- Email: [email protected]
A triple-flow gas-sheathed d.c. arc for spectrochemical analysis
Amlyticrt
@:>Elscvicr
Chirrricw Acts. 74 (1975) 247-252 Scientific Publishing Company.
A TRIPLE-FLOW ANALYSIS
I-I. K. EL-KNOLY*.
GAS-SHEATHED
J. C. BURRIDGE
The Mrr~cwlc~y Itrstitwe (Reccivcd
24th July
247 Amstcrdnm
- Printed
D.C.
ARC
in The
Ncthcrlands
FOR
SPECTROCHEMICAL
cmd R. 0. SCOTT
jbr Soil Resectrch.
Crrrigiehlrcl~ler.
Ahertlret~
A BY 2Q.l (Scwtlrrrtl)
1974)
Although Stallwood’ introduced hisgas-sheathedarcmainly to reduce selective volatilization and matrix effects, stabilization of the arc and the exclusion of air from the arc plasma are the more usual advantages claimed for the commerciallyavailable Stallwood Jets. Boumans2 has recently summarized some applications of special excitation sources including jet devices. The reduction of cyanogen (CN) band emission is often of particular importance because these bands can prevent the analytical use of lines of several elements. Weber and Darr’, among other workers, used chambers to surround the arc with inert gas in order to reduce CN radiation to an acceptable level. Such chambers are inconvenient for routine analytical use on account of the frequent cleaning required, particularly when arcs are operated at currents above about 10 A. The gas-sheathed arc described in this paper was developed during the course of an investigation” into the effects of d.c. arc conditions on the determination of trace elements in soils and rocks, with a Hilger and Watts Large Quartz Spectrograph, and combines a twin-jet around the anode with an independent gas sheath around the cathode counter-electrode. It was found that when argon-oxygen mixtures are used to enclose the arc column, CN emission is virtually eliminated without the use of a surrounding chamber; the arc sheathed in this manner has good stability at currents up to 20 A; and the limits of detection for trace elements in soils and rocks are better than are obtained with a cathode-layer arc. SOURCE
DESCRIPTION
The triple-flow gas-sheath is maintained around the arc by means of a twin-jet device and a counter-flow silica tube, illustrated in Fig. 1 by a cross-sectional diagram. Two gas streams (I3 and C. Fig. 1) are directed in opposite directions around the anode through independent chambers (G and H) of an annular twin-jet device. Stream (B), an upward flowing mixture of argon and oxygen, provides the atmosphere for the arc plasma while the downward flowing stream of argon (C) reduces the consumption of the electrode material. The third gas stream (E), a downward current of argon in the silica tube (F), ensures that the arc strikes only to the tip of the cathode counter-electrode (A). The flow rates normally used are (B) 10 1 min - l of an argon-oxygen mixture obtained by * Present
address:
Fuculty
of Science.
University
of Sanaa.
Yemen.
248 A------
1-l. K. EL-KHOLY.
J. C. BURRIDGE.
R. 0.
SCOTT
e --e e_--__-
E
F
---_--G
--H
Fig. l.‘Schematic cross-scction of the arc system. The twin-jet was made from nn uluminium alloy (BS147G. HE14-W). (A) Curbon counter-ctcctrodc (cathode): (B) argon-oxygen inlet to upper chamber of twin-jet: (C) argon inlet to lower chamber of twin-jet: (D) insulating unodc-support guide; (E) argon inlet to sheath counter-electrode: (F) silica tube: (G) twin-jet upper annular chamber: (1-l) twin-jet lower annulor chamber: (I) argon outlets: (J) aluminium alloy unodc-support (sliding lit in D) for graphite unode.
combining flows of argon at 7 1 min-’ and oxygen at 3 1 min-‘, (C) 5 1 mine1 argon and (E) 3 1 min- l argon. An arc sheathed’ in the manner described above has been used chiefly for analyzing finely ground soil or rock samples. These are mixed 1:l with carbon powder and packed into a cavity 4.7-mm diameter by G-mm deep in r?. 6.15-mm diameter anode of graphite (National Carbon Company, grade AGKS). This cavity holds about 40 mg of sample mixture. A 5.5-mm diameter carbon rod (Morganite, grade SG9053) is used as the cathode and this projects some 2-3 mm beyond the silica tube (F in Fig. 1). Loss of sample when striking the arc is minimized by applying a few drops of a 5% solution of polymethyl methacrylate (Perspex) in chloroform to the filled electrode and allowing it to dry overnight, Throughout an exposure the tip of the sample electrode is kept 1 mm above the top surface of the twin-jet and a constant 5 mm arc gap is maintained by observation of a projected image of the arc. At 20 A the anode spot covers the sample completely and uniformly, gas flow (B) preventing the arc from wandering to the outer surface of the sample electrode. The rate of sample consumption is approximately ,proportional to the oxygen content of flow (B), and samples are consumed in about s min when a current bf 20 A is used with flow (B) containing 30”/, oxygen. The prototype twin-jet used in this study was made from an aluminiumbased alloy (Duralumin type ES1476 HE14-W) and despite the heat from the arc, water-cooling was not required. For the present application to the trace element analysis of soils and rocks,
A GAS-SI-IEATHED
DC.
ARC
FOR
249
SPECTROANALYSlS
silica was selected as the most suitable material to sheathe the counter-electrode (A), as it does not’create contamination problems. It would be possible to economize in the consumption of argon by using preset gas-control valves to reduce flow rates while electrodes are being changed. ARC
STABILI-&
Although the choice of internal standard is probably the major single factor affecting analytical precision. the stabilities of both arc current and plasma position are also of considerable importance and a smoothly burning arc is generally essential. Good stabilization of both plasma position and arc current has been achieved in the present work by the USC of the counter-flow (E). The .effect of this counter-flow on arc current is shown in Fig 2. Without the counter-flow (Fig. 2. curve a), there is a marked tendency for the arc to wander up the side of the counter-electrode, sometimes several centimetres from the tip, and a large reduction of arc current occurs when the arc plasma lengthens in this way. The strong upward flow of gas from the twin-jet hinders the return of the cathode spot to the tip of the counter-electrode. The improvement in current stability achieved by means of the counter-flow (Fig. 2. curve b). is very apparent. The a
b
------_a
20A-
2A
M
30.5
Fig. 2. Arc current fluctuation clcctrodc as shown in Fig. 1.
c
cl with
time:
Fig. 3. Uppa elcctroclcs 5.5 mm diameter and (b) with the counter-flow gas sheath. arced with the shplc cone-jet shown in and (c) arced in twin-jet shown in Fig. I.
(n)
without
and
(b)
with
d
gas
‘.c‘
sheath
around
atthodcs ;trced with twin-jet source, carbon: Lower elcctrodcs 6.15 Inm diumctcr graphite: Fig. 4, (d) un-arcctl saniplc clcctrodc cut to
countcr-
(a) without anodes. (c) show
crater
250
H. K. EL-KHOLY.
J. C. BURRIDGE,
R. 0. SCOTT
recordings shown were made while the arc was being controlled by means of manually operated current regulators. Only 10 A was obtained from a d-c. supply with normal resistance ballasting, the other 10 A being supplied by an unballasted rectifier fed from a hand-adjusted variable auto-transformer. A solid-state stabilizcdcurrent supply capable of continuous operation at selected currents up to 30 A, at present under construction, should further improve the arc stability. The efficiency with which the arc is confined to the counter-electrode tip is apparent in the photographs (Fig. 3) by comparing the shapes and surlice textures of electrodes (a) and (b). Both had flat ends before arcing and it is obvious from the photographs that electrode (a), arced without counter-flow, has been heated and eroded over a greater area than has electrode(b). with counter-flow. It can be seen from the photographs of arced anodes (Fig. 3) that the consumption of anode material is much lower with the triple-flow arc (electrode e) than with a Stallwood jet (electrode c) of the cone-shape illustrated in Fig. 4. The same argon-oxygen mixture and flow rate were used in both cases. Apart from the necessary consumption of the crater walls, shown in the cut-away unarced electrode (d) and weighing about 120 mg, there was an additional consumption of 280 mg of graphite for the Stallwood jet compared with one of 70 mg with the triple-flow arc.
/ B ,& ,; t.__.___--_--.
c
I
p
/>
~_...-.._
_-.-.._-
..q_
_
lcm
*--_.--
t ._
._.._____..
E
-
Fig. 4. Schematic cross-section of u simple cone-shaped jet. (A) Pcrrorntcd plate: (B) insulating anodesupport yuidc: (C) counter-clcctrodc (cathode): (D) argon--oxygen inlet; (E) nnodc support (sliding fit in B).
This action of the lower half of the twin-jet, in reducing the consumption of sample electrode material, helps to improve the trace element determination limits in two ‘different ways. First, the contribution of any contamination in the electrode material to the analysis is reduced, resulting in lower blank levels, and so the determination of low trace element contents in samples becomes possible without high-purity electrodes. Secondly, the smaller consumption of electrode carbon results in less background radiation with a consequent improvement in line-to-background ratios. A further measure of source stability is the reproducibility of emitted line intensity. This was determined from twelve replicate arcings of a sample containing
A GAS-SWEATHED
DC. ARC FOR SPECTROANALYSIS
251
100 p.p.m. of several trace elements in a synthetic soil matrix; a Hilger and Watts E492 Large Quartz Spectrograph was used with a seven-step rotating sector at the slit and a non-recording microphotometer. Line intensit coefficients of variation found were cobalt (3453.5 A)* 13.5, chromium (3578.7 x )_t 14.7, molybdenum (3170.4 A)-+_ 12.5 and tin (3175.0 A)* 21.9’Y. With palladium as internal standard for cobalt, and the line ratio Co 3453.5 &Pd 3460.8 A, the coefficient of variation for the cobalt content of 100 p.p.m. was t-4.5’%‘,. REDUCTION
OF CYANOGEN
BAND EMISSION
The United States Geological Survey standard diabase W-l was arced in the a tracing of part of the spectrum is shown in Fig. 5. The triple-flow source; absence of the ,main CN band systems, whose positions in a carbon arc in air arc indicated by the broken lines, demonstrates.the efficient suppression of cyanogen emission. Neither the band heads nor the main lines of the (0.0). (1.1) and (2,2) CN systems could be identified when the spectra were examined with a Judd-Lewis has reported the scandium content of this material as comparator. Flanagan” about 35 p.p,m. and the distinct tracing of the line ScII 3911.8 A in Fig. 5 illustrates that the triple-flow arc has reduced backgroun’d emission in this wavelength region sufficiently for such lines to be readily measured. D 00
OS 0.6
0.3
0.1
C’
Ll-___I CNbh(2.2) 3661.9J3
CNbh(l.1) 3671.48
I I
CNbh(O.O) 3663.4 8
--
Fe1 3886.7
8
SCIL 3911.6
--iG 8
---3920.3
Can
s,
3933.7
&
Fig. 5. Microphotometer tract of 11 spectrum from USGS rock W-l arced with twin-jet and shcathcd mixed with oxygen at 3 I min-’ was used to suppress counter-electrode; argon at 7 I min-’ cyanogen emission. The broken line indicates the positions of CN bands cmittcd by a curbon arc in ilir. Equipment: I-Ii&r and Watts E492 Large Quartz Spectrograph : arc imaged on the collimator: Word G30 Chromatic plate developed in ID2 1:2: Leeds Northrup microphotometcr with Spccdomax recorder. APPLICATION
TO SOIL ANALYSlS
The practical
application
of the triple-flow
arc has been
tested
by analyzing
I-I. K. EL-KHOLY.
252
J. C. BURRIDGE.
R. 0. SCOTT
some 60 soils and rocks for a wide range of trace elements; the detailed results are lower than can be have been presented elsewhere e. Limits of determination obtained with a cathode layer arc in air (Mitchell”). Some approximate values of these limits for the triple-flow arc, with a step sector at the spectrograph slit and with the arc imaged on the collimator, are arsenic (2349.8 A), 300 p.p.m.; bismuth (3067.7 A) and zinc (3345.0 A), 30 p.p.m.; tin (3175.0 A) and zirconium (3392.0 %(), lOp.p,m.: germanium (3039.1 A), indium (3039.4 A), lead (2833.1 A) and yttrium 3327.9 A), 3 p.p.m.: cobalt (3453.5 A), copper (3247.5 k), lanthanum (4333.7 A ). nickel (3414.8 A , scandium (3911.8 A) and vanadium (4379.2 A), 1 p.p.m.; beryllium (3130.4 d ). gallium (2943.6 A), manganese (2801.1 A) and silver (3280.7 A), 0.3 p.p.m. The lim’its of detection for the same lines, when spectrograms taken with the arc imaged on the spectrograph slit are examined with a Judd-Lewis comparator, are about one third of these quantitative determination limits. The improved detection limit for zinc is of particular value, as the zinc content of most soils and rocks is below the detection limit of about 300 p.p.m. the quantitative determination limit to in the cathode layer arcG. B y improving around 30 p.p.m., the triple-flow arc has already provided useful new information about the zinc content of Scottish soils4. The stability of this triple-flow arc, confirmed by the coeflicient of variation of less than 5”/;;for cobalt, and the good limits of determination indicated above for a range of elements, demonstrate the considerable potential value of this arc arrangement for the analysis of non-conducting powders. The twin-jet device was constructed in the Macaulay Institute instrument workshop and the authors wish to acknowledge the contribution of the workshop staff, in particular J. H. Normington and A. M. Fraser. SUMMARY
A d.c. arc source sheathed bjr three independent gas streams is described. A mixed flow of argon at 7 1 min- ’ and oxygen at 3 I min- ’ is used to support the arc plasma and isolate it from the surrounding air, while separate streams of argon are used to sheathe the base of the anode and lower part of the csithode counter-electrode, The anode holds about 40 mg of a sample mixture. The source is stable when operated with currents up to 20 A, cyanogen emission is eliminated, and the limits of detection of a wide range of trace elements in soils and rocks are better than with a cathode-layer arc in air. REFERENCES B. J. Stallwood. J. O/u. Sot. Arnw.. 44 (1954) 171. P. W. J. M. Boumuns in E. L. Grove (Ed.), Awlyticul Ewissio~r Specrroscopy, ptrrt II. Dckker. New York, 1972. Ch. 1. J. L. Weber and M. M. Durr. Ntrr. Bttr. Stur7tl. (U.S.) Twh. Note 582. 1972. p. 18. H. K. El-Kholy. Ph. D. Theis. University of hbcrdccn. 1972. F. J. %nagnn, Geoultiur. Cosmx*/tir~~, Ada. 37 ( 1973) 1 189. R. L. Mitchell. The Spectroclter~~ictrl Arral~sis (!f Soils. PImt,v crr~tl Reltrtetl Mnteriuls. Tech. Cornmrn. C’OI~WOIIH~. Bur.
Soils. MA.
1964.
@:>Elscvicr
Chirrricw Acts. 74 (1975) 247-252 Scientific Publishing Company.
A TRIPLE-FLOW ANALYSIS
I-I. K. EL-KNOLY*.
GAS-SHEATHED
J. C. BURRIDGE
The Mrr~cwlc~y Itrstitwe (Reccivcd
24th July
247 Amstcrdnm
- Printed
D.C.
ARC
in The
Ncthcrlands
FOR
SPECTROCHEMICAL
cmd R. 0. SCOTT
jbr Soil Resectrch.
Crrrigiehlrcl~ler.
Ahertlret~
A BY 2Q.l (Scwtlrrrtl)
1974)
Although Stallwood’ introduced hisgas-sheathedarcmainly to reduce selective volatilization and matrix effects, stabilization of the arc and the exclusion of air from the arc plasma are the more usual advantages claimed for the commerciallyavailable Stallwood Jets. Boumans2 has recently summarized some applications of special excitation sources including jet devices. The reduction of cyanogen (CN) band emission is often of particular importance because these bands can prevent the analytical use of lines of several elements. Weber and Darr’, among other workers, used chambers to surround the arc with inert gas in order to reduce CN radiation to an acceptable level. Such chambers are inconvenient for routine analytical use on account of the frequent cleaning required, particularly when arcs are operated at currents above about 10 A. The gas-sheathed arc described in this paper was developed during the course of an investigation” into the effects of d.c. arc conditions on the determination of trace elements in soils and rocks, with a Hilger and Watts Large Quartz Spectrograph, and combines a twin-jet around the anode with an independent gas sheath around the cathode counter-electrode. It was found that when argon-oxygen mixtures are used to enclose the arc column, CN emission is virtually eliminated without the use of a surrounding chamber; the arc sheathed in this manner has good stability at currents up to 20 A; and the limits of detection for trace elements in soils and rocks are better than are obtained with a cathode-layer arc. SOURCE
DESCRIPTION
The triple-flow gas-sheath is maintained around the arc by means of a twin-jet device and a counter-flow silica tube, illustrated in Fig. 1 by a cross-sectional diagram. Two gas streams (I3 and C. Fig. 1) are directed in opposite directions around the anode through independent chambers (G and H) of an annular twin-jet device. Stream (B), an upward flowing mixture of argon and oxygen, provides the atmosphere for the arc plasma while the downward flowing stream of argon (C) reduces the consumption of the electrode material. The third gas stream (E), a downward current of argon in the silica tube (F), ensures that the arc strikes only to the tip of the cathode counter-electrode (A). The flow rates normally used are (B) 10 1 min - l of an argon-oxygen mixture obtained by * Present
address:
Fuculty
of Science.
University
of Sanaa.
Yemen.
248 A------
1-l. K. EL-KHOLY.
J. C. BURRIDGE.
R. 0.
SCOTT
e --e e_--__-
E
F
---_--G
--H
Fig. l.‘Schematic cross-scction of the arc system. The twin-jet was made from nn uluminium alloy (BS147G. HE14-W). (A) Curbon counter-ctcctrodc (cathode): (B) argon-oxygen inlet to upper chamber of twin-jet: (C) argon inlet to lower chamber of twin-jet: (D) insulating unodc-support guide; (E) argon inlet to sheath counter-electrode: (F) silica tube: (G) twin-jet upper annular chamber: (1-l) twin-jet lower annulor chamber: (I) argon outlets: (J) aluminium alloy unodc-support (sliding lit in D) for graphite unode.
combining flows of argon at 7 1 min-’ and oxygen at 3 1 min-‘, (C) 5 1 mine1 argon and (E) 3 1 min- l argon. An arc sheathed’ in the manner described above has been used chiefly for analyzing finely ground soil or rock samples. These are mixed 1:l with carbon powder and packed into a cavity 4.7-mm diameter by G-mm deep in r?. 6.15-mm diameter anode of graphite (National Carbon Company, grade AGKS). This cavity holds about 40 mg of sample mixture. A 5.5-mm diameter carbon rod (Morganite, grade SG9053) is used as the cathode and this projects some 2-3 mm beyond the silica tube (F in Fig. 1). Loss of sample when striking the arc is minimized by applying a few drops of a 5% solution of polymethyl methacrylate (Perspex) in chloroform to the filled electrode and allowing it to dry overnight, Throughout an exposure the tip of the sample electrode is kept 1 mm above the top surface of the twin-jet and a constant 5 mm arc gap is maintained by observation of a projected image of the arc. At 20 A the anode spot covers the sample completely and uniformly, gas flow (B) preventing the arc from wandering to the outer surface of the sample electrode. The rate of sample consumption is approximately ,proportional to the oxygen content of flow (B), and samples are consumed in about s min when a current bf 20 A is used with flow (B) containing 30”/, oxygen. The prototype twin-jet used in this study was made from an aluminiumbased alloy (Duralumin type ES1476 HE14-W) and despite the heat from the arc, water-cooling was not required. For the present application to the trace element analysis of soils and rocks,
A GAS-SI-IEATHED
DC.
ARC
FOR
249
SPECTROANALYSlS
silica was selected as the most suitable material to sheathe the counter-electrode (A), as it does not’create contamination problems. It would be possible to economize in the consumption of argon by using preset gas-control valves to reduce flow rates while electrodes are being changed. ARC
STABILI-&
Although the choice of internal standard is probably the major single factor affecting analytical precision. the stabilities of both arc current and plasma position are also of considerable importance and a smoothly burning arc is generally essential. Good stabilization of both plasma position and arc current has been achieved in the present work by the USC of the counter-flow (E). The .effect of this counter-flow on arc current is shown in Fig 2. Without the counter-flow (Fig. 2. curve a), there is a marked tendency for the arc to wander up the side of the counter-electrode, sometimes several centimetres from the tip, and a large reduction of arc current occurs when the arc plasma lengthens in this way. The strong upward flow of gas from the twin-jet hinders the return of the cathode spot to the tip of the counter-electrode. The improvement in current stability achieved by means of the counter-flow (Fig. 2. curve b). is very apparent. The a
b
------_a
20A-
2A
M
30.5
Fig. 2. Arc current fluctuation clcctrodc as shown in Fig. 1.
c
cl with
time:
Fig. 3. Uppa elcctroclcs 5.5 mm diameter and (b) with the counter-flow gas sheath. arced with the shplc cone-jet shown in and (c) arced in twin-jet shown in Fig. I.
(n)
without
and
(b)
with
d
gas
‘.c‘
sheath
around
atthodcs ;trced with twin-jet source, carbon: Lower elcctrodcs 6.15 Inm diumctcr graphite: Fig. 4, (d) un-arcctl saniplc clcctrodc cut to
countcr-
(a) without anodes. (c) show
crater
250
H. K. EL-KHOLY.
J. C. BURRIDGE,
R. 0. SCOTT
recordings shown were made while the arc was being controlled by means of manually operated current regulators. Only 10 A was obtained from a d-c. supply with normal resistance ballasting, the other 10 A being supplied by an unballasted rectifier fed from a hand-adjusted variable auto-transformer. A solid-state stabilizcdcurrent supply capable of continuous operation at selected currents up to 30 A, at present under construction, should further improve the arc stability. The efficiency with which the arc is confined to the counter-electrode tip is apparent in the photographs (Fig. 3) by comparing the shapes and surlice textures of electrodes (a) and (b). Both had flat ends before arcing and it is obvious from the photographs that electrode (a), arced without counter-flow, has been heated and eroded over a greater area than has electrode(b). with counter-flow. It can be seen from the photographs of arced anodes (Fig. 3) that the consumption of anode material is much lower with the triple-flow arc (electrode e) than with a Stallwood jet (electrode c) of the cone-shape illustrated in Fig. 4. The same argon-oxygen mixture and flow rate were used in both cases. Apart from the necessary consumption of the crater walls, shown in the cut-away unarced electrode (d) and weighing about 120 mg, there was an additional consumption of 280 mg of graphite for the Stallwood jet compared with one of 70 mg with the triple-flow arc.
/ B ,& ,; t.__.___--_--.
c
I
p
/>
~_...-.._
_-.-.._-
..q_
_
lcm
*--_.--
t ._
._.._____..
E
-
Fig. 4. Schematic cross-section of u simple cone-shaped jet. (A) Pcrrorntcd plate: (B) insulating anodesupport yuidc: (C) counter-clcctrodc (cathode): (D) argon--oxygen inlet; (E) nnodc support (sliding fit in B).
This action of the lower half of the twin-jet, in reducing the consumption of sample electrode material, helps to improve the trace element determination limits in two ‘different ways. First, the contribution of any contamination in the electrode material to the analysis is reduced, resulting in lower blank levels, and so the determination of low trace element contents in samples becomes possible without high-purity electrodes. Secondly, the smaller consumption of electrode carbon results in less background radiation with a consequent improvement in line-to-background ratios. A further measure of source stability is the reproducibility of emitted line intensity. This was determined from twelve replicate arcings of a sample containing
A GAS-SWEATHED
DC. ARC FOR SPECTROANALYSIS
251
100 p.p.m. of several trace elements in a synthetic soil matrix; a Hilger and Watts E492 Large Quartz Spectrograph was used with a seven-step rotating sector at the slit and a non-recording microphotometer. Line intensit coefficients of variation found were cobalt (3453.5 A)* 13.5, chromium (3578.7 x )_t 14.7, molybdenum (3170.4 A)-+_ 12.5 and tin (3175.0 A)* 21.9’Y. With palladium as internal standard for cobalt, and the line ratio Co 3453.5 &Pd 3460.8 A, the coefficient of variation for the cobalt content of 100 p.p.m. was t-4.5’%‘,. REDUCTION
OF CYANOGEN
BAND EMISSION
The United States Geological Survey standard diabase W-l was arced in the a tracing of part of the spectrum is shown in Fig. 5. The triple-flow source; absence of the ,main CN band systems, whose positions in a carbon arc in air arc indicated by the broken lines, demonstrates.the efficient suppression of cyanogen emission. Neither the band heads nor the main lines of the (0.0). (1.1) and (2,2) CN systems could be identified when the spectra were examined with a Judd-Lewis has reported the scandium content of this material as comparator. Flanagan” about 35 p.p,m. and the distinct tracing of the line ScII 3911.8 A in Fig. 5 illustrates that the triple-flow arc has reduced backgroun’d emission in this wavelength region sufficiently for such lines to be readily measured. D 00
OS 0.6
0.3
0.1
C’
Ll-___I CNbh(2.2) 3661.9J3
CNbh(l.1) 3671.48
I I
CNbh(O.O) 3663.4 8
--
Fe1 3886.7
8
SCIL 3911.6
--iG 8
---3920.3
Can
s,
3933.7
&
Fig. 5. Microphotometer tract of 11 spectrum from USGS rock W-l arced with twin-jet and shcathcd mixed with oxygen at 3 I min-’ was used to suppress counter-electrode; argon at 7 I min-’ cyanogen emission. The broken line indicates the positions of CN bands cmittcd by a curbon arc in ilir. Equipment: I-Ii&r and Watts E492 Large Quartz Spectrograph : arc imaged on the collimator: Word G30 Chromatic plate developed in ID2 1:2: Leeds Northrup microphotometcr with Spccdomax recorder. APPLICATION
TO SOIL ANALYSlS
The practical
application
of the triple-flow
arc has been
tested
by analyzing
I-I. K. EL-KHOLY.
252
J. C. BURRIDGE.
R. 0. SCOTT
some 60 soils and rocks for a wide range of trace elements; the detailed results are lower than can be have been presented elsewhere e. Limits of determination obtained with a cathode layer arc in air (Mitchell”). Some approximate values of these limits for the triple-flow arc, with a step sector at the spectrograph slit and with the arc imaged on the collimator, are arsenic (2349.8 A), 300 p.p.m.; bismuth (3067.7 A) and zinc (3345.0 A), 30 p.p.m.; tin (3175.0 A) and zirconium (3392.0 %(), lOp.p,m.: germanium (3039.1 A), indium (3039.4 A), lead (2833.1 A) and yttrium 3327.9 A), 3 p.p.m.: cobalt (3453.5 A), copper (3247.5 k), lanthanum (4333.7 A ). nickel (3414.8 A , scandium (3911.8 A) and vanadium (4379.2 A), 1 p.p.m.; beryllium (3130.4 d ). gallium (2943.6 A), manganese (2801.1 A) and silver (3280.7 A), 0.3 p.p.m. The lim’its of detection for the same lines, when spectrograms taken with the arc imaged on the spectrograph slit are examined with a Judd-Lewis comparator, are about one third of these quantitative determination limits. The improved detection limit for zinc is of particular value, as the zinc content of most soils and rocks is below the detection limit of about 300 p.p.m. the quantitative determination limit to in the cathode layer arcG. B y improving around 30 p.p.m., the triple-flow arc has already provided useful new information about the zinc content of Scottish soils4. The stability of this triple-flow arc, confirmed by the coeflicient of variation of less than 5”/;;for cobalt, and the good limits of determination indicated above for a range of elements, demonstrate the considerable potential value of this arc arrangement for the analysis of non-conducting powders. The twin-jet device was constructed in the Macaulay Institute instrument workshop and the authors wish to acknowledge the contribution of the workshop staff, in particular J. H. Normington and A. M. Fraser. SUMMARY
A d.c. arc source sheathed bjr three independent gas streams is described. A mixed flow of argon at 7 1 min- ’ and oxygen at 3 I min- ’ is used to support the arc plasma and isolate it from the surrounding air, while separate streams of argon are used to sheathe the base of the anode and lower part of the csithode counter-electrode, The anode holds about 40 mg of a sample mixture. The source is stable when operated with currents up to 20 A, cyanogen emission is eliminated, and the limits of detection of a wide range of trace elements in soils and rocks are better than with a cathode-layer arc in air. REFERENCES B. J. Stallwood. J. O/u. Sot. Arnw.. 44 (1954) 171. P. W. J. M. Boumuns in E. L. Grove (Ed.), Awlyticul Ewissio~r Specrroscopy, ptrrt II. Dckker. New York, 1972. Ch. 1. J. L. Weber and M. M. Durr. Ntrr. Bttr. Stur7tl. (U.S.) Twh. Note 582. 1972. p. 18. H. K. El-Kholy. Ph. D. Theis. University of hbcrdccn. 1972. F. J. %nagnn, Geoultiur. Cosmx*/tir~~, Ada. 37 ( 1973) 1 189. R. L. Mitchell. The Spectroclter~~ictrl Arral~sis (!f Soils. PImt,v crr~tl Reltrtetl Mnteriuls. Tech. Cornmrn. C’OI~WOIIH~. Bur.
Soils. MA.
1964.