Provo, Utah 84602 resolved in about 20 mm under standard. IC conditions. Here, we describe a modified ion-chromatographic system. (2), which shortens the ...
Table 1. Mg2 and Ca2 Concentrations in Filtrates of Reconstituted Serum as a Function of pH, as Determined by AAS and by Modified IC AAS
IC (proced.
IC (proced.
a)
Ca2 SD of method pH
±0.5 mg/I
6.43 5.28
16
87
19
106 122 121 122
1.42
21
1.17 1.05
21 22
a
±1 mg/L
SD of method ±0.4 mg/L ±0.8 mg/I ConcentratIon, mg/I
17.6
in about
IC conditions.
20 mm under
Here,
Mg2+ Ca2+ SD of method ±0.2 mg/I ±0.8 mg/I
90
16.7
89
20.8
118
19.8
110
21.7 20.8 20.9
124 123 117k
21.7 19.9 21.6
127 113 115k
standard
we describe
a
modified ion-chromatographic system (2), which shortens the time required to resolve Mg2+ and Ca2+. In addition, concentrations of these cations in serum as determined by IC are compared with results obtained by atomic absorption spectroscopy (AAS). A Model 10 Ion Chromatograph (Dionex Corp., Sunnyvale, CA 94086) was equipped with standard 3 X 150 mm pre-column and 6 X 250 mm cation separator column, both in the Ba2+ form. Two different eluents were tested: (a) 1.0 mmol/L aqueous Ba(N03)2 and (b) 1.0 mmol/L Ba(N03)2 in 0.1 mmol/L HNO3. We used flow rate of 138 mL/h (30% pump rate). For both eluents the suppressor was a 6 X 250 mm column containing Bio-Rad AG1-X 10 cationexchange resin (SO42- form). With the second eluent, a 3 X 150 mm post-column containing Bio-Rad 50W-X16 cation-exchange resin (H+ form) followed the suppressor column. A 0.1-mL sample of a 10-fold dilution of serum ultrafiltrate was injected onto the columns and the chromatogram was recorded with a strip-chart recorder.
Measurements
were made in triplicate
(AAS) or duplicate (IC). Serum specimens were ultrafiltered to prevent clogging of IC columns. We acidified the sample with HC1 before ultrafiltration, to ensure release of calcium and magnesium from protein. Total calcium and magnesium were accounted for in the ultrafiltrate at pH 1.5 and below (Table 1). Centriflo membrane cones (CF-25; Amicon Corp., Lexington, MA 02173) which retain molecules of Mr 25000 and greater were soaked in redistilled water for 1 h and dried by centrifugation. We applied 5 mL of serum to the cones and obtained 1 to 2 mL of ultrafiltrate after 5 mm of centrifugation. For this preliminary study, we used reconstituted human lyophilized serumm (Dade MoniTrol#{174}). Under the IC conditions described, the Mg2 and Ca2 peaks eluted at 5.6 and 8.1 mm, respectively, or about 0.2
J. Rehfeld Hans F. Loken
Selwyn
b)
Glin. V.A. 4150 San
Pat hol. Service Med. Center Clement St. Francisco, CA 94121
Francis
Single measurement.
resolved
by ion chromatography employing sulfatesuppressed barium and lead eluent. Anal. Chem., submitted for publication.
mm later when the post-column was used, with baseline resolution. The peaks were sharp and peak height reliably indicated concentration over a wide range (0.2 to 20 mg/L for Mg2 and 0.3 to 30 mg/L for Ca2). Monovalent cations such as Na+ and K+ eluted in a single large peak before Mg2 and Ca2 and did not interfere with the divalent peaks. The limit of detectability of this method is 80 .tgfL for Mg2 and 180 g/L for Ca2, with the first eluent, 10 and 24 tg/L, respectively, for the second eluent (2). Table 1 compares the Mg2+ and Ca2+ concentrations as determined by the modified IC techniques and by AAS Model 603 (Perkin-Elmer, Norwalk, CT 06856). Agreement of results for Mg2 is especially good. Note that IC reproducibly gives greater precision for Mg2 than does AAS. Agreement for Ca2+ between AAS and IC is also good, although a shifting baseline probably accounts for the poorer reproducibility of Ca2+ peaks with the second eluent. One potential source of interference with the modified IC technique is the close elution times of Ca2+ and Zn2+, but the agreement between IC and AAS indicates that this interference, if any, is too small to detect. Of the two IC eluent systems used, sensitivity is superior for the second eluent, with use of the post-column. However, for the concentrations of Mg2+ and Ca2+ normally found in serum this added sensitivity is unnecessary. The first eluent gives better precision without requiring use of a post column. Our results lead us to believe that IC will soon be attractive as a rapid procedure for measurement of both Ca and Mg in the clinical laboratory. This study was supported in part by the Veterans Administration. References 1. Anderson, C., Ion chromatography: A new technique for clinical chemistry. Glin. Ghem. 22, 1424-1426 (1976). 2. Nordmeyer, F. G., Hansen, L. D., Batough, D. J., et a]., Determination of alkaline earth and divalent transition metal cations
R. Nordmeyer John D. Lamb
Dept. of Chemistry Thermochemical Institute Brigham Young University Provo, Utah 84602
An improved Automated Method for Serum Chloride To the Editor: The widely used mercuric thiocyanate method (SCN) for the automated determination of serum chloride (1) suffers from three major defects: it is not linear, mercury is discharged to waste, and the very much higher sensitivity for iodide and bromide can seriously interfere with determination of electrolyte balance. The automated method of Feldkamp et al. (2) involving mercuric 2,4,6-tripyridyl-s-triazine (TPTZ) as proposed by Fried et al. is excessively noisy owing to the very high dilution required by this sensitive reagent, is critically dependent on mercuric nitrate flow, and does not decrease the amount of mercury used. The method, however, is linear and neither iodide nor bromide react preferentially.
We have accordingly
modified
the
above automated method for the Technicon SMAC and AutoAnalyzer II, to improve the shortcomings of the method. In our technique (Figure 1) two dialyzers are used in series for dilutions of 1 in 3000-4000 on AutoAnalyzer II and 1 in 600 on SMAC, thereby avoiding all the difficulties involved in conventional continuous-flow techniques involving high dilution. Reagents. Reagents for the standard SCN method on AAII and SMAC were obtained commercially (Technicon Chemicals Co., Orcq, Belgium): 0.67 g of TPTZ (Sigma Chemical Co., St. Louis, MO), dissolved in 5 mL of 1 mol/L HNO3, was added to 400 mL of a solution of 0.78 g of pure ammonium sulfate (FeSO4 - (NH4)2SO4 . 6HO; E. Merck A.G., Darmstadt, F.R.G.) per litre of de-ionized water. The absorbance of the TPTZ solution was adjusted to 1 A600 by dropwise addition of a 0.2 mol/L mercuric nitrate solution. After addition of 4 mL of Brij-35 surfactant (30% solution; Technicon) the solution was diluted to 2 L with de-ionized water.
CLINICALCHEMISTRY,Vol. 26, No. 8,
1980
1233
SAMPLER 150/hour
orn/orn Waste
Waste
I 10 turns
if
II
2Oturns1
I
FIowceII and PnINTER
0.Sxl0ttyn
Fig. 1. SMAC manifold
14.3di01yZ
38 uI/mm.
385
I
uI/mm.
90/miri. red/red
C merane
482 uI/mm.
I 90/mm. _____________________ C mera
[ __________________________________
600rwn
90/mm. orn/grn
wht/wht
1 type
287 uI/mm.
blk/blk
226
uI/mm.
smple
0 I
diluent
predi
air
IutiOfl
6:1
snple
recipient
(1)
air recipient
(2)
air TPTZ
reagent
Ior TPTZ chloride method
Sample diluent = 0.25 mol/L HNO3 Recipients = de-lonized water with 1 mL Brij 35 (300 mLIL solution) per liter AutoAnalyzer II manifold is exactly similar, except that a 12- and 6-inch dialyzer are used and pump lines are (in mL/min flow rate): I = 1.20, 2 0.60, 3 = 0.23. 4 = 2.00. 5 = 0.75. 6 = 2.00, 7 = 0.70, 8 = 1.00. The 20-turn coil is replaced with a 70-turn Coil. Sampling rate is 60/h, with a 5:1 sample/wash time rate
The reagents are stable for at least one month at room temperature. Sodium iodide, chloride, fluoride (E. Merck A.G., Darmstadt, W. Germany), and bromide (Brocades, Delft, The Netherlands) were of “pure” grade. Bovine albumin was from Poviet Production B.V., Oss, The Netherlands. Standards. To avoid any protein interference with dialysis, we prepared all standards in a 50 gIL solution of albumin. Equipment. SMAC and AutoAnalyzer II were standard Technicon equipment, except that we used a sampler specially developed in this laboratory to improve timing and reduce carryover, by use of a microprobe. The SMAC computer program was not changed. Procedure. The published thiocyanate procedure (1) was compared on both SMAC and AutoAnalyzer II with the TPTZ method with use of the man#{182}‘Reco,der Units bramde iodide
chloride
20 rrrnol/l
40
60
80
100
120
ifold and reagents shown in Figure 1. Sampling rate on SMAC remained at 150/h and 60/h on the AAII with the
TPTZ method. Results. Contrary to the report of Feldkamp et al., we found a definite difference between the analog response of chloride and iodide or bromide (Figure 2), although the difference is clinically insigificant in any calculation of electrolyte balance. Of perhaps greater importance is the linearity of the chloride response on SMAC and AutoAnalyzer II with use of TPTZ. On the other hand, the curves were of the same good quality as obtained with the thiocyanate method. Normal values obtained with SMAC and TPTZ for 216 supposedly healthy male blood donors between 22 and 56 years old was 99.3 (SD 4.8) mmol/L while for 183 women of similar age the value was 100.3 (SD 4.3) mmol/L. We saw no significant difference in precision with AAII between thiocyanate and TPTZ method (20 samples each day for 20 days, range 90-115 mmol/L), but the SMAC precision was markedly improved by using TPTZ (within-run CV = 0.61%, between-run CV = 0.79% vs SCN within-run = 1.90%). Drift was the same in the case of AAII with SCN and TPTZ but less with TPTZ (1.0 mmol/L per 48 samples) than SCN (1.5 mmol/L per 48 samples) on SMAC. Carryover also was improved only on SMAC (TPTZ 2.2%, SCN 3.8%). Mercury consumption with TPTZ was less than half that of SCN. We have operated this method routinely now for three years on the AAII and 18 months on the SMAC without any difficulty.
halogen
Fig. 2. Linearity of response for bromide,
References
iodide, and chloride
1. Method
1234
CLINICAL
CHEMISTRY, Vol. 26. No. 8, 1980
lyzerII, 1974, Technicon Tarrytown, NY 10591.
Instruments
2. Feldkamp, C. S., Palmer, D. J., Salancy, J. A., and Zak, B., Interference by other halides in the automation of chloride. Contributions to the general methodology of continuous flow systems. J. Gun. Chem. Gun. Bioche,n. 4, 146-150
(1974).
3. Fried, R., Hoefimayr, J., and Vel#{246}sy, G., A new highly sensitive method for the determination of chloride in body fluids without protein precipitation. J. Glin. Chem. Gun. Biochem. 6, 280 (1972).
E.B.M.deJong H. M. J. Goldschmidt A. C. C. M. van Aiphen J. A. Loog St. Elisabeth Hosp., Dept. of Clin. Chem. Jan van Beverwijchstraat 2a, 500 LC Tilburg, The Netherlands
NATIONAL IN CLINICAL Certification
REGISTRY CHEMISTRY
of:
Clinical Chemists Clinical Chemistry
Technologists
For free brochure
anti application
materials contact: Executive Secretary 1155 Sixteenth Street, NW Washington, D. C. 20036 Phone: (202) 659-9660
No. SE. 4-0005 FD4, AutoAna-
Corp.,
