Talonra Vol. 28, pp. 177 t0 181 0 Pergamon Press Ltd 1981. Printed in Great Britain

0039.9140/81/030177-05$02.00/O

ELECTROTHERMAL ATOMIZATION OF CALCIUM AND STRONTIUM IN A MOLYBDENUM

MICRO-TUBE

MASAM~ SUZUKI* and KIYOHISA OHTA

Department of Chemistry, Faculty of Engineering, Mie University, Kamihama-cho, Tsu, Mie-ken 514, Japan

(Received 24 April 1980. Reoised 28 May 1980. Accepted 29 September 1980)

Summary-The excitation and ionization of calcium and strontium in a molybdenum micro-tube atomizer and their use in atomic-absorption spectrometry are described. Increasing hydrogen flow led to complex calcium atomization profiles for absorption measurements, but decreasing hydrogen flow resulted in higher atomic emission. Ionization of calcium and its suppression by potassium were also observed. Strontium was measured effectively by atomic absorption because of the higher sensitivity. Increasing hydrogen flow resulted in a lower atomization temperature and higher absorption for stron- tium, while decreasing hydrogen flow resulted in higher atomic emission. No interference from lOO-fold amounts of magnesium, calcium and sodium was found for atomization of strontium, but lOO-fold amounts of aluminium shifted the peak temperature for strontium though with no variation of appear- ance temperature and peak absorption. A small shift in peak temperature was observed for strontium in the presence of lOO-fold amounts of phosphate.

Although the use of electrothermal atomization in atomic-absorption spectrometry is now widespread little work has been published on electrothermal atomization of calcium and strontium. Cragin et al.’ have determined calcium at below the 5 pg/l. level in glacial snow samples with a graphite atomizer and Berggren et al.’ have used a carbon-rod atomizer for the determination of calcium in biological materials. A graphite atomizer has also been used to measure strontium in blood serum3 and soi1.4 Approximately, 20 pg of calcium and 5 pg of strontium can be deter- mined in the graphite tube furnace. The excitation potentials of the resonance lines for alkaline earth elements are relatively low. This fact allows sensitive measurements of these elements to be made by flame emission spectrometry. The ionization potentials are also relatively low.

Recently, atomic emission obtained with an electri- cally heated carbon furnace has been used for deter- mination of some elements and the potential of this technique as an analytical tool has been recognized.’ Significant improvements in detection limits are obtained for alkali metals, compared to those obtained by carbon-furnace atomioabsorption spec- trometry. Ionization has been shown to occur when carbon furnaces are used.6

However, the formation of stable carbides in car- bon atomizers results in the release of alkaline-earth elements in the atomic form being slow. Attempts have been made to prevent carbide formation by lining the graphite tube with tantalum foil.’ Metal

*To whom correspondence and requests for reprints should be addressed.

atomizers reduce the possibility of formation of car- bides and metal atomizers appear to be attractive for atomization of alkaline-earth metals. The metal microtube atomizer needs only a short heating period to provide an environment with a uniform tempera- ture throughout the atomizer. The maximum atom cloud concentration is reached in an extremely short period of time, in contrast to conventional carbon atomizers.

This paper describes measurements of the exci- tation and ionization of calcium and strontium taking place in molybdenum micro-tube atomizers’ and the resultant effects on atomic-absorption measurements. Atomic emission from a molybdenum micro-tube atomizer is also discussed as a possible analytical technique.

Apparatus

EXPERIMENTAL

All absorption and emission measurements were made with a Nippon Jarrell-Ash 0.5-m Ebert-type monochroma- tor coupled to an R106 photomultiplier tube (Hamamatsu TV Co.) and a fast-response amplifier.’ The output signal from the amplifier was monitored on a Memoriscope (Iwatsu MS-5021) with a time constant of 1 psec.

A molybdenum micro-tube (15 mm long and 2 mm bore) was used as the atomizer. It was made from molybdenum sheet (0.05 mm thick). The micro-tube was mounted on two copper supports so that there was no localized vari- ation in tube temperature. The micro-tube atomizer was enclosed in a Pyrex chamber (300 ml volume) which had two silica end-windows to allow transmission of the light- beam. The chamber was purged with argon and hydrogen.

Two light apertures (0.5 mm diameter) were positioned in front of the monochromator entrance slit so that the light-beam from the hollow-cathode lamp passed through the centre of the micro-tube and through the apertures.

111

178 MASAMI SUZUKI and KIYOHISA OHTA

The apertures reduced the micro-tube emission to a negli- gible extent.

The atomizer surface temperature was measured with a photodiode (Hamamatsu TV Co., S641). The signal from the photodiode was calibrated with an optical pyrometer. The signal from the photodiode and the absorption or emission signal were recorded simultaneously by using the two beams of the Memoriscope.

A calcium hollow-cathode lamp (Hamamatsu TV Co.) was used for absorption measurements of atoms at 422.67 nm and a strontium hollow-cathode lamp (Hamamatsu TV Co.) for absorption measurements of atoms at 460.73 nm and ions at 407.77 nm. Atomic emission measurements were also made at 422.67 nm and 460.73 nm for calcium and strontium, respectively. Ionic emission for calcium was measured at 393.37 nm.

All sample solutions were injected with a glass micro- pipette (1 ~1) into the micro-tube through a 0.3-mm hole, drilled at the mid-point of the tube.

Reagents

Stock solutions (1 mg/ml) of calcium and strontium were prepared by dissolving their carbonates in hydrochloric acid and diluting with demineralized water. Dilute sol- utions were prepared immediately before use. Demineral- ized water gave no signal for alkaline-earth elements. All reagents were of analytical reagent grade and checked for calcium or strontium as impurities.

Procedure

Samples of 1 ~1 of the element of interest were injected into the micro-tube and dried at 100” for 30 sec. Samples were then atomized by heating to a final temperature of 2300”. The measurements of atomic emission for strontium and ionic emission for calcium and strontium were achieved by increasing the sensitivity of the vertical deflec- tion system of the Memoriscope. Operation of the photo- multiplier at increased gain was necessary for measure- ments of the ionic absorption of calcium and strontium.

RESULTS AND DISCUSSION

Electrothermal atomization of calcium

The addition of hydrogen to argon as the purge gas in the absorption chamber has a favourable effect on the electrothermal atomization of some elements with metal micro-tube atomizers. lo Hydrogen also serves

13SCP+ P ,p, .:.. a ‘:i “_\.:. ._,._. -. -“. k -;_ fl :j i.‘\_~ _-b

0 I i,’

. . . . . .._ .._

‘C

0 1.0 Time , set

Fig. 1. Memoriscope traces showing effect of flow-rate of hydrogen on atomic absorption of calcium (100 pg): a, Ar 500 ml/min; b, Ar 480 ml/min and H, 20 ml/min; c, Ar 100 ml/min and Hz 400 ml/min; d, HZ 500 ml/min; e, tempera- ture increase. Photomultiplier voltage 770 V; vertical scale

of Memoriscope 1 V/div.

0 1.0 Time . set

Fig. 2. Memoriscope traces showing effect of flow-rate of hydrogen on atomic emission of calcium (100 pg): a, Ar 480 ml/min and H, 20 ml/mitt; b, Ar 400 ml/min and H, 100 ml/min; c, Ar 300 ml/min and H2 200 ml/min; d, Ar 100 ml/min and H2 400 ml/min; e, H, 500 ml/min;1; tem- perature increase; g, background at Ar = 500 ml/min. Photomultiplier voltage 770 V; vertical scale of

Memoriscope 0.5 V/div.

to protect the atomizer from oxidation by traces of oxygen in the argon. Therefore, the dependence of atomic absorption and emission profiles on hydrogen flow-rate was examined for calcium. The results are shown in Figs. 1 and 2. Atomic absorption and emission profiles were traced with different vertical deflection scales on the Memoriscope at the same photomultiplier voltage. Complex absorption profiles were seen on addition of hydrogen to the argon. Increasing hydrogen flow resulted in reduction of the atomic emission. In pure argon no atomic emission was observed, although a higher background still remained. The reason for this observation is not clear. Figure 3 shows the variation of atomic-emission sig- nals with increasing amounts of calcium. This result suggests the possibility of the determination of cal- cium by atomic emission in a molybdenum micro-

tube atomizer. However, improvements in atomizer design will be necessary for greater sensitivity of atomic emission. Accurate measurements of atomic absorption were difficult to make without contribu-

h 205.0’ e

I I,, , I,, I I_ 0 1.0

Time . set Fig. 3. Memoriscope traces for atomic emission of different amounts of calcium: a, 0.1 pg; b, 1 pg; c, 10 pg; d, 50 pg; e, temperature increase. Gas flow-rates, Ar 480 ml/min and

H, 20 ml/min.

Atomization of calcium and strontium 179

I I I I I I. I L 4

0 1.0 Time , set

Fig 4. Memoriscope traces for ionic emission of calcium (100 pg): a, no potassium; b, 500 ng of potassium added; c, temperature increase. Gas flow-rates, Ar 480 ml/min and

HZ 20 ml/min.

tions from atomic emission when the present detec- tion system was used, because an unmodulated light source was employed for fast tracing of the atomiz- ation signals on the Memoriscope.

Calcium has an ionization potential of 6.11 eV and measurements of the emission of ionized species could be made. Figure 4 shows the profiles recorded during the atomization stage for calcium ionic emission. In flame spectrometry ionization is often suppressed by use of elements with a low ionization potential. The suppression of the calcium ionic emission (during the micro-tube atomization) by the addition of potassium (ionization potential 4.339 eV) to the calcium solution is analogous to that used in flame spectrometry. No molecular emission was observed for calcium.

Detection limits were calculated from peak heights in the atomization profiles and defined as that concentration giving an absorption or emission signal:background ratio of 2. In metal micro-tube atomization the detection limit is restricted by the small sample volume (1 ~1). Therefore the detection limits are given in absolute terms (pg). A detection limit of 12 pg was obtained under the optimized con- ditions for the atomic emission of calcium. The detec- tion limit was difficult to calculate from the peak height for atomic absorption because of the complex absorption profiles.

Electrothermal atomization of strontium

Figures 5 and 6 show the Memoriscope traces dur- ing the electrothermal atomization of strontium. These demonstrate the effect of hydrogen flow-rate on the atomic absorption and emission of strontium. The atomic absorption profiles were traced with the emission signals attenuated by reducing the sensitivity for the vertical deflection of the Memoriscope by a factor of about 5. The emission signals were negligible for the strontium concentrations tested. The atomic- absorption profiles are characterized by sharper and narrower curves, showing higher peak absorption with increasing hydrogen flow. The increased reduc- ing nature of the gas phase may be responsible for the

‘I. A

0 1.0 Time , set

Fig. 5. Memoriscope traces showing effect of flow-rate of hydrogen on atomic absorption of strontium (50 pg): a, Ar 500 ml/mitt; b, Ar 480 ml/min and H, 20 ml/min; c, 450 ml/min and Hz 50 ml/min; d, Ar 400 ml/min and Hz 100 ml/min; e, Ar 300 ml/min and HZ 200 ml/min; J Ar 100 ml/min and H2 400 ml/min; g, H2 500 ml/mitt; h, tempera- ture increase. Photomultiplier voltage 730 V; vertical scale

of Memoriscope 0.5 V/div.

improvement in the atomization characteristics. No background was observed in the atomic-absorption measurements.

Higher atomic emission for strontium was observed at flow-rates of hydrogen lower than those used for calcium. The profiles for both atomic absorption and emission in pure argon were more rounded, with the decay from the maximum having lower slope. Maxi- mum emission signals were shown at the same tem- perature at which atomic-absorption signals were a maximum. Ottaway and Shaw5 have shown that emission maxima in graphite-furnace atomic emission spectrometry occur later in time than do the absorp tion maxima, since atomization takes place at a lower temperature than that which is optimum for exci- tation. Alder et al. I1 found a similar phenomenon for aluminium and suggested that this was a result of

2120. 2270'

I I I, I I I , I,

0 1.0 Time , set

Fig 6. Memoriscope traces showing etlect of flow-rate of hydrogen on atomic emission of strontium (50 pg): a, Ar 500 ml/min; b, Ar 480 ml/min and H, 20 ml/min; c, Ar 450 ml/min and Hz 50 ml/min; d, 400 ml/min and H2 100 ml/min; e, Ar 300 ml/min and HZ 200 ml/min;f, Ar 100 ml/min and H1 400 ml/mitt; g. Hz 500 ml/min; h, tempera- ture increase. Photomultiplier voltage 30 V; vertical scale

of Memoriscope 0.1 V/div.

180 MASAMI SUZUKI and KIYOHISA OHTA

50 4% .z

6; C

0

0 1 .o Time , set

Fig. 7. Memoriscope traces for atomic absorption and emission of different amounts of strontium: a and d, 10 pg; b and e, 25 pg; c andf, 50 pg; g, 100 pg. a-c for absorption and d-g for emission. Detection conditions as for Fig. 5 for absorption and Fig. 6 for emission. Gas flow-rates, Ar 450

ml/min and H2 50 ml/min.

atoms being formed in the ground state at a tempera- ture below that at which significant emission occurs. The results for molybdenum micro-tube atomizers differed from those for carbon atomizers. This is pre- sumably due to the fast uniform heating of micro-tube atomizers compared with that of carbon atomizers. No molecular emission was observed for strontium. The atomic absorption and emission signals obtained with increasing amounts of strontium are presented in Fig. I.

Strontium has a lower ionization potential (5.692 eV) and therefore it seems possible that some degree of ionization might occur during the atomization pro- cess. Figure 8 shows the Memoriscope traces for strontium ionic absorption at different hydrogen flow- rates. Measurements were made at an increased photomultiplier voltage so as to observe the ionic absorption profiles with the Memoriscope. Higher

% A 100

50

0

2120’ 2210-

rg

0 1.0 Time , set

Fig. 8. Memoriscope traces showing effect of flow-rate of hydrogen on ionic absorption of strontium (50 pg): a, Ar 500 ml/min; b, Ar 480 ml/min and H2 20 ml/min; c, Ar 450 ml/min and H, 50 ml/min; d, Ar 400 ml/min and Hz 100 ml/min; e, Ar 300 ml/min and H, 200 ml/min; f; Ar 100 ml/min and H2 400 ml/min, and H, 500 ml/min; g. tem- perature increase. Photomultiplier voltage 780 V; vertical

scale of Memoriscope 0.5 V/div.

absorption was shown at lower hydrogen flow-rates, but weak absorption was observed in pure argon. Changes in hydrogen flow-rate have different effects on absorption and emission and this phenomenon requires further investigation. The suppression of ion- ization with potassium was also observed for the ionic absorption and emission of strontium. Lower ionic absorption and emission have little effect on measure- ments of atomic absorption for strontium.

Absolute detection limits of 1.1 and 5 pg were obtained under the optimized conditions for atomic absorption and emission of strontium, respectively. The coefficient of variation was 2.6% (11 determi- nations) for atomic absorption of 50 pg of strontium. The poor detection limit for atomic emission resulted from the higher noise level compared with that for atomic absorption. The atomic-absorption mode is recommended for the measurement of strontium because of its higher sensitivity and lower noise levels in absorption profiles.

Aluminium, calcium, magnesium and phosphate are some of the most important interfering substances in atomization of strontium in an air-acetylene flame. The effects of lOO-fold amounts of aluminium, magne- sium, calcium, sodium and phosphate on electrother- mal atomization of 50 pg of strontium were tested. The metals tested were added as their chlorides (and phosphate as the sodium salt). All the samples were checked for background effects and reagent blanks. Aluminium shifted the peak temperature (the tem- perature at which maximum absorption is observed) for strontium, although the appearance temperature and peak height remained the same (Fig. 9). Phos- phate also gave a small shift in the peak temperature. No interferences from calcium, sodium and magne- sium were observed. Almost identical effects were shown on the measurements of atomic emission of strontium, although a small depressant effect was observed for calcium.

Sample analysis

Some samples were analysed for strontium and the

Time , set Fig. 9. Memoriscope traces showing effect of aluminium on strontium absorption: a, 50 pg of Sr; b, 50 pg of Sr and 5 ng of Al; c, temperature increase. Gas flow-rates, Ar 450

ml/min and H2 50 ml/min.

Atomization of calcium and strontium 181

Table 1. Determination of strontium in some samples

Strontium, ~/ml

Sample Added Found4

Milk* - 0.53 0.10 0.64 0.20 0.76

Tap water? 0.02 I 0.006 0.028

Sea-water 6.0 2.5 8.7

* IOO-fold dilution. t 500-fold dilution. 5 Mean of three determinations.

Recovery, %

102 104

104

102

results are listed in Table 1. Dilution was required, because of the high sensitivity of strontium atomic absorption. Samples were diluted 100- and W-fold with demineralized water for milk and tap water re- spectively. Samples were spiked with strontium and the recovery found to be almost quantitative. Alkaline-earth elements are readily subject to con- tamination, so care must be taken in handling samples.

CONCLUSIONS

Atomic emission and ionization were observed in the electrothermal atomization of calcium and stron-

tium with a molybdenum micro-tube. Atomic absorp- tion of strontium could be measured without interference from these phenomena, but complex absorption profiles and pronounced atomic emission made it difficult to measure calcium accurately. The application of atomic emission in a molybdenum micro-tube appears to be worthy of further study for the determination of alkaline-earth metals.

Acknowledgement-This work was supported in part by a Grant-in-Aid for Special Project Research from the Min- istry of Education, Science and Culture, Japan.

1.

2.

3.

4.

5. 6. 7.

8. 9.

10. 11.

REFERENCES

J. H. Cragin and M. M. Meuron, At. Absorpr. NewsL, 1973, 12, 37. P. 0. Berggren, 0. Berglund and B. Hellman, Anal. Biochem., 1978, 84, 393. F. Bek, J. Janowskova and B. Moldam, At. Absorpt. Newsl., 1974, 13, 47. E. S. Gladney, L. E. Wangen and R. D. Aguilar, Anal. Lett., 1977, 10, 1083. J. M. Ottaway and F. Shaw, Analyst, 1975, 100, 438. Idem, ibid., 1976, 101, 582. B. Weltz, Atomic Absorption Spectroscopy, p. 137. Ver- lag Chemie, Weinheim, 1976. K. Ohta and M. Suzuki, Talanta, 1976, 23, 560. Idem, ibid., 1979, 26, 207. Idem, Anal. Chim. Acta, 1979, 104, 293. J. F. Alder, A. J. Samuel and R. D. Snook, Spectrochim. Acta, 1976, 31B, 509.