was necessary to select an electrolyte system by which the ammonium ion could be isotachophoretically separated from the potassium ion present as an. impurity in the reagents. We have already described” how the isotachophoretic separation of low concentrations of these ions can be achieved with a leading electrolyte containing l-3 mM I&crownd, with linear response for ammonium ion up to 1.Omg/l..

range 0.3-0.7 ml). During the circulation, magnesium hydroxide precipitated when the volume of alkali added was below the range above, because significant amounts of the magnesium ion remained in solution, and calcium carbonate precipitated when volumes >0.7 ml were used. Therefore, 0.5 ml of 5M sodium hydroxide was selected as the volume to be added to the 45 ml of sample solution. Concentration of sulphuric acid

Artificial sea-water samples containing 100 pug/l. ammonium ion were analysed by use of both the gas-liquid separators described above. The recovery of ammonium ion was 20% when the older (glass) type was used and 69% with the modified (FEP) type, i.e., the increase in the surface area of the inner PTFE tube resulted in a corresponding increase in the recovery factor. The modified gas-liquid separator was, therefore, used in all subsequent experiments.

The concentration of the sulphuric acid was varied in the range 0.5-3.0 x 10m4M. The zone length for ammonium ion increased with the sulphuric acid concentration, but only slightly when this was > 1.0 x 10w4M, as shown in Fig. 2. The isotachophoretic measurement took longer when the sulphuric acid concentration was higher than 2.5 x 10m4M, because of the excess of hydrogen ion in the treated solution. Therefore, 2.0 x 10m4M was adopted as the optimum sulphuric acid concentr+Qion.

Volume of sodium hydroxide solution

Flow-rate of sample circulation

The zone length for ammonium ion in the isotachopherograms was almost constant irrespective of the volume of 5M sodium hydroxide added (over the

The flow-rate of sample solution was varied in the range 2-10 ml/min. The zone length for ammonium

Gas-liquid separator

0'

0.5

I

1.0

1

1.5 HzSO,,

I

I

20

2.5 1O-L M

I

3.0

Fig. 2. Effect of sulphuric acid concentration.

lj--_-10

15

20

Circulation

25

30

time, min

Fig. 4. Effect of circulation time.

Determination of ammonium ion in sea-water

‘fL---30

LO

Temperature

50

60

801

y = 0.395x - 0.8 (correlation coefficient 1.OOO)where x is the ammonium concentration in pg/l. and y the zone length in mm. The performance was estimated from the variation in zone length per 1.0 pg/I., calculated for each point on the calibration graph. The standard deviation was found to be 0.033 mm (n = 12). The lower determination limit for ammonium ion was calculated by substituting 0.1 mm for y in the regression equation, and was taken as 0.25 pg/l. The recovery of ammonium ion, at 67 f 3%, was less than complete, but this does not matter in practice, because the recovery is almost constant in the range O-300 pg/l. It may, however, be increased by use of a longer microporous PTFE tube. In addition, a similar calibration was obtained for sea-water samples appropriately diluted with artificial sea-water. The regression equation of this graph was y = 0.379x + 0.6 (correlation coefficient 0.999).

70

of water-bath, ‘C

Fig. 5. Effecctof temperature.

ion increased with flow-rate up to 4 ml/min, but then almost levelled off, as shown in Fig. 3. At higher flow-rate, sample solution as well as ammonia is liable to permeate through the wall of the inner PTFE tube, so 6 ml/min was adopted as the optimum flow-rate for sample circulation. Duration of sample circulation

Analysis of sea -water samples

The proposed method was applied to the determination of ammonium ion in surface and bottom sea-water samples collected around the coastal area of Osaka Bay between Port of Amagasaki and Rokko Island on 6 February 1988. The concentrations of ammonium ion in these samples were so high that

The circulation time was varied in the range 10-30 min. The zone length for ammonium ion increased with circulation time, but not linearly, as shown in Fig. 4. To shorten the analysis time, 20 min was adopted as the circulation time. Eflect of temperature

The temperature of the water-bath was varied in the range 30-70”. The zone length for ammonium ion increased linearly with temperature up to 60”, as shown in Fig. 5, because of decreased solubility of ammonia in the alkaline solution.*’ Although the recovery of ammonia was greater when a temperature of 70” was used, 60” was preferred because temperature control was easier. Calibration graph

A linear calibration graph was obtained for 12 artificial sea-water samples containing 25-300 pg/l. ammonium ion. The regression equation was

I

I

I

I

22

23

2L

25

Time, min

Fig. 6. Isotachopherogram of sea-water sample treated by the proposed method. Q, NH: ; b, K+ ; c, Ca*+ ; d, Na+

Table 1. Results for ammonium ion in sea-water Sampling site Port of Amagasaki Port of Amagasaki Mouth of Muko river Mouth of Muko river Nishinomiya harbour Nishionmiya harbour Pond at KUMM Pond at KUMM Rokko Island Rokko Island

Depth, M 0

5.0 0 3.0 0 1.5 0 5.0 0 5.5

Temp., “C

pH

Salinity, %o

DO, mgll.

NH: found, /xgll.

10.5 10.7 9.5 9.2 10.1 9.6 9.3 9.2 9.2 9.1

7.50 7.64 7.79 7.85 7.78 7.83 7.82 7.80 8.01 8.08

15.1 24.0 26.0 26.5 25.6 26.6 22.6 26. I 27.6 27.9

7.11 7.00 10.05 9.09 8.85 8.25 9.89 7.89 10.77 9.80

3750 1740 1060 1060 1200 987 4580 1040 1010 970

Sampling date: 6 February 1988; KUMM = Kobe University of Mercantile Marine.

KEIICHI FIJKUSHI and

802

they were diluted with artificial sea-water before analysis. An isotachonheroaram of surface sea-water from the mouth of-the Muko river is shown in Fig. 6. The concentrations of ammonium ion listed in Table 1 were calculated by use of the calibration graph prepared with artificial sea-water samples. From these results, it was found that eutrophication was progressing rapidly in these sea-areas. Tanaka et aL9 found that methylamine interfered positively with the determination of ammonium ion by the automated measuring apparatus with a gaspermeable membrane. In the proposed method, this interference can be eliminated because methylammonium, dimethylammonium and trimethylammonium ions are separated from the ammonium ion by isotachophoresis with the electrolyte system described above.** For the determination of ammonium ion in seawater, the proposed method is simple, has a high sensitivity and precision and no interferences. Acknowledgemenfs-The authors express their gratitude to Drs. E. Sekido, Y. Masuda and T. Tanaka for their kind encouragement and suggestions, and to Mr. S. Miyamichi for his experimental help.

REFERENCES 1. S. Tsunogai and S. Noriki, in Kaiyo-Kaguku, M. Nishimura (ed.), p. 106. Sangyo-Tosho, Tokyo, 1986. 2. K. Matsunaga and M. Nishimura, Bunseki Kaguku, 1971, 20, 993.

Kazuo

HIIRO

3. J. D. H. Strickland and T. R. Parsons, A Practical Analysis, 2nd Ed., p. 87. Fisheries Research Board of Canada, Ottawa, 1972. F. Koroleff, in Merhoats of Seawarer Analysis, 2nd Ed., K. Grasshoff, M. Ehrhardt and K. Kremlinn feds.). n. 150. Verlag Chemie, Weinheim, 1983. Y. Miyake and Y. Kitano, Shin-Suishirsu-KagakuBunseki-Ho, p. 135. Chijin-Shokan. Tokyo, 1985. M. A. Brzezinski, Mar. Chem., 1987, 20; 277. G. Catalano. ibid. 1987. 20. 289. T. Aoki, S. Uemura and M. Munemori, Anal. Chem., Handbook ofSeowarer

4. 5.

6. 7. 8.

1983, 55, 1620.

9. T. Tanaka, A. Kawahara, S. Wakida and K. Hiiro, Kankyo Gijursu, 1986, 15,46. 10. R. Nakata and H. Sakashita, Proc. 3Srh Annual Meeting Japan Society for Analytical Chemisrrv, Okavama, 11-14 October i986, p. 556. Anal. Chem.. 11. Y. M. Fraticelli and M. E. Meverhoff. < 1981, 53, 992. 12. T. Aoki, S. Uemura and M. Munemori, Bunseki Kagaku,

1984, 33, 505.

13. D. Midgley and K. Torrance, Analyst, 1972, 97, 626. 14. T. R. Gilbert and A. M. Clay, Anal. Chem., 1973,4S, 1757. 15. Y. Takase, H. Hashitani and M. Okumura, Proc. Symposium Chemical Society of Japan (ChugokuShikoku Brunch), Matsue, 22-23 November 1987,

16. 17. 18. 19.

p. 228. K. Fukushi and K. Hiiro, J. Chromarog., 1987,393,433. I&m, Z. Anal. Chem., 1987, 328, 247. I&m, Bunseki Kagaku, 1987, 36, 712. Test Method for Rust-Preventing Characrerisrics of Lubricaring Oil, JIS K 2510-1987, p. 7. Japanese

Standards Association, Tokyo, 1987. 20. K. Fukushi and K. Hiiro, Talanra, 1988, 35, 55. 21. W. E. van der Linden, Anal. Chim. Acre, 1983,151,359. 3rd Ed., p. 460. Chemical Society of 22. Kagaku-Binran,

Japan, Tokyo, 1984.