Separation of heavy metals by high-speed countercurrent

Separation of Heavy Metals by High-Speed Countercurrent. Chromatography. Eiichi Kitazume* and Nobuyoshi Sato. College of Humanities and Social Science...
1 downloads 0 Views 355KB Size
Anal. Chem. 1993, 65, 2225-2228

2225

Separation of Heavy Metals by High-speed Countercurrent Chromatography Eiichi Kitamme*and Nobuyoshi Sat0 College of Humanities and Social Sciences, Iwate University, Ueda, Morioka, Iwate, 020 Japan

Yoshihiro Saito Analytical Instruments Research Laboratory, Shimadzu Corporation, 1, Nishinkyo Kuwahara-cho, Nakagyo-ku, Kyoto, 604 Japan

Yoichiro Ito Laboratory of Biophysical Chemistry, National Heart, Lung, and Blood Institute, National Institutes of Health, Building 10, Room 7N322, Bethesda, Maryland 20892

Separation of heavy metal elements such as Co, Cu, Fe(II),Fe(III),Mg, andNi was performed with a high-speedCCC coil planet centrifuge equipped with a multilayer coil. The two-phase solvent systems used for the separation were composed of m-heptane containing bis(2-ethylhexyl) phosphoric acid (stationary phase) and diluted citric acid (mobile phase), where the distribution ratios of the elements can be optimized by selectinga proper acid concentration. The mobile phase was eluted through the column at a flow rate of 5.0 mL/min while the apparatus was rotated at 800 rpm. Continuous detection of the metal elements was effected by measurement of the emissionintensity with direct current plasma atomic emission spectrometry (DCP-AES),where the elution curve was obtained by on-line monitoring of each emission signal. Each metal element was well resolved at high partition efficiencies ranging from 200 to 3600 theoretical plates.

INTRODUCTION Countercurrent chromatography (CCC)is a useful method of separating many organic materials, such as biologically active substances,natural and synthetic peptides, and various plant hormones. Also it has been applied to separation of inorganic elementa.1*2 Recently, high-speed countercurrent chromatography (HSCCC) has been widely applied to the separation of various organic materials.= Ita advantages over other countercurrent chromatographic methods include high partition efficiency, speedy separation, and versatility in performing liquid-liquid dual CCC and foam CCC. Like other CCC methods, it is also free from problems arising from the use of solid support, such as adsorption or irreversible binding and contamination ofthe sample. In addition, highly (1)Kabasawa, Y.; Tanimura, T.; Tamura, Z. Bumeki Kagaku 1974,

23,89-90. (2) Ferreira, H. A.; Braga, G. L.; Shuhama, T. Rev. Bras. Farm. 1974, 55,83-88. (3) Ito, Y.CRC Crit. Rev. Anal. Chem. 1986,17,66-143. (4) Ito, Y. Countercurrent Chromatography; Marcel Dekker: New York, 1988; Chapter 3. (6) Ito, Y.; Sandlin, J.; Bowers, W. G. J. Chromatogr. 1982,244,247-

268.

efficient chromatographic separation has been achieved using a set of three multilayer coils.6-8 In spite of these advantages, there have been no applicationa of HSCCC to the separation of inorganic elements until quite recently. Radioactive calcium and strontium were separated by Zolotov et al.9 In previous reports we have demonstrated the efficient separation of the adjacent 14lanthanide elements in a single run with an exponential gradient elution10 and also determined the trace impurities in rare earth elements using HSCCC.11 In the present paper, the capability of the single-column HSCCC centrifuge was demonstrated in the separation of several metals including divalent and trivalent irons. Under optimum conditions, excellent separation of metal elements was achieved in an isocratic elution.

EXPERIMENTAL SECTION Apparatus. A Shimadzu countercurrent chromatograph (HSCCC-lA, prototype) was used which holds a multilayer coil separation column on the rotary frame at a distance of 10.0 cm from the central axis of the centrifuge. The column was prepared from a single piece of approximately 150-m-long, 1.60-mm4.d. poly(tetrafluoroethy1ene) (PTFE) tubing by winding it directly onto the holder hub (10-cmdiameter). The B value, an importaut parameter used to determine the hydrodynamic distribution of the two solvent phases in the rotating coil, ranged from 0.5 at the internal terminal to 0.6 at the external terminal. (B = r/R,where r is the distance from the column holder axis to the centrifuge.) The coil capacity is approximately 300 mL. The HSCCC can be operated at a speed of 800 rpm. The flow diagram of the experimentalassembly is illustrated in Figure 1. A Shimadzu LC-6A pump was used to pump the mobile phase; a stream splitter was used for delivering an appropriateamount of effluenttothe direct currentplasma (DCP) via a peristaltic pump equipped in the DCP. The DCP atomic emission spectrometry(AES)system is a SpectraMetricsModel SpectraSpanIIIB system with 20 fiied-wavelengthchannels for observation of the elution profile. To facilitate monitoring the (6) Ito, Y.; Oka, H.; Slemp, J. L. J. Chromatogr. 1989,475, 219-227. (7) Ito, Y.; Oka, H.; b e , Y.-W. J. Chromatogr. 1990, 498, 169-178.

(8)Ito, Y.; Oka, H.; Kitazume, E.; Bhatnagar, M.; Lee, Y.-W. J. Liq. Chromatogr. 1990,13, 2329-2349. (9) Zolotov,Yu. A.; Spivakov,B. Ya.; Maryutina,T.A.;Bashlov,V.L.; Pavlenko, I. V. Z . Anul. Chem. 1989,335,938-944. (10) Kitazume, E.; Bhatnagar, M.; Ito, Y. J. Chromatogr. 1991,538, 133-140.

(11) Kitazume, E.; Bhatnagar, M.; Ito, Y. Proceedings, International Trace Analysis Symposium, Sendai, Japan, 1990, pp 103-110.

0003-2700/B3/0365-2225$04.00/0 0 1003 American Chemical Society

2228

ANALYTICAL CHEMISTRY, VOL. 65, NO. 17, SEPTEMBER 1, 1993

d Rinter

Column

ri Digital Multimeter Waste

Mobile Phase

Station

Plotter

Figure 1. Flow diagram of Instrumentation assembly.

Table I. Wavelengths Used for Present Work element line (nm) element Cd co

cu Fe

226.5 345.3 324.7 259.9

Mg Mn Ni

Measurement of Distribution Ratio. The distributionratio, line (nm)

KT,of each metal element was obtained using a simple test tube

280.2 257.6 341.4

KT = (AT - &)/A,

elution profile of the single element, an analog recorder signal from the DCP system was converted into a digital signal using a Hewlett-Packard Model 3478A digital multimeter. The digital data were stored in a Hewlett-Packard Vectra-D work station through an HP-IB bus, and the elution profile was plotted using a Hewlett-Packard 7440A graphics plotter and Advanced Graphics Software Slidewrite Plus. For simultaneous multielement measurements, the emission signal for each channel was integrated for 10s at intervals of 20 a, and each set of integrated data was printed out. The data were entered manually into the work station and the elution profie was plotted. Wavelengths used for the present work are listed in Table I. Reagents. For the preparation of the two-phase solvent system, analytical reagent grade n-heptane and citric acid were purchased from Kanto Chemical Co., Inc., Tokyo, Japan, and bis(2-ethylhexyl)phosphoric acid (DEHPA)was purchased from Nacalai Tesque Inc., Kyoto, Japan. Acid treatmentlOJ1was not necessary for DEHPA in the present work. Samples used in the present study include Co(NO&6H20 and Ni(NOd2.6H20 (Kanto Chemical), FeC12.4H20 and FeC4.6H20 (Merk, Darmstadt, Germany), and CuS04.5H20 (Kojima Kagaku Co., Ltd. Tokyo, Japan). Other metal standards were obtained as 0.1 M nitric acid solutions (lo00 ppm) from Kanto Chemical. Preparation of Two-Phase Solvent and SampleSolutions. DEHPA was dissolved in n-heptane (stationary phase). Citric acid was dissolved in deionized water (mobile phase). Sample solutions were prepared by dissolving various nitride, chloride, and sulfate in the aqueous citric acid. Procedure. Eachseparation was initiated by filling the entire column with the stationary phase followed by sample injection through the samplingport (100rL). Then, the mobile phase was eluted through the column at a flow rate of 4.5 or 5 mL/min in a head to tail elution mode while the apparatus was rotated at 800 rpm. The elution curve was obtained by means of on-line monitoring of the effluent using a DCP emission spectrometer. The effluent from the column was divided into two streams with a tee adapter: a peristaltic pump equipped in the DCP emission spectrometer was used to pump the effluent into the DCP at a flow rate of 1.2 mL/min, while the rest of the effluent may be fractionated with a fraction collector or discarded.

method (1)

where AT is the total concentration of the sample in the lower phase before equilibration with the upper phase; AL, the concentration of the sample in the lower phase after equilibration; AT- AL,the concentration of the sample in the upper phase after equilibration. The distribution ratio, KE,was also calculated from the elution curve using the following equation described previous1y:"J

Measurement of Partition Efficiency. Partition efficiencies in terms of theoretical plate numbers (N) and peak resolution (R,) were computed from the obtained chromatograph according to the following equations:lO N = (4R/W2 R, = 2(R1 -RJ/(W1 + Wz)

(3)

(4)

RESULTS AND DISCUSSION The distribution ratios of the metal ions were measured at various sample concentration using 0.0024.2 M DEHPA in n-heptane as the upper phase and 7-140 mM citric acid as the lower phase. The results are shown in Figure 2, and the range of KT values observed is in the last column in Table 11. Figure 2 shows the effects of sample concentration on the distribution ratios of nickel, cobalt, magnesium, and copper obtained with the test tube method using 0.2 M DEHPA in n-heptane and 7 mM citric acid. Except for the nickel ion, which elutes first, the KT values of these metal ions slightly increase as the sample concentration is decreased. Similar results were obtained for the distribution ratio of copper, cadmium, and manganese when 50 mM citric acid was used as the mobile phase. Metal ions eluted by the aqueous mobile phase (diluted citric acid) were detected by their emission intensity with DCP. Figure 3 shows a typicalchromatogram of nickel, cobalt, magnesium, and copper obtained by eluting 7 mM citric acid at a flow rate of 5 mL/min. The partition efficiencies ranged from 280 TP (Co) to 200 TP (Cu).

2227

ANALYTICAL CHEMISTRY, VOL. 65, NO. 17, SEPTEMBER 1, 1993

1'

2 0

F

1.00

d

0

-....................

-..................

t

cu

T

i

-... .................................... .....................

20

10

30

50

40

1

60

70

80

90

100 0

S A M X E CONCU.ITRATICN (ppml

20

10

30

40

Fbwr 2. Effects of sample concentration on distribution ratio.

solvent system element Ni

acid, mM)

0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.002 0.002

co Mg cu

cu Cd Mn Fe(I1) Fe(II1) 4

LPb (citric

(DEHPA,M)

7 7 7 7

50 50 50

140 140

KE (av)

KT

0.02 0.15 0.37 1.76 0.09 0.69 1.10 0.04 0.20

0.00-0.02 0.074.09 0.124.25 0.96-1.30 0.114.14 0.474.64 0.81-1.44 0.014.03 0.214.40

Upper phase (heptane). b Lower phase (aqueous solution).

t

70

00

90

Flgure 4. Isocratic separation of copper, cadmium, and manganese by HSCCC. The experimental conditions were as follows: apparatus, HSCCC centrifugewlth lO.&mrevoiutlon radius; column, one multileyer coil, 150 m X 1.6 mm Ld., 300-mL capacity; stationary phase, 0.2 M MHPA in heptane; mobile phase, 50 mM citric acid; sample, each 14.1 pg: revolution, 800 rpm: flow rate, 5.0 mL/min; pressure, 140 psi. 20

-

3

J >

G 5 5

10-

6

a

z

Ni

0

Co

0

.oo

1

60

TIME (min)

Table 11. Distribution Ratio Values UPd

50

10

20

30

50

40

TIME (mid

Flgure5. Isocratic separation of iron(I1)andiron(II1) by HSCCC. The experimental condltionswere as follows: apparatus, HSCCC centrifuge with 10.0-cwevoiutlon radius; column, one multilayer coil, 150 m X 1.6 mm i.d., 300-mL capacity; stationary phase, 0.002 M DEHPA in heptane: mobile phase, 140 mM citric acid: sample, each 0.5 mg; revolution, 800 rpm; flow rate, 4.5 mL/min; pressure, 140 psi.

0.50

0.00 0

20

40

60

00

100

120

T I M (rnin)

5. Isocratlcseparationof nlckei, cobalt, magnesium, and copper by HSCCC. The experimental condltions were as follows: apparatus, HSCCC centrifugewlth lo.&radius; cdunn, one multilayer cdi, 150 m X 1.6 mm i.d., 300-mL capacity; Stationary phase, 0.2 M DEHPA in heptane; mobile phase, 7 mM cltric acid; sample, 10 pg of Ni, 20 pg each of Co and Mg, 40 pg of Cu; revolution, 800 rpm: flow rate, 5.0 mumin; pressure, 140 psi.

Cadmiumand manganesewere separated from copper using 50 mM citric acid as an eluent. Figure 4 shows a chromato-

gram of copper, cadmium and manganese. The partition efficiencies ranged from 630 TP (Cu) to 270 TP (Mn). Separation of iron(I1) and iron(II1) was achieved a t a lower DEHPA concentration in the stationary phase and a higher acid concentration in the mobile phase, where iron(II1) was strongly retained in the stationary phase (Figure 5). The partition efficiency of the separation was 3700 TP for the first peak (FeC12)and 410 TP for the second peak (Feels), while the peak resolution between the two peaks was 2.46. The great difference in the plate count for iron(I1) and iron(III)probably indicates slow kinetics involved in the distribution of iron(II1).

Table 111. Reproducibility of Peak Height of Nickel, Cobalt, Magnesium, and Copper

emission intensity

element

Ni Co

Mg Cu

32.0 53.0 76.0 10.5

29.0 64.0 17.9 10.8

27.5 30.0 54.5 52.0 77.5 75.5 10.9 11.3

27.3 56.0 19.5 11.0

mean 29.2 53.9 77.3 10.9

RSDb(%)

5.9 2.5 1.8 2.4

'Measurement of intensity (au) was performed by successive sample charge without renewing column content. b Relativestandard deviation.

Table I1 presents the average KE computed from two or three chromatograms, while KT was obtained from the test tube method. With a few exceptions, these two distribution ratio values are in a close agreement with each other. Although HSCCC is generally used for preparative separations, the method has the capability of performing efficient analytical separations especially when combined with a continuous detection system such as spectrophotometry or spectroscopy. By use of a sample injection loop set between the elution pump and the HSCCC, continuous repetitive measurements can be carried out as in high-performance liquid chromatography (HPLC). In the present work, the reproducibility of this method was studied using standard

2228

ANALYTICAL CHEMISTRY, VOL. 65, NO. 17, SEPTEMBER 1, 19Q3

solutions containing 100 ppm nickel, 200 ppm each of cobalt and magnesium, and 400ppm copper. After the steady-state hydrodynamic equilibrium between the mobile and stationary phases was established in the column, 100 pL of the sample solution was successively injected five times a t given intervals and the peak height resulting from each injection was measured. As shown in Table 111, the values are quite reproducible with small relative standard deviations except for nickel, indicating that HSCCC may be used for the quantitative analysis of microgram to milligram amounts of metal elements.

The variation for detection of nickel may be related to its sharp peak and early elution. Although there may not be any difficulties when HSCCC is coupled with a DCP-AES, the detection limits for this system are influenced by that of DCPAES. If a more sensitive detector such as ICP-MS could be coupled with HSCCC, the present system would be a highly sensitive and selective determination technique.

RECEIVED for review November 16, 1992. Accepted May 20, 1993.