- Email: [email protected]
Continuous spectrophotometric determination of copper, nickel and zinc in copper-base alloys by flow-injection analysis
Julunru. Vol. 28. pp. 389 to 393. I981 Printedin Great Britain.All rights reserved
0039-914c!8, @50389-0%0?.00~0 CopyrIght 0 1981 Pergamon Press Ltd
CONTINUOUS SPECTROPHOTOMETRIC DETERMINATION OF COPPER, NICKEL AND ZINC IN COPPER-BASE ALLOYS BY FLQ_W-INJECTION ANALYSIS R~KU&~URODA* and TADASHI
MOCHIZUKI
Laboratory for Analytical Chemistry, Faculty of Engineering, University of Chiba, Yayoi-cho, Chiba, Japan (Receired
l?LNorember
1980. Accepted
24 December
1980)
Summary-Flow-injection methods have been developed for the determination of copper, nickel and zinc in copper-base alloys, including several types of brasses, deoxidized copper, beryllium copper and German silver. The system for copper and nickel involves the measurement of the absorbance of the copper(H) and nickel(H) aquo-complexes at 805 and 410 nm, respectively, after simple dissolution of the sample in a nitric acid-phosphoric acid mixture. The system needs no further reagents and the sample solutions can be analysed at rates of up to 280 /hr for copper (or nickel in German silver) without any carry-over. The system for zinc consists of automatic dilution of the injected sample with a thiosulphateacetate buffer solution and the subsequent measurement of the absorbance of the zinc-Xylenol Orange complex at 568 nm. This system permits analysis rates of up to 9O/hr for zinc solutions. with no carry-over. The procedures have been applied to standard copper-base alloys. The results agreed satisfactorily with the certified values. The precision ranges are 0.24.7”., for copper and nickel and 0.5-0.8”; for zinc.
Since its introduction by RtiiEka Stewart in 1975, flow-injection
and Hansen and by analysis has found
wide application, especially in agricultural, environmental, clinical and pharmaceutical analyses. Recent progress has been reviewed in several articles.‘4 Despite its great potential for automatic analysis of alloys, flow-injection analysis has been little applied in this area.5 We have attempted to develop rapid procedures for the determination of copper, nickel and zinc in a variety of copper-base alloys by means of spectrophotometric flow-injection analysis. A simple flow system allows the rapid determination of copper (O-99.9%) and nickel (in German silver only) by direct measurement of the absorb#roe of their aquo-complexes. Spectrophotometric detection with Xylenol Orange (X0) permits the rapid determination of zinc, with minimum consumption’ of reagents because of the small sample size.
metal in 20 ml of 7M nitric acid and diluting exactly to 100 ml with distilled water [Zn(II) 2 mg/ml]. Masking agent-bufir solution. Ten g of sodium thiosulphate dissolved in IM acetic acid-IM sodium acetate buffer solution (pH 5.9) and diluted to 1 litre with the buffer. Xylenol Orange solution. A 0.012% aqueous solution. Apparatus
EXPERIMENTAL
Sample dissolution
Reagents Standard solutions. Standard copper solution was prepared by dissolving 1 g of pure copper metal in 20 ml of 7M nitric acid and diluting exactly to 100 ml with distilled water [Cu(II) 10 mg/ml]. Standard nickel solution was prepared by dissolving 12 g of nickel nitrate in 1.4M nitric acid and diluting to 100 ml with the same acid [Ni(II) 25 mg/ml] and standardized compleximetrically. Standard zinc solution was prepared by dissolving 0.2 g of pure zinc
*To whom correspondence should be sent.
The systems for the determination of copper, nickel and zinc are shown in Fig. 1. The systems were assembled from l.O-mm Teflon tubing except for a back-pressure coil (0.5 mm), valve and connectors. An SJ-121 IH peristaltic pump (Atto. Japan) was operated with Tygon pump tubing (l/8 or l/l6 in. bore) to obtain the desired flow-rates. A sampleinjection valve (Nihon Seimitsu NV-508-6M) made of Teflon with a Teflon sample loop (bore 1.0 mm) was used for sampling. A Shimazu Model W-180 spectrophotometer equipped with a Nihon Seimitsu NS-FC-210 flowthrough cell (volume 31.4 ~1, light-path 10 mm) was used. coupled with a Shimazu Chromatopac C-RIA data-processer to record the peaks and simultaneously calculate the peak heights or areas.
and requests for reprints 389
Determination of copper and zinc. A sample weighing about 60 mg is dissolved in a mixture of 2 ml of O.OlM phosphoric acid and 10 ml of 7M nitric acid and diluted exactly to 50 ml with distilled water. A further tenfold dilution is made for the determination of zinc. Determination ofnickel. A 1.0-g of German silver is taken and treated as for copper determination, Flow-injection
analysis
Carrier solution C (1.4M nitric acid) is pumped into the analytical line at a flow-rate of 8.0 ml/min with a peristaltic pump [Fig. l(a)]. The sample solution (318 ~1 in our apparatus) is introduced into the carrier stream by a six-way loop-valve injector V and the Copper
and nickel.
ROKURO KURODA and TADASHI MOCHIZUKI
390
(b)
P, ml / min
P, ml/min
Fig. 1. Flow diagrams. (a) System for determination of copper and nickel. C: carrier (I.4M HNO,); V: sample-injection valve; S: sample; P: peristaltic pump; D: spectrophotometer; Cs: back-pressure coil; W: waste; Ws: sample waste. (b) Automatic sample-dilution system. C,: carrier (distilled water); R,: reagent (distilled water). (c) System for the determination of zinc. Cr: carrier (0.14M HN03); RI: masking agent-buffer solution (1% Na,S,O,-acetate buffer, pH 5.9); R3: calorimetric reagent (0.012% Xylenol Orange solution); A, B: confluence points; C,,, Cm2 : mixing coils. The numerals under the tubes or coils refer to their lengths in cm and their internal diameters in mm.
absorbance of the aquo-complex is monitored in the flowthrough cell D at 805 nm for copper(H) and 410 nm for nickel(II), respectively, against the carrier solution as reference. A narrow coil Cs (50 cm long) is attached to the exit of the flow-through cell, as a back-pressure coil. Zinc. Carrier solution C2 (0.14M nitric acid) is pumped into the analytical line at a flow-rate of 2.0 ml/min and the sample solution S (148 ~1) is introduced into the carrier solution by the injection valve V [Fig. l(c)]. The sample slug meets the masking agent-buffer solution R, at the point A. In the mixing coil C,, the sample is diluted, copper(I1) and lead(I1) being masked at the same time. The mixture and the calorimetric reagent flow R, (0.012”/;, Xyleno1 Orange solution) merge at the point B. The colour reaction proceeds in the mixing coil Cm2, and the absorbance of the chelate is measured at 568 nm against the waste solution containing no sample constituents. For calibration, similarly treated standard solutions are injected into the line between the sample runs. The area or the height of the absorbance peak can be used for the calibration.
50
40
IyA
z t
I
OO RESULTS AND DISCUSSION For
rapid
continuous
spectrophotometric
analysis
of the alloys for the major components the preparation of the sample solution should be as simple as possible, and the same solution should be suitable for both the copper and the zinc determination. For this purpose a mixture of nitric and phosphoric acids is satisfactory, yielding no precipitation of hydrated
I 0
5 Flow - rota
lo
,
ml / min
2Oo Injection volume
, pl.
Fig. 2. Etfects of flow-rate and injection volume on peak height-aquo-complex method. Sample concentration: A: Cu(II), 3 mg/ml; B: Cu(II), 2 mg/ml. Sample injection volume: 318 ~1 for A and B. C: Cu(II), 2 mg/ml. Flow-rate of carrier keot constant at 8.0 ml/min.
Continuous spectrophotometric
determination
stannic oxide and providing effective masking of iron with phosphate. Copper(H) aquo-complexes exhibit an absorption maximum at 8OS810 nm, with molar absorptivity 12.9 l.mole- ’ .cm- ‘. Such a low sensitivity is well suited to the determination of copper in copper-base alloy@. The absorbance is not affected by variation in the nitric acid concentration. The nickel(H) aquo-complexes also absorb at 805 nm to some extent, causing a 7% positive error when present
at the same w/v concentration as the copper(H). The nickel(I1) has an absorption maximum at 393 nm, but the nitrous acid produced in the dissolution process interferes at this wavelength so the absorbance is measured at 410 nm, at which the molar absorptivity is 3.8 l.mole- 1.cm-I, providing a basis for the determination of nickel in German silver. Xylenol Orange reacts with zinc to form a coloured complex,’ which exhibits an absorption maximum at 568 nm. The chelate is best formed at pH 5.7-5.8. Copper(I1) and lead(U) interfere, but they are effectively masked with thiosulphate. However, nickel(H) still interferes seriously, and the method cannot be applied to nickel alloys unless a suitable masking agent can be found. Optimization
offlow
0.2%, a45 0.63 ‘9 \ \
0.49
:a, 0.04
a45
\
.--a 0.43
&i_
0.09 -. ow----,g --._ --._ 6, a04 --oy6
0.04 011
aoi-----o 0.13
01
0
eo-
z
- 60E .o e x e a 40-
20-
01 0
loo
xx)
330
lr@ction volume, pl. Fig. 4. Effect of injection volume-zinc-Xylenol Orange method. Sample concentration: A: Zn(II) 40 pg/ml; B: Zn(II) 20 pg/ml.
system
The effects of flow-rate and injection volume on the peak height are shown in Fig. 2 for the determination of copper. Flow-rates greater than 5.0 ml/min give a constant absorbance peak-height for an injection
-
391
of copper, nickel and zinc
1
I
lo s Ratio of flow-rotor, u/x
1
IS
Fig. 3. EtTect of flow-rate of carrier and reagent solutionszinc-Xylenol Orange method. x, y: respective flow-rates of carrier and reagent solutions; mixing coil length, z: 100 cm (U), 300 cm (A), 500 cm (0); total Row-rate: 12.0 ml/min (-), 8.0 ml/min (---). The numerals at each point indicate the standard deviation.
volume of 318 ~1. The injection volume has a significant effect, yielding increased peak height and reproducibility with increasing injection volume. However, increasing the injection volume widens the sample zones and lowers the sampling rate, so that 318 ~1 is taken as a compromise. This allows analysis of 280 copper solutions per hour, without any carry-over. The optimum conditions for nickel are the same as those for copper. The Xylenol Orange method is too sensitive to be applied directly to the analysis for zinc. To simplify the sample dilution we decided to dilute the sample within the flow system with a combined masking agent-buffer solution, as illustrated in Fig. l(b). TO examine the effectiveness of this dilution system, the 0.01% Xylenol Orange solution (78 ~1) was introduced by the sample-injection valve V into the analytical line, where the total flow rate (x + y) was kept constant at either 8.0 or 12.0 ml/min and the ratio of flow-rates y/x 0, is the flow-rate of the diluent) and the length of coil C,, were varied over the ranges $15 and 100-500 cm, respectively. The results are illustrated in Fig. 3. Increasing either the coil length or the dilution ratio lowered the peak height but improved the reproducibility. The 12-ml/min total flowrate gives better precision than the 8 ml/min rate. On the basis of these observations we decided to set C,, = 500 cm, x = 2.0 ml/min and y = 10.0 ml/min and incorporated this flow-dilution system into the total flow system as indicated in Fig. l(c). A long mixing coil Cm2 (900 cm) is necessary to obtain a stable base-line. The flow-rate of the Xylenol Orange solution R3 was varied over the range 2.WlO.O ml/min with a sample-injection volume of 78 ~1 (zinc
ROKURO KURODA and TADASHIM~CHIZUKI
392
(a)
(b)
40
80
1: 5 min
30
i .P ru
60
20
40
IO
20
0
0
x
:: a
Fig. 5. Determination of copper and zinc in copper-base alloys. (a) Runs for copper. Sampling rate 280/hr with the flow diagram shown in Fig. l(a). From left to right is shown a series of aqueous copper standards (0.8. 1.6 and 2.4 mg/ml), followed by six copper-base alloys and additional calibration runs. Each standard or sample is analysed in triplicate. (b) Runs for zinc. Sampling rate 90/hr with flow diagram shown in Fig. I(c). From left to right is shown a set of zinc standards (26, 45 and 64 ng/ml) and four copper-base alloys, followed by another calibration run. Each is analysed in triplicate.
concentration 20 or 40 pg/ml), and the optimum was found to be 6 ml/min; lower rates made the base-line unstable and higher rates broadened the peaks. The sample-injection volume also has a significant effect on the peak height, which increases sharply with increasing injection volume, as illustrated in Fig. 4. Larger injection volumes increase the sensitivity and precision, but limit the speed of analysis owing to broadening of the peaks. We chose an injection volume of 148 ~1, which gives no carry-over. This allows analyses of 90 sample solutions for zinc per hour. Determination of copper, nickel and zinc reference samples of copper-base alloys
in
standard
Routine runs for the determination of copper and zinc in the copper-base alloys tested are illustrated in Fig. 5, together with those for a series of aqueous
and zinc standards. The results are given in Table 1. The copper content of the alloys tested ranges from 60 to almost lOOS, (deoxidized copper) and can be determined with relative standard deviation (rsd) of 0.2-0.79,. Zinc is determined over the range 3&409, with an rsd of 0.550.80& Agreement of the results with the certified values for copper and zinc is satisfactory. Only German silver was analysed for nickel, and again the result and certified value agree satisfactorily. From the practical point of view, the rate-determining step in the alloy analysis is not the flow-injection system, but the other steps including sample weighing, dissolution and dilution to standard definite volume. The precision of the flow-injection analyses described here is rather good, owing to the simple flowsystems developed, but the specifications for copperbase alloys often have comparatively narrow tolercopper
Table 1. Determination of copper, nickel and zinc in standard samples of copper-base alloys by flow-injection analysis
cu.%
Ni 3
Samplet
Found*
Certified
Brass”. class 1 Bras?. class 3 Naval brass’, class 1 Leaded bras@, class 2 Phosphorus-deoxidized copper’
70.5,(0.5) 60.3,(0.3) 62.6,,(0.7) 59.6,(0.2) 99.8s (0.6)
70.61 60.62 62.16 59.63 99.94
Beryllium copper’ German silverg
97.6,(0.5)
97.71
Found*
18.3r (0.4)
O/ 10
Zn. % Certified
18.21
Found*
Certified
29.3s (0.8) 39.3,,(0.6) 36.8,(0.5) 39.2,(0.5) -
29.36 39.35 36.82 39.12
-
*Values in parentheses are the relative standard deviations. t Provided by Japan Brass Makers’ Association. Other constituents (“,) a Pb 0.009. Fe O.Ols. b Pb 0.01,. Fe 0.01,. ’ Pb 0.00664, Sn 1.002. Fe O.OlOt.d Pb 1.211, Sn 0.009, Fe 0.02.,. ’ P 0.0192. ‘Si 0.0555, Fe 0.0343. Co 0.26s. Be 1.894. e Pb 0.0029. Mn 0.26s. Fe 0.014s. Co 0.0048, Cu 55.1 I,, Zn 26.39,.
Continuous spectrophotometric determination of copper, nickel and zinc ante limits, e.g. 34% for copper in brasses and 4--6x for copper in German silvers. The precision should be high enough to ensure maximum economy in the production of alloys containing high-cost constituents such as copper and especially nickel. If a precision of 0.1% is desired, comparable to that of standard volumetric methods, replicate analysis and averaging are recommended. Ten consecutive runs on the same sample will result in an improvement in the precision by a factor of 3, i.e., to 0.074235; for copper and 0.17-0.27”/:, for zinc. The high sampling rate of the flow-injection system compensates for the need for replication, and the overall analysis rate is about 28 samples per hour for copper or nickel and 9 per hour
393
for zinc, which is in line with most industrial laboratory requirements.
REFERENCES
,
J. RtiiiEka and E. H. Hansen, Ann/. Chim. Acta, 1978, ’ 99,37. 2, D. Retteridge, Anal. Chem., 1978,50, 832A. 3. J. RdiiZka and E. H. Hansen, NBS Spa. Pub/., No. 519, 1979,p. 501. z: Idem, Anal. Chim. Acta, 1980, 114, 19. R. Ku&a, T. Mochizuki and K. Oguma, Eunseki Kagaku, 1980,29, T73. 6. M. Namiki and J. Kimura, ibid., 1978, 27, T39. 7. K. Studlar and I. Janousek, Takanta, 1961.8, 203.
0039-914c!8, @50389-0%0?.00~0 CopyrIght 0 1981 Pergamon Press Ltd
CONTINUOUS SPECTROPHOTOMETRIC DETERMINATION OF COPPER, NICKEL AND ZINC IN COPPER-BASE ALLOYS BY FLQ_W-INJECTION ANALYSIS R~KU&~URODA* and TADASHI
MOCHIZUKI
Laboratory for Analytical Chemistry, Faculty of Engineering, University of Chiba, Yayoi-cho, Chiba, Japan (Receired
l?LNorember
1980. Accepted
24 December
1980)
Summary-Flow-injection methods have been developed for the determination of copper, nickel and zinc in copper-base alloys, including several types of brasses, deoxidized copper, beryllium copper and German silver. The system for copper and nickel involves the measurement of the absorbance of the copper(H) and nickel(H) aquo-complexes at 805 and 410 nm, respectively, after simple dissolution of the sample in a nitric acid-phosphoric acid mixture. The system needs no further reagents and the sample solutions can be analysed at rates of up to 280 /hr for copper (or nickel in German silver) without any carry-over. The system for zinc consists of automatic dilution of the injected sample with a thiosulphateacetate buffer solution and the subsequent measurement of the absorbance of the zinc-Xylenol Orange complex at 568 nm. This system permits analysis rates of up to 9O/hr for zinc solutions. with no carry-over. The procedures have been applied to standard copper-base alloys. The results agreed satisfactorily with the certified values. The precision ranges are 0.24.7”., for copper and nickel and 0.5-0.8”; for zinc.
Since its introduction by RtiiEka Stewart in 1975, flow-injection
and Hansen and by analysis has found
wide application, especially in agricultural, environmental, clinical and pharmaceutical analyses. Recent progress has been reviewed in several articles.‘4 Despite its great potential for automatic analysis of alloys, flow-injection analysis has been little applied in this area.5 We have attempted to develop rapid procedures for the determination of copper, nickel and zinc in a variety of copper-base alloys by means of spectrophotometric flow-injection analysis. A simple flow system allows the rapid determination of copper (O-99.9%) and nickel (in German silver only) by direct measurement of the absorb#roe of their aquo-complexes. Spectrophotometric detection with Xylenol Orange (X0) permits the rapid determination of zinc, with minimum consumption’ of reagents because of the small sample size.
metal in 20 ml of 7M nitric acid and diluting exactly to 100 ml with distilled water [Zn(II) 2 mg/ml]. Masking agent-bufir solution. Ten g of sodium thiosulphate dissolved in IM acetic acid-IM sodium acetate buffer solution (pH 5.9) and diluted to 1 litre with the buffer. Xylenol Orange solution. A 0.012% aqueous solution. Apparatus
EXPERIMENTAL
Sample dissolution
Reagents Standard solutions. Standard copper solution was prepared by dissolving 1 g of pure copper metal in 20 ml of 7M nitric acid and diluting exactly to 100 ml with distilled water [Cu(II) 10 mg/ml]. Standard nickel solution was prepared by dissolving 12 g of nickel nitrate in 1.4M nitric acid and diluting to 100 ml with the same acid [Ni(II) 25 mg/ml] and standardized compleximetrically. Standard zinc solution was prepared by dissolving 0.2 g of pure zinc
*To whom correspondence should be sent.
The systems for the determination of copper, nickel and zinc are shown in Fig. 1. The systems were assembled from l.O-mm Teflon tubing except for a back-pressure coil (0.5 mm), valve and connectors. An SJ-121 IH peristaltic pump (Atto. Japan) was operated with Tygon pump tubing (l/8 or l/l6 in. bore) to obtain the desired flow-rates. A sampleinjection valve (Nihon Seimitsu NV-508-6M) made of Teflon with a Teflon sample loop (bore 1.0 mm) was used for sampling. A Shimazu Model W-180 spectrophotometer equipped with a Nihon Seimitsu NS-FC-210 flowthrough cell (volume 31.4 ~1, light-path 10 mm) was used. coupled with a Shimazu Chromatopac C-RIA data-processer to record the peaks and simultaneously calculate the peak heights or areas.
and requests for reprints 389
Determination of copper and zinc. A sample weighing about 60 mg is dissolved in a mixture of 2 ml of O.OlM phosphoric acid and 10 ml of 7M nitric acid and diluted exactly to 50 ml with distilled water. A further tenfold dilution is made for the determination of zinc. Determination ofnickel. A 1.0-g of German silver is taken and treated as for copper determination, Flow-injection
analysis
Carrier solution C (1.4M nitric acid) is pumped into the analytical line at a flow-rate of 8.0 ml/min with a peristaltic pump [Fig. l(a)]. The sample solution (318 ~1 in our apparatus) is introduced into the carrier stream by a six-way loop-valve injector V and the Copper
and nickel.
ROKURO KURODA and TADASHI MOCHIZUKI
390
(b)
P, ml / min
P, ml/min
Fig. 1. Flow diagrams. (a) System for determination of copper and nickel. C: carrier (I.4M HNO,); V: sample-injection valve; S: sample; P: peristaltic pump; D: spectrophotometer; Cs: back-pressure coil; W: waste; Ws: sample waste. (b) Automatic sample-dilution system. C,: carrier (distilled water); R,: reagent (distilled water). (c) System for the determination of zinc. Cr: carrier (0.14M HN03); RI: masking agent-buffer solution (1% Na,S,O,-acetate buffer, pH 5.9); R3: calorimetric reagent (0.012% Xylenol Orange solution); A, B: confluence points; C,,, Cm2 : mixing coils. The numerals under the tubes or coils refer to their lengths in cm and their internal diameters in mm.
absorbance of the aquo-complex is monitored in the flowthrough cell D at 805 nm for copper(H) and 410 nm for nickel(II), respectively, against the carrier solution as reference. A narrow coil Cs (50 cm long) is attached to the exit of the flow-through cell, as a back-pressure coil. Zinc. Carrier solution C2 (0.14M nitric acid) is pumped into the analytical line at a flow-rate of 2.0 ml/min and the sample solution S (148 ~1) is introduced into the carrier solution by the injection valve V [Fig. l(c)]. The sample slug meets the masking agent-buffer solution R, at the point A. In the mixing coil C,, the sample is diluted, copper(I1) and lead(I1) being masked at the same time. The mixture and the calorimetric reagent flow R, (0.012”/;, Xyleno1 Orange solution) merge at the point B. The colour reaction proceeds in the mixing coil Cm2, and the absorbance of the chelate is measured at 568 nm against the waste solution containing no sample constituents. For calibration, similarly treated standard solutions are injected into the line between the sample runs. The area or the height of the absorbance peak can be used for the calibration.
50
40
IyA
z t
I
OO RESULTS AND DISCUSSION For
rapid
continuous
spectrophotometric
analysis
of the alloys for the major components the preparation of the sample solution should be as simple as possible, and the same solution should be suitable for both the copper and the zinc determination. For this purpose a mixture of nitric and phosphoric acids is satisfactory, yielding no precipitation of hydrated
I 0
5 Flow - rota
lo
,
ml / min
2Oo Injection volume
, pl.
Fig. 2. Etfects of flow-rate and injection volume on peak height-aquo-complex method. Sample concentration: A: Cu(II), 3 mg/ml; B: Cu(II), 2 mg/ml. Sample injection volume: 318 ~1 for A and B. C: Cu(II), 2 mg/ml. Flow-rate of carrier keot constant at 8.0 ml/min.
Continuous spectrophotometric
determination
stannic oxide and providing effective masking of iron with phosphate. Copper(H) aquo-complexes exhibit an absorption maximum at 8OS810 nm, with molar absorptivity 12.9 l.mole- ’ .cm- ‘. Such a low sensitivity is well suited to the determination of copper in copper-base alloy@. The absorbance is not affected by variation in the nitric acid concentration. The nickel(H) aquo-complexes also absorb at 805 nm to some extent, causing a 7% positive error when present
at the same w/v concentration as the copper(H). The nickel(I1) has an absorption maximum at 393 nm, but the nitrous acid produced in the dissolution process interferes at this wavelength so the absorbance is measured at 410 nm, at which the molar absorptivity is 3.8 l.mole- 1.cm-I, providing a basis for the determination of nickel in German silver. Xylenol Orange reacts with zinc to form a coloured complex,’ which exhibits an absorption maximum at 568 nm. The chelate is best formed at pH 5.7-5.8. Copper(I1) and lead(U) interfere, but they are effectively masked with thiosulphate. However, nickel(H) still interferes seriously, and the method cannot be applied to nickel alloys unless a suitable masking agent can be found. Optimization
offlow
0.2%, a45 0.63 ‘9 \ \
0.49
:a, 0.04
a45
\
.--a 0.43
&i_
0.09 -. ow----,g --._ --._ 6, a04 --oy6
0.04 011
aoi-----o 0.13
01
0
eo-
z
- 60E .o e x e a 40-
20-
01 0
loo
xx)
330
lr@ction volume, pl. Fig. 4. Effect of injection volume-zinc-Xylenol Orange method. Sample concentration: A: Zn(II) 40 pg/ml; B: Zn(II) 20 pg/ml.
system
The effects of flow-rate and injection volume on the peak height are shown in Fig. 2 for the determination of copper. Flow-rates greater than 5.0 ml/min give a constant absorbance peak-height for an injection
-
391
of copper, nickel and zinc
1
I
lo s Ratio of flow-rotor, u/x
1
IS
Fig. 3. EtTect of flow-rate of carrier and reagent solutionszinc-Xylenol Orange method. x, y: respective flow-rates of carrier and reagent solutions; mixing coil length, z: 100 cm (U), 300 cm (A), 500 cm (0); total Row-rate: 12.0 ml/min (-), 8.0 ml/min (---). The numerals at each point indicate the standard deviation.
volume of 318 ~1. The injection volume has a significant effect, yielding increased peak height and reproducibility with increasing injection volume. However, increasing the injection volume widens the sample zones and lowers the sampling rate, so that 318 ~1 is taken as a compromise. This allows analysis of 280 copper solutions per hour, without any carry-over. The optimum conditions for nickel are the same as those for copper. The Xylenol Orange method is too sensitive to be applied directly to the analysis for zinc. To simplify the sample dilution we decided to dilute the sample within the flow system with a combined masking agent-buffer solution, as illustrated in Fig. l(b). TO examine the effectiveness of this dilution system, the 0.01% Xylenol Orange solution (78 ~1) was introduced by the sample-injection valve V into the analytical line, where the total flow rate (x + y) was kept constant at either 8.0 or 12.0 ml/min and the ratio of flow-rates y/x 0, is the flow-rate of the diluent) and the length of coil C,, were varied over the ranges $15 and 100-500 cm, respectively. The results are illustrated in Fig. 3. Increasing either the coil length or the dilution ratio lowered the peak height but improved the reproducibility. The 12-ml/min total flowrate gives better precision than the 8 ml/min rate. On the basis of these observations we decided to set C,, = 500 cm, x = 2.0 ml/min and y = 10.0 ml/min and incorporated this flow-dilution system into the total flow system as indicated in Fig. l(c). A long mixing coil Cm2 (900 cm) is necessary to obtain a stable base-line. The flow-rate of the Xylenol Orange solution R3 was varied over the range 2.WlO.O ml/min with a sample-injection volume of 78 ~1 (zinc
ROKURO KURODA and TADASHIM~CHIZUKI
392
(a)
(b)
40
80
1: 5 min
30
i .P ru
60
20
40
IO
20
0
0
x
:: a
Fig. 5. Determination of copper and zinc in copper-base alloys. (a) Runs for copper. Sampling rate 280/hr with the flow diagram shown in Fig. l(a). From left to right is shown a series of aqueous copper standards (0.8. 1.6 and 2.4 mg/ml), followed by six copper-base alloys and additional calibration runs. Each standard or sample is analysed in triplicate. (b) Runs for zinc. Sampling rate 90/hr with flow diagram shown in Fig. I(c). From left to right is shown a set of zinc standards (26, 45 and 64 ng/ml) and four copper-base alloys, followed by another calibration run. Each is analysed in triplicate.
concentration 20 or 40 pg/ml), and the optimum was found to be 6 ml/min; lower rates made the base-line unstable and higher rates broadened the peaks. The sample-injection volume also has a significant effect on the peak height, which increases sharply with increasing injection volume, as illustrated in Fig. 4. Larger injection volumes increase the sensitivity and precision, but limit the speed of analysis owing to broadening of the peaks. We chose an injection volume of 148 ~1, which gives no carry-over. This allows analyses of 90 sample solutions for zinc per hour. Determination of copper, nickel and zinc reference samples of copper-base alloys
in
standard
Routine runs for the determination of copper and zinc in the copper-base alloys tested are illustrated in Fig. 5, together with those for a series of aqueous
and zinc standards. The results are given in Table 1. The copper content of the alloys tested ranges from 60 to almost lOOS, (deoxidized copper) and can be determined with relative standard deviation (rsd) of 0.2-0.79,. Zinc is determined over the range 3&409, with an rsd of 0.550.80& Agreement of the results with the certified values for copper and zinc is satisfactory. Only German silver was analysed for nickel, and again the result and certified value agree satisfactorily. From the practical point of view, the rate-determining step in the alloy analysis is not the flow-injection system, but the other steps including sample weighing, dissolution and dilution to standard definite volume. The precision of the flow-injection analyses described here is rather good, owing to the simple flowsystems developed, but the specifications for copperbase alloys often have comparatively narrow tolercopper
Table 1. Determination of copper, nickel and zinc in standard samples of copper-base alloys by flow-injection analysis
cu.%
Ni 3
Samplet
Found*
Certified
Brass”. class 1 Bras?. class 3 Naval brass’, class 1 Leaded bras@, class 2 Phosphorus-deoxidized copper’
70.5,(0.5) 60.3,(0.3) 62.6,,(0.7) 59.6,(0.2) 99.8s (0.6)
70.61 60.62 62.16 59.63 99.94
Beryllium copper’ German silverg
97.6,(0.5)
97.71
Found*
18.3r (0.4)
O/ 10
Zn. % Certified
18.21
Found*
Certified
29.3s (0.8) 39.3,,(0.6) 36.8,(0.5) 39.2,(0.5) -
29.36 39.35 36.82 39.12
-
*Values in parentheses are the relative standard deviations. t Provided by Japan Brass Makers’ Association. Other constituents (“,) a Pb 0.009. Fe O.Ols. b Pb 0.01,. Fe 0.01,. ’ Pb 0.00664, Sn 1.002. Fe O.OlOt.d Pb 1.211, Sn 0.009, Fe 0.02.,. ’ P 0.0192. ‘Si 0.0555, Fe 0.0343. Co 0.26s. Be 1.894. e Pb 0.0029. Mn 0.26s. Fe 0.014s. Co 0.0048, Cu 55.1 I,, Zn 26.39,.
Continuous spectrophotometric determination of copper, nickel and zinc ante limits, e.g. 34% for copper in brasses and 4--6x for copper in German silvers. The precision should be high enough to ensure maximum economy in the production of alloys containing high-cost constituents such as copper and especially nickel. If a precision of 0.1% is desired, comparable to that of standard volumetric methods, replicate analysis and averaging are recommended. Ten consecutive runs on the same sample will result in an improvement in the precision by a factor of 3, i.e., to 0.074235; for copper and 0.17-0.27”/:, for zinc. The high sampling rate of the flow-injection system compensates for the need for replication, and the overall analysis rate is about 28 samples per hour for copper or nickel and 9 per hour
393
for zinc, which is in line with most industrial laboratory requirements.
REFERENCES
,
J. RtiiiEka and E. H. Hansen, Ann/. Chim. Acta, 1978, ’ 99,37. 2, D. Retteridge, Anal. Chem., 1978,50, 832A. 3. J. RdiiZka and E. H. Hansen, NBS Spa. Pub/., No. 519, 1979,p. 501. z: Idem, Anal. Chim. Acta, 1980, 114, 19. R. Ku&a, T. Mochizuki and K. Oguma, Eunseki Kagaku, 1980,29, T73. 6. M. Namiki and J. Kimura, ibid., 1978, 27, T39. 7. K. Studlar and I. Janousek, Takanta, 1961.8, 203.