Analyriccr
C’ki~?~lcrrAcra.
,(T Elscvicr
Scicntilic
77 ( 1975
Publishing
65
J 65-70
Company.
Amstcrditm
DETERMINATION OF SILVER FREQUENCY PLASMA-TORCH RYOZO
NAKASHIMA.
Go~*ewtmw Imiltstritrl (Rcccivcd
13th January
SHOZO
in The Ncthcrlanda
IN BIOLOGICAL MATERIALS EMISSION SPECTROMETRY
SASAKI und SHOZO
Resecnd~ Irrsrirrtre.
- Printed
BY HIGH-
SHIBATA
Nt~go~v~. Nirtrre-mrclri
Kitty-h.
NN~JO.W (Jtrpm)
1975)
High-frequency plasma torches have recently become more widely studied for excitation in cm&ion spcctrometry. They provide better reproducibility than conventional methods of excitation. and much higher temperatures than flames, and they can be applied to gaseous, liquid and solid samples. In previous work from this Laboratory’ . a Hitachi UHF plasma Spectrascan was used for the determination of niobium, titanium and zirconium in steels. The method has the advantages outlined above. but when aqueous solutions are used. the moisture which is not trapped in the condenser at the front of the atomizer. can lower the efficiency of the plasma and thus the emission intensity. Silver can be detected with very high sensitivity when a plasma torch is used. but even better sensitivity may be necessary when the amount of sample available is limited. A considerable improvement in sensitivity can be achieved if the silver spectral line is enhanced by the presence of bismuth. and if a second at the front of the atomization chamber condenser is maintained at about -3-5-C Silver can be collected from biological samples by codescribed previously’. precipitation with bismuth iodide, before plasma-torch emission spectrometry. The combination of techniques allows the detection limit to be improved to 0.5 ng of silver for 0.2 ml of solutions of samples such as whole blood. EXPERIMENTAL
Apparatus
A Hitachi UHF-Plasma Spectruscan Model 300, which has already been described ‘, was used. In preliminary experiments, spectral intensities were measured by wavelength scanning during continuous introduction of the sample solution. The instrumental conditions were as follows: wavelength, 328.07 nm: plasma gas. sheath gas, 3.5 I Ar min- ’ ; cooling water. 5°C: heating chamber. 3.0 1 Ar min-I; l80& 10°C; entrance slit, 30 jtm: exit slit. 50 Alrn: pneumatic nebulizer, 3.5 ml min-’ with water at 3.0 I mine’* , anode current. 300 mA: lield current. 400 mA: photomultiplier. Hamamatsu TV. RlO6. at 600 V. For actual analyses. the wavelength of the scanning monochromator was fixed at 328.06 nm and a delinite aliquot of sample solution was introduced. When the second condenser was used, the field current was set at 270 mA, which was the maximum value possible for a stable plasma. The nebulizer was provided with a two-way tap. so that either rinsing
R. NAKASHIMA.
66
S. SASAKI.
S. SHIBATA
water or sample solution from a 2-ml vessel could be introduced. An aliquot of solution was pipetted into the vessel. and the sample was introduced by quickly turning the tap. At the end of the sa’rnplc introduction, the tap was turned as soon as possible, to minimize introduction of air. so that distilled water was introduced through the other arm of the tap. The sampling vessel was then rinsed with distilled water for the next sample. Starularti
0.1 M nitric twice-distilled
silver
solution
acid (Wako water.
(1000
Pure
pprn.).
Chemical
A standard silver nitrate solution in Industries) was suitably diluted with
Stmtlnrcl hisrmrttt solutior7 (5000 p.pm). Bismuth metal ( > 99.999%; Mitsubishi Metal Mining Co.) was dissolved in nitric acid and the solution was suitably diluted. Other standard solutions (2000 p,p.m.) were prepared from iron, mnnganesc or thallium wire (Johnson-Matthey). antimony and tin (a 99.999”/0,: Mitsubishi Metal Mining) and yttrium oxide (Shinetsu Chemicals). These metals and oxide were dissolved in as little hydrochloric or nitric acid (super-special-grade: Wako Chemicals) as possible, and the solutions were diluted suitably with twice-distilled water in volumetric flasks. All other reagents used were of high purity or superior to analytical-grade chemicals. Twice-distilled water was used throughout,
Pipette a l-2-ml aliquot of sample, 0.5 ml of perchloric acid (70’%) and I.0 ml of concentrated nitric acid into a IO-ml beaker. and cover with a watchglass. Heat gently on a hot plate: when decomposition is complete. continue heating until the perchloric acid has almost evaporated, but do not evaporate to dryness. Dissolve the residue in 2-3 ml of water, and transfer the solution quantitatively to a 15-ml centrifuge tube. Add 1.0 ml of 0.4 M sodium iodide and 1.0 ml of bismuth solution while stirring. After IO min. centrifuge, decant the solution and rinse the precipitate several times with 2 ml of water. Pour 0.1-0.2 ml of concentrated nitric acid on the collected precipitate, and heat until violet fumes cease to appear. Transfer the solution to a 2-ml measuring cylinder (graduated at 0.05-ml intervals) and rinse the tube into the cylinder at least five times with the minimal volume of water. Then. read the volume exactly. and stopper and shake the cylinder. Analyze this solution by the standard addition method. RESULTS
AND DISCUSSION
The resonance line of silver at 328.07 nm was compared with the 338.28nin line for sensitivity; the former line had the better detection limit and was therefore used. The profile of the emission intensity over the width of the plasma (Fig. I) showed that intensity was maximal at 3.54.0 mm from the plasma axis at a fixed height of 15-30 mm from the tip of the electrode. The effect of inorganic acids on the emission intensity is shown in Fig, 2.
SILVER
BY
PLASMA-TORCH
SPECTROMETRY
67
Q20-
0 Distance
Fig.
1. The
Fig. (Q)
2. Effects HISOL.
As the smallest
profile
( mm
acidity effect
1
of the emission
of acid
I 050
I 025
0
conccntmtions
I 074
I 100
M intensity on
at 328.07
the silver
nm over the plusma lint
intensity:
(0)
width
HNOJ,
increased. the intensity gradually decreased. and was therefore used in subsequent work.
for 0. (0)
Nitric
I p.p.m. silver.
HCI.
acid
(0)
HCIO.,.
had
the
Whole blood usually contains about 3000 pp.m, of sodium and 300 p.p,m. The effects of these metals on the emission intensity for silver must bc known. as must their effects when traces of silver are precipitated with meta13-S or an organic reagcnt6*’ and the precipitate is dissolved before with the plasma torch, Although it was not explicitly stated earlier’*“. material tends to decrease intensities, probably because of scattering of by carbonized particles in the outer region of the plasma torch: organic were therefore not examined further. The cfl’ccts of diverse metals (1000 p.p.m.) on the emission intensity of 0.1 p.p,m. of silver were studied. Antimony. bismuth. iron, lead. tin and yttrium all enhanced the emission intensity. but the intensity profile was unchanged. With sodium. as the concentration increased, the intensity was enhanced gradually. and the position of maximum intensity shifted towards the plasma axis. coinciding with it at 200 p.p.m. of sodium: at higher concentrations. the intensity decreased gradually. but the intensity profile was unchanged. At sodium concentrations of 1000-3000 p.p,m., the intensity was less than one tenth that in the absence of sodium. Similar behavior was found with thallium. Manganese showed such a strong line at 328.08 nm that the silver lint could not be measured. Iron showed a weak line at 328.03 nm. but the emission intensity for 250 p.p.m. of iron was only a few percent of that for 0.1 p.p.m. of silver. In the analytical method. the silver line could be recorded without interference from iron. except for increased background noise. These results indicate that iron, manganese and sodium must be eliminated from the final solution. Other elements with ncighbouring lines. i.e. molybdenum. titanium, vanadium and zirconium. did not interfere at the lo-p.p.m. level.
of iron. therefore a carrier analysis organic radiation reagents
R. NAKASHIMA.
68
OL
’
1
I’l*ll’
I
Silcer rcrket1 (llfl) _--_
10.0
*
I
,,,,I
I
S. SHIBATA
It111111 10.
(PPm)
of bismuth on the emission intensity of 0. I p.p.m. silver. (0)
In lhc ubscncc
1
PRECISION
2.0 4.0
I
10'
Bi Fig. 3. EfTcct ofconccntration of bismuth. TABLE
,
102
10
S. SASAKI.
AND ACCURACY
FOR SILVER
A wrcrye /iJld (II
C’ki~?~lcrrAcra.
,(T Elscvicr
Scicntilic
77 ( 1975
Publishing
65
J 65-70
Company.
Amstcrditm
DETERMINATION OF SILVER FREQUENCY PLASMA-TORCH RYOZO
NAKASHIMA.
Go~*ewtmw Imiltstritrl (Rcccivcd
13th January
SHOZO
in The Ncthcrlanda
IN BIOLOGICAL MATERIALS EMISSION SPECTROMETRY
SASAKI und SHOZO
Resecnd~ Irrsrirrtre.
- Printed
BY HIGH-
SHIBATA
Nt~go~v~. Nirtrre-mrclri
Kitty-h.
NN~JO.W (Jtrpm)
1975)
High-frequency plasma torches have recently become more widely studied for excitation in cm&ion spcctrometry. They provide better reproducibility than conventional methods of excitation. and much higher temperatures than flames, and they can be applied to gaseous, liquid and solid samples. In previous work from this Laboratory’ . a Hitachi UHF plasma Spectrascan was used for the determination of niobium, titanium and zirconium in steels. The method has the advantages outlined above. but when aqueous solutions are used. the moisture which is not trapped in the condenser at the front of the atomizer. can lower the efficiency of the plasma and thus the emission intensity. Silver can be detected with very high sensitivity when a plasma torch is used. but even better sensitivity may be necessary when the amount of sample available is limited. A considerable improvement in sensitivity can be achieved if the silver spectral line is enhanced by the presence of bismuth. and if a second at the front of the atomization chamber condenser is maintained at about -3-5-C Silver can be collected from biological samples by codescribed previously’. precipitation with bismuth iodide, before plasma-torch emission spectrometry. The combination of techniques allows the detection limit to be improved to 0.5 ng of silver for 0.2 ml of solutions of samples such as whole blood. EXPERIMENTAL
Apparatus
A Hitachi UHF-Plasma Spectruscan Model 300, which has already been described ‘, was used. In preliminary experiments, spectral intensities were measured by wavelength scanning during continuous introduction of the sample solution. The instrumental conditions were as follows: wavelength, 328.07 nm: plasma gas. sheath gas, 3.5 I Ar min- ’ ; cooling water. 5°C: heating chamber. 3.0 1 Ar min-I; l80& 10°C; entrance slit, 30 jtm: exit slit. 50 Alrn: pneumatic nebulizer, 3.5 ml min-’ with water at 3.0 I mine’* , anode current. 300 mA: lield current. 400 mA: photomultiplier. Hamamatsu TV. RlO6. at 600 V. For actual analyses. the wavelength of the scanning monochromator was fixed at 328.06 nm and a delinite aliquot of sample solution was introduced. When the second condenser was used, the field current was set at 270 mA, which was the maximum value possible for a stable plasma. The nebulizer was provided with a two-way tap. so that either rinsing
R. NAKASHIMA.
66
S. SASAKI.
S. SHIBATA
water or sample solution from a 2-ml vessel could be introduced. An aliquot of solution was pipetted into the vessel. and the sample was introduced by quickly turning the tap. At the end of the sa’rnplc introduction, the tap was turned as soon as possible, to minimize introduction of air. so that distilled water was introduced through the other arm of the tap. The sampling vessel was then rinsed with distilled water for the next sample. Starularti
0.1 M nitric twice-distilled
silver
solution
acid (Wako water.
(1000
Pure
pprn.).
Chemical
A standard silver nitrate solution in Industries) was suitably diluted with
Stmtlnrcl hisrmrttt solutior7 (5000 p.pm). Bismuth metal ( > 99.999%; Mitsubishi Metal Mining Co.) was dissolved in nitric acid and the solution was suitably diluted. Other standard solutions (2000 p,p.m.) were prepared from iron, mnnganesc or thallium wire (Johnson-Matthey). antimony and tin (a 99.999”/0,: Mitsubishi Metal Mining) and yttrium oxide (Shinetsu Chemicals). These metals and oxide were dissolved in as little hydrochloric or nitric acid (super-special-grade: Wako Chemicals) as possible, and the solutions were diluted suitably with twice-distilled water in volumetric flasks. All other reagents used were of high purity or superior to analytical-grade chemicals. Twice-distilled water was used throughout,
Pipette a l-2-ml aliquot of sample, 0.5 ml of perchloric acid (70’%) and I.0 ml of concentrated nitric acid into a IO-ml beaker. and cover with a watchglass. Heat gently on a hot plate: when decomposition is complete. continue heating until the perchloric acid has almost evaporated, but do not evaporate to dryness. Dissolve the residue in 2-3 ml of water, and transfer the solution quantitatively to a 15-ml centrifuge tube. Add 1.0 ml of 0.4 M sodium iodide and 1.0 ml of bismuth solution while stirring. After IO min. centrifuge, decant the solution and rinse the precipitate several times with 2 ml of water. Pour 0.1-0.2 ml of concentrated nitric acid on the collected precipitate, and heat until violet fumes cease to appear. Transfer the solution to a 2-ml measuring cylinder (graduated at 0.05-ml intervals) and rinse the tube into the cylinder at least five times with the minimal volume of water. Then. read the volume exactly. and stopper and shake the cylinder. Analyze this solution by the standard addition method. RESULTS
AND DISCUSSION
The resonance line of silver at 328.07 nm was compared with the 338.28nin line for sensitivity; the former line had the better detection limit and was therefore used. The profile of the emission intensity over the width of the plasma (Fig. I) showed that intensity was maximal at 3.54.0 mm from the plasma axis at a fixed height of 15-30 mm from the tip of the electrode. The effect of inorganic acids on the emission intensity is shown in Fig, 2.
SILVER
BY
PLASMA-TORCH
SPECTROMETRY
67
Q20-
0 Distance
Fig.
1. The
Fig. (Q)
2. Effects HISOL.
As the smallest
profile
( mm
acidity effect
1
of the emission
of acid
I 050
I 025
0
conccntmtions
I 074
I 100
M intensity on
at 328.07
the silver
nm over the plusma lint
intensity:
(0)
width
HNOJ,
increased. the intensity gradually decreased. and was therefore used in subsequent work.
for 0. (0)
Nitric
I p.p.m. silver.
HCI.
acid
(0)
HCIO.,.
had
the
Whole blood usually contains about 3000 pp.m, of sodium and 300 p.p,m. The effects of these metals on the emission intensity for silver must bc known. as must their effects when traces of silver are precipitated with meta13-S or an organic reagcnt6*’ and the precipitate is dissolved before with the plasma torch, Although it was not explicitly stated earlier’*“. material tends to decrease intensities, probably because of scattering of by carbonized particles in the outer region of the plasma torch: organic were therefore not examined further. The cfl’ccts of diverse metals (1000 p.p.m.) on the emission intensity of 0.1 p.p,m. of silver were studied. Antimony. bismuth. iron, lead. tin and yttrium all enhanced the emission intensity. but the intensity profile was unchanged. With sodium. as the concentration increased, the intensity was enhanced gradually. and the position of maximum intensity shifted towards the plasma axis. coinciding with it at 200 p.p.m. of sodium: at higher concentrations. the intensity decreased gradually. but the intensity profile was unchanged. At sodium concentrations of 1000-3000 p.p,m., the intensity was less than one tenth that in the absence of sodium. Similar behavior was found with thallium. Manganese showed such a strong line at 328.08 nm that the silver lint could not be measured. Iron showed a weak line at 328.03 nm. but the emission intensity for 250 p.p.m. of iron was only a few percent of that for 0.1 p.p.m. of silver. In the analytical method. the silver line could be recorded without interference from iron. except for increased background noise. These results indicate that iron, manganese and sodium must be eliminated from the final solution. Other elements with ncighbouring lines. i.e. molybdenum. titanium, vanadium and zirconium. did not interfere at the lo-p.p.m. level.
of iron. therefore a carrier analysis organic radiation reagents
R. NAKASHIMA.
68
OL
’
1
I’l*ll’
I
Silcer rcrket1 (llfl) _--_
10.0
*
I
,,,,I
I
S. SHIBATA
It111111 10.
(PPm)
of bismuth on the emission intensity of 0. I p.p.m. silver. (0)
In lhc ubscncc
1
PRECISION
2.0 4.0
I
10'
Bi Fig. 3. EfTcct ofconccntration of bismuth. TABLE
,
102
10
S. SASAKI.
AND ACCURACY
FOR SILVER
A wrcrye /iJld (II
