Publisher’s version / Version de l'éditeur:
Journal of Analytical Atomic Spectrometry, 14, 7, pp. 1019-1024, 1999
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE. https://nrc-publications.canada.ca/eng/copyright
Vous avez des questions? Nous pouvons vous aider. Pour communiquer directement avec un auteur, consultez la
première page de la revue dans laquelle son article a été publié afin de trouver ses coordonnées. Si vous n’arrivez
Questions? Contact the NRC Publications Archive team at
PublicationsArchive-ArchivesPublications@nrc-cnrc.gc.ca. If you wish to email the authors directly, please see the first page of the publication for their contact information.
NRC Publications Archive
Archives des publications du CNRC
This publication could be one of several versions: author’s original, accepted manuscript or the publisher’s version. / La version de cette publication peut être l’une des suivantes : la version prépublication de l’auteur, la version acceptée du manuscrit ou la version de l’éditeur.
For the publisher’s version, please access the DOI link below./ Pour consulter la version de l’éditeur, utilisez le lien DOI ci-dessous.
https://doi.org/10.1039/a901850f
Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
Reduction of metal oxides by carbon in graphite furnaces. Part 1.
Temporal oscillations of atomic absorption in the process of slow
evaporation of Al, Bi, Cr, In, Mg, Mn, Pb, Sb, Sn and Te oxides
L'vov, B. V.; Vasilevich, A. A.; Dyakov, A. O.; Lam, J.; Sturgeon, R.
https://publications-cnrc.canada.ca/fra/droits
L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.
NRC Publications Record / Notice d'Archives des publications de CNRC:
https://nrc-publications.canada.ca/eng/view/object/?id=68689467-6a93-47d8-af49-02405eebce14
https://publications-cnrc.canada.ca/fra/voir/objet/?id=68689467-6a93-47d8-af49-02405eebce14
Reduction of metal oxides by carbon in graphite furnaces
Part 1. Temporal oscillations of atomic absorption in the process of slow evaporation
of Al, Bi, Cr, In, Mg, Mn, Pb, Sb, Sn and Te oxides
Boris V. L’vov,a Andrey A. Vasilevich,a Alexey O. Dyakov,a Joseph W. H. Lamb and
Ralph E. Sturgeonb
aDepartment of Analytical Chemistry, St. Petersburg State Technical University, St.
Petersburg 195251, Russia
bInstitute for National Measurement Standards, National Research Council, Ottawa, Ontario
K1A 0R6, Canada
Received 9th March 1999, Accepted 13th May 1999
Temporal oscillations in the kinetics of carbothermal reduction of oxides of 10 elements (Al, Bi, Cr, In, Mg, Mn, Pb, Sb, Sn and Te) have been observed by electrothermal atomic absorption spectrometry. Microgram amounts of the elements, as their nitrates, were evaporated with slow heating of the graphite tube. With the exception of Al, Cr and Mn, these oscillations were observed for the first time. Some of the features of this phenomenon are described, including the effects of sample mass, acidity of the solution, nature of the graphite substrate, the influence of the presence of Sr and the type of sheathing gas ( He and Ar) on the characteristics of Al spikes. Similar results were obtained in two different laboratories.
The appearance of spikes on absorption signals under con-
Experimental
ditions of comparatively slow heating (5–20 °C s−1) ofmicro-Instrumentation and reagents gram masses of alumina in a graphite furnace (GF ) was first
reported in 1981.1 This phenomenon was subsequently sub- Measurements made in both laboratories used Perkin-Elmer jected to a number of studies, mainly by L’vov et al.,2–27 as a (Norwalk, CT, USA) Model 5000 spectrometers equipped result of which a gaseous carbide mechanism for the reduction with HGA-500 graphite furnaces and AS-40 autosamplers. of oxides by carbon (ROC ) was proposed.21,22 The foundation Hollow-cathode lamps (HCL’s), electrodeless discharge lamps of this mechanism is the assumption that the oxide is directly ( EDL’s) and a deuterium lamp (D
2) were used as light sources. reduced by gaseous molecules of metal carbides which, in their Absorption signals were registered using personal computers turn, form by the interaction of metal vapor with carbon, e.g. and laboratory developed programs. Standard pyrolytic graph-by the reactions: ite coated tubes were employed in most cases. Argon contain-ing not more than 1×10−3 % oxygen served as a sheath gas. M(g)+2C(s)MC
2(g) (1) Helium was used for the internal purge gas in some MC
2(g)+2MO(s)3M(g)+2 CO (2) experiments. The first of these reactions, initiated by defects in the carbon
structure, occurs on the graphite surface whereas the second
Procedures occurs on that of the oxide. Reactions (1) and (2) determine
the boundary conditions for the fluxes of gaseous products The first stage of the study consisted of the selection of optimum experimental conditions for each of the investigated between the solid reactants. Based on this model, many specific
features of the process, in particular, the periodicity of the elements in order to record the absorbance signal during the process of slow evaporation of microgram masses of metal spikes noted for Al,21 Mn,21,26 Tm,2,21,27 and Yb21,25 oxides
were explained.21,22 For this reason, it is appropriate to oxides. For this purpose, secondary low-sensitivity lines were used. In some cases, such lines are absent (Mg and Mn) or consider below not only the appearance of fast single spikes,
but more generally, temporal oscillations in the kinetics of are located in the visible part of the spectrum (Cr), incon-venient for the absorption measurement at high tube tempera-carbothermal reduction of oxides. In the early 90’s, a number
of publications appeared,28–33 wherein various groups of tures. Because of this, a D
2-lamp, in combination with a narrow slit width, was used in these cases. The second stage researchers critically discussed details of the gaseous-carbide
mechanism, as applied to reduction of alumina, and questioned of the study consisted of the selection of optimum conditions for the generation of temporal oscillations. Several factors its validity. L’vov34 subsequently responded to all of the
arguments presented against it. were varied: the concentration, acidity and volume of solution injected; substrate type (uncoated and pyrolytic graphite The purpose of this study was to investigate temporal
oscillations for some elements from different Groups of the coated, new and old tubes); wall and platform evaporation; rate of heating during the dry and atomization stages; and Periodic Table (in addition to Al, Cr and Mn) and to examine
the general features of this phenomenon. One may speculate internal flow rate of argon. In some cases, 0.1–10 mg of Sr, as its nitrate, was added to the sample. All experiments were that the effects noted should also be evident for additional
elements, but are unlikely to arise for those of the iron and repeated at least 5–10 times to obtain a measure of the reproducibility of the results.
platinum groups, alkali metals or Ag, Au, Cu and Hg. A
second thrust was to compare the experimental results obtained The parameters for atomic absorption measurements and the heating programs used in both laboratories are given in for the same elements in two different laboratories: the St.
Petersburg State Technical University in Russia ( lab 1) and Tables 1 and 2 (only those parameters were included which correspond to the results presented below).
the National Research Council of Canada ( lab 2).
istics of these temporal oscillation profiles: mean values of spike widths and intervals between the spikes. The scatter in the values of these measured parameters is in the range 20–30% and is typical of the reproducibility of these phenomena. Activation energies
The most convincing argument for the ROC origin of these temporal oscillations is the apparent value of the activation energy, Ea (summarised in Table 4), associated with the rising part of the spike22:
E
a=R ln (A2/A1)/(T1−1−T2−1) (3) Here R is the gas constant and the absorbance and correspond-ing temperature values used for the calculation of E
a(A1, A2 and T 1 , T2) are taken from the profiles presented in Figs. 1–10. The error in calculations associated with an uncer-tainty in temperature calibration of the HGA-500 power supply (within ±100 °C ) is not higher than 20%. As evident from Table 4, these apparent E
avalues for the spikes are much higher than the theoretical values of activation energies calcu-lated35 for the process of thermal dissociation of oxides in the Fig. 1 Oscillation profiles for Al recorded under conditions given in equimolar mode for all the elements investigated. This clearly Tables 2 (a) and 1 (b).
demonstrates the accelerating character of the process, which is typical for the autocatalytic ROC mechanism. A substantial
Results and discussion
longitudinal non-isothermality in the temperature of the graph-Atomic absorption spike profiles ite tube occurs when HGA-type atomizers are used and thus one must entertain the possibility that the oscillations are due Figs. 1–10 contain some typical spike profiles for Al, Bi, Cr,
to analyte condensation/re-vaporisation phenomena. In light In, Mg, Mn, Pb, Sb, Sn and Te. For 7 of these 10 elements
of the element specificity of the effect, however, it is unlikely (exceptions being Al, Cr and Mn), these temporal oscillations
that this can be considered as correct. This problem has earlier are reported here for the first time. For all elements, the
been discussed in detail by L’vov.34 oscillation profiles obtained in the two different laboratories
were very similar (see Figs. 3, 4, 6, 7 and 9, for illustration).
Characteristics of oxides From a cursory examination of these data, some general
differences in the shape and position of the spikes becomes It is interesting to correlate the differences in spike profiles for the elements investigated with any differences in the character-apparent. For Al, Cr, In, Sb, Sn and Te, they are located on
the front of the thermal dissociation pulse, but for Bi, Mg, Mn, istics of their oxides. In Table 5, the melting points and enthalpies of formation for these oxides36 are presented. As Pb, on its tail. Additionally, two different types of ‘spikes’ arise:
fast and sharp spikes for Al and Mn and smooth oscillations can be seen from the comparison of the melting points with the temperatures corresponding to the oscillation process, only for the other elements. In some cases (Sb, for example), spikes
of both types occur. Table 3 contains some additional character- oxides of Mg, In and Cr remain in the solid state during the Table 1 Experimental conditions used in lab 1 for observation of spike profiles presented in Figs. 1–10
Heating program
Heating
Wave- Light Current/ Mass/ Volume/ Temperature/ Ramp time/ Hold rate/ Ar flow/
Element length/nm sourcea mA Slit/nm mg ml °C s time/s °C s−1 ml min−1 Solution
Al 309.3 HCL(LSP-1) 15 0.07 11 5 130 20 40 1%HNO 3 1500 1 5 2300 80 0 10 50 Bi 227.7 HCL(LSP-1) 20 0.7 4 20 150 10 40 20%HNO3 800 1 5 1700 75 0 12 50 Cr 429.0 HCL(LSP-1) 25 0.07 2 20 150 10 40 10%HNO3 1500 1 5 2300 70 0 11 300 In 410.2 HCL(LSP-1) 20 0.07 3 20 150 10 40 10%HNO3 1200 1 5 1900 60 0 12 20 Mg 285.2 D2 — 0.07 11 5 150 10 40 1%HNO3 1500 1 5 2400 60 0 15 20 Mn 222.2 HCL(P-E) 20 0.2 2 20 150 10 40 3%HNO 3 1300 1 5 2100 80 0 10 10 Pb 205.3 HCL (P-E ) 12 0.7 5 20 150 10 40 20%HNO 3 600 1 10 1800 70 0 17 30
Sb 231.1 EDL ( P-E) 400 0.2 19 20 200 1 5 20%HNO
3 110 1 50 1200 5 5 1900 60 0 12 50 Sn 254.7 HCL(P-E) 20 0.2 13 20 150 10 40 5%H2SO4 1100 5 5 2200 70 0 16 30 Te 225.9 EDL(P-E ) 400 0.2 6 20 130 1 3 10%HNO3 150 30 20 800 1 5 2000 60 0 20 50
aP-E: Perkin-Elmer; LSP-1: manufactured in Russia.
Table 2 Experimental conditions used in lab 2 for observation of spike profiles presented in Figs. 1–14 Heating program
Heating Internal Wavelength/ Light Current/ Slit/ Mass/ Volume/ Temperature/ Ramp Hold rate/ gas flow/
Element nm sourcea mA nm mg ml °C time/s time/s °C s−1 ml min−1 Solution
Al 309.3 HCL( H ) 15 1 5 150 1 40 1%HCl 0.07 (+2 mgSr) (+5 mlSr) 1700 1 5 2100 60 0 7 20(Ar) Al 309.3 HCL( H ) 15 0.2 9 20 150 1 40 20%HNO 3 (+1 mgSr) (+5 mlSr) 1950 1 5 2250 50 0 6 40( He) Cr 357.9 D2 — 17 20 150 10 40 4%HNO3 0.07 1500 1 5 1900 30 0 13 300(Ar) In 410.2 HCL(C ) 6 0.2 4 10 150 10 40 10%HNO3 1100 1 5 1800 60 0 12 30(Ar) Mn 279.5 D 2 — 0.07 19(+1 mgSr) 15(+5 mlSr) 1400150 101 405 3%HNO3 2100 60 0 12 50(Ar) Pb 205.3 HCL( H ) 15 0.2 5 20 150 10 40 5%HNO3 700 1 10 1700 60 0 17 30(Ar) Sn 254.7 HCL( H ) 15 0.7 10 20 150 10 40 30%HNO 3 1100 5 5 2200 60 0 18 50(Ar) Sn 254.7 HCL( H ) 15 0.7 10 20 180 1 40 4%H2SO4 300 5 5 500 1 5 2200 50 0 34 20( He)
aC: Cathoden; H: Hamamatsu.
Fig. 2 Oscillation profile for Bi recorded under conditions given in Table 1.
Fig. 4 Oscillation profiles for In recorded under conditions given in Table 2 (a) and 1 (b).
Fig. 3 Oscillation profiles for Cr recorded under conditions given in Fig. 5 Oscillation profile for Mg recorded under conditions given in Table 1.
Tables 2 (a) and 1 (b).
Fig. 8 Oscillation profile for Sb recorded under conditions given in Table 1.
Fig. 6 Oscillation profiles for Mn recorded under conditions given in Tables 2 (a) and 1 (b).
Table 3 Spike characteristics
Meana Meana temperature Heating spike interval Mass/ rate/ Ar flow/ width/ between
Element mg °C s−1 ml min−1 °C spikes/°C Reference
Al 1 10 20 12 89 Fig. 1a Bi 4 12 50 29 67 Fig. 2 Cr 2 11 300 20 57 Fig. 3b In 3 12 20 35 88 Fig. 4b Mg 11 15 20 30 60 Fig. 5b Mn 2 10 10 17 38 Fig. 6b Pb 5 17 30 47 75 Fig. 7a Sb 19 12 50 9 30 Fig. 8 Sn 13 16 30 26 72 Fig. 9b Te 10 20 50 42 107 Fig. 10 aTypical reproducibility is ±20–30% RSD.
Fig. 7 Oscillation profiles for Pb recorded under conditions given in Tables 2 (a) and 1 (b).
oscillation period. Oxides of Pb, Sb, Bi and Te are in a liquid state from the very beginning of the oscillations and oxides of Al, Sn and Mn are in a solid state in the beginning, and in a liquid state by the end of the oscillation period. Despite these significant differences in the physical states of the oxides during the oscillation process, no differences in the character of the oscillations can be noted.
Very significant differences in the enthalpies of formation for these oxides also exist. For Al
2O3 (the most stable) and PbO (the most unstable), the enthalpies (per one metal atom)
differ by about 4-fold. This explains, firstly, the difference in Fig. 9 Oscillation profiles for Sn recorded under conditions given in Tables 2 (a) and 1 (b).
the appearance temperatures of the spikes for these elements (2000 K and 1100 K, respectively) and, secondly, the dramatic difference in associated activation energy values ( Table 4). It
Effect of sample mass is also possible to relate this enthalpy distinction to the
difference in the width of the spikes ( Table 3). Nevertheless, As an illustration of the effect of analyte mass on the width of the spikes, Fig. 11 and Table 6 present the results obtained this last correlation does not appear to be very reliable or, in
any case, unambiguous. As shown later, the width of the for aluminium. In all cases, the width of the first spike was utilized. It is clear that the width varies nearly proportionally spikes, to a large extent, is a function of sample mass and
graphite substrate conditions. to the analyte mass in the range 10 to 200 mg Al.
Effect of solution acidity
In some cases an increase in solution acidity stimulated temporal oscillations. Therefore, practically all solutions were acidified to contain up to 20–30% HNO
3or H2SO4(for tin). This effect may be associated with the activation of the graphite surface and, as a result, with formation of gaseous carbides.7 Effect of graphite substrate
Graphite substrate conditions are of primary importance among other parameters which affect the generation of tem-poral oscillations. Pyrolytic graphite coated tubes are prefer-able in all cases. For most of the elements tested, new tubes (up to 20–30 firings) promote better results than used ones.
In some cases, the influence of the graphite substrate on the width of the spikes was evident. The width of spikes generated in the process of carbothermal reduction of MnO from a pyrolytic graphite platform (Fig. 12) is about 6-fold less than those from a pyrolytic graphite coated tube wall ( Fig. 6). Fig. 10 Oscillation profile for Te recorded under conditions given
Additionally, a difference in the position of spikes on the front in Table 1.
( Fig. 12) and on the tail ( Fig. 6) is evident. Unfortunately, temporal oscillations for samples deposited on the platform Table 4 Activation energies for the elements investigated are more difficult to generate, likely as a result of a difference in the structure of the pyrolytic graphite. This difference in
E
aa/ Eab/ the spike shapes requires further consideration. Element T
1/K T2/K A1 A2 kJ mol−1 kJ mol−1 Al 2059 2064 0.05 1.71 25000 651
Effect of sheath gas
Bi 1301 1307 0.66 0.87 650 257
Cr 1783 1848 0.05 0.55 1010 537
In 1523 1583 0.02 0.30 900 394 Experiments on the observation of temporal oscillations were Mg 1783 1938 0.02 0.81 690 489 performed using helium as a sheath gas for Al and Sn. For
Mn 1400 1803 0.02 0.76 700 436
Pb 1057 1103 0.02 0.35 600 268 Al, the spike width in He ( Fig. 13) is less than that in Ar Sb 1378 1403 0.04 0.66 1800 329
( Fig. 1b). In contrast, for Sn, the spike width in He ( Fig. 14)
Sn 1618 1626 0.13 1.28 6250 432
Te 1638 1663 0.54 0.91 470 233 is larger than that in Ar ( Fig. 9b). An interesting feature of aEa: apparent activation energy; bEa: calculated for thermal dissociation (e-mode).35 this trace is the appearance of a very sharp spike at 2000 °C. In both cases, the shift of oscillations in He to about 200– 300 °C higher apparent tube wall temperature is connected Table 5 Thermal characteristics of oxides36
with the much higher thermal conductivity of He and, as a
Group Oxide Tm/K -DfH°298/kJ mol−1
IIa MgO 3100 601.5 IIIa Al2O3 2327 1675.7 IIIa In2O3 2186 926.3 IVa SnO2 1903 557.6 IVa PbO 1160 218.6 Va Sb2O3 928 699.6 Va Bi2O3 1090 578.2 VIb Cr 2O3 2705 1140.6 VIa TeO 2 1006 322.0 VIIa MnO 2058 385.4
Table 6 Dependence of spike width on mass of Ala
Mass/mg Width/s 10 1.5 20 2.8 40 6.0 80 11.0 100 14 200 26
Fig. 12 Oscillation profile for 3.5 mg Mn evaporated from a pyrolytic aTypical reproducibility is 20–30% RSD. graphite platform at 10 °C s−1 heating rate and 300 ml min−1 internal
flow of Ar.37
Fig. 11 First spike of aluminium for (a) 10; (b) 40 and (c) 100 mg Al.
tive or phenomenological investigation, with little comprehen-sive interpretation of the data. The primary purpose of this report was to stimulate further research into this effect by the analytical community.
Acknowledgement
Financial support from the Royal Society of Chemistry to B. V. L. for his visit to the Institute for National Measurement Standards, NRCC, is gratefully acknowledged.
References
Fig. 13 Oscillation profile for Al recorded under conditions given in
1 B. V. L’vov, in Proceedings of the XII Mendeleev Congress on Table 2 (in He).
General and Applied Chemistry, Part I, Nauka, Moscow, 1981, p. 286.
2 B. V. L’vov and A. S. Savin, Zh. Anal. Khim., 1982, 37, 2116. 3 B. V. L’vov and A. S. Savin, Zh. Anal. Khim., 1983, 38, 1925. 4 B. V. L’vov and A. S. Savin, Zh. Anal. Khim., 1983, 38, 1933. 5 B. V. L’vov, Dokl. Akad. Nauk SSSR, 1983, 271, 119.
6 B. V. L’vov and L. F. Yatsenko, Zh. Anal. Khim., 1984, 39, 1773. 7 B. V. L’vov, Zh. Anal. Khim., 1984, 39, 1953.
8 B. V. L’vov, Izv. Akad. Nauk SSSR, Ser. Metally., 1984, 5, 3. 9 B. V. L’vov, Dokl. Akad. Nauk SSSR, 1985, 283, 1415.
10 B. V. L’vov, A. S. Savin and L. F. Yatsenko, Zh. Prikl. Spektrosk., 1985, 43, 887.
11 B. V. L’vov, Izv. Vuzov, Chern. Metall., 1986, 1, 4.
12 B. V. L’vov and L. F. Yatsenko, Izv. Vuzov, Chern. Metall., 1986, 5, 1.
13 B. V. L’vov, Kinet Katal, 1986, 27, 1513. 14 B. V. L’vov, Kinet Katal, 1986, 27, 1513. 15 B. V. L’vov, J. Anal. At. Spectrom.,1987, 2, 95. Fig. 14 Oscillation profile for Sn recorded under conditions given in
16 B. V. L’vov, Khim. Zh., 1987, 5, 30. Table 2 (in He).
17 B. V. L’vov, V. G. Nikolaev, A. V. Ilyukhin and H.-Y. Stark, Khim. Tverd. Topliva, 1988, 1, 114.
18 B. V. L’vov, V. G. Nikolaev, A. V. Ilyukhin and H.-Y. Stark, Khim. result, with the difference in the real tube temperature in
Tverd. Topliva, 1988, 1, 118.
different gases at the same input power. 19 B. V. L’vov and A. V. Ilyukhin, Izv. Vuzov, Chern. Metall., 1988, 9, 10.
Effect of Sr addition 20 B. V. L’vov, L. K. Polzik and A. I. Yuzefovskii, Zh. Anal. Khim., 1989, 44, 794.
As noted earlier,34 addition of Sr greatly stimulates temporal 21 B. V. L’vov, Spectrochim. Acta, Part B, 1989, 44, 1257.
oscillations in the process of carbothermal reduction of Al 22 B. V. L’vov, L. K. Polzik, N. P. Romanova and A. I. Yuzefovskii, J. Anal. At. Spectrom., 1990, 5, 163.
and Mn oxides in a spatially isothermal transverse-heated
23 B. V. L’vov and N. P. Romanova, Zh. Prikl. Spectrosk., 1990, graphite atomizer ( THGA). The frequency of oscillations and
53, 664. their regularity in the presence of a small addition of Sr (as
24 B. V. L’vov and N. P. Romanova, Zh. Anal. Khim., 1990, 45, 506. its nitrate) were much higher than those in the absence of Sr. 25 B. V. L’vov, L. K. Polzik and A. I. Yuzefovskii, Zh. Anal. Khim., Such comparative experiments were repeated with all the 1990, 45, 920.
26 B. V. L’vov and N. P. Romanova, Zh. Anal. Khim., 1991, 46, 461. elements using non-isothermal HGA tubes. The same
stimula-27 B. V. L’vov, L. K. Polzik and N. P. Romanova, Zh. Anal. Khim., tion of oscillations was observed for Al (Fig. 1) and Mn
1991, 46, 837.
( Fig. 6) but no effect of Sr on oscillations for other elements 28 K. E. Ohlsson and W. Frech, Spectrochim. Acta, Part B, 1991, was evident. No reasonable explanation for this can be 46, 559.
advanced at this time. 29 J. A. Holcombe, D. L. Styris and J. D. Harris, Spectrochim. Acta, Part B, 1991, 46, 629.
30 A. Kh. Gilmutdinov, Yu. A. Zakharov, V. P. Voloshin and K.
Conclusions
Dittrich, J. Anal. At. Spectrom., 1992, 7, 675.31 D. A. Katskov, A. M. Shtepan, I. L. Grinshtein and A. A. The most significant result of these investigations is the obser- Pupushev, Spectrochim. Acta, Part B, 1992, 47, 1023.
vation of temporal oscillations in the kinetics of carbothermal 32 K. E. Ohlsson, E. Iwamoto, W. Frech and A. Cedergren, Spectrochim. Acta, Part B, 1992, 47, 1341.
reduction of oxides of different elements from several Groups
33 M. M. Lamoureux, C. L. Chakrabarti, J. C. Hutton, A. Kh. of the Periodic Table. The mechanism of gaseous carbide
Gilmutdinov, Yu. A. Zakharov and D. C. Gregoire, Spectrochim. reduction, which was applied to Al
2O3reduction, can thus be Acta, Part B, 1995, 50, 1847.
extended to the oxides of many other elements (Bi, In, Mn, 34 B. V. L’vov, Spectrochim. Acta, Part B, 1996, 51, 533. Pb, Sb, Sn, Te, etc.), suggesting that the formation of gaseous 35 B. V. L’vov, Spectrochim. Acta, Part B, 1997, 52, 1.
36 Thermodynamic Properties of Individual Substances, ed. L. V. carbides is a much more typical phenomenon than generally
Gurvich, G. A. Khachkurusov and V. A. Medvedev, Nauka, accepted. The most efficient approach to verify this implication
Moscow, 1978–1982. is the investigation of the gas phase composition by means of
37 L. F. Yatsenko, Dissertation, Leningrad State University, 1985. mass spectrometry, as so clearly demonstrated more than 10 38 D. L. Styris and D. A. Redfield, Anal. Chem., 1987, 59, 2891. years ago by Styris et al. with oxides of Al,38 Be,39 Se,40 Ga 39 D. L. Styris and D. A. Redfield, Anal. Chem., 1987, 59, 2897.
40 D. L. Styris, Fresenius’ Z. Anal. Chem., 1986, 323, 710. and In,41 and Mg, Ca, Sr and Ba.42 The observation of an
41 D. L. Styris, Personal communication, 1989. over-equilibrium concentration of gaseous carbides in the
42 L. J. Prell, D. L. Styris and D. A. Redfield, J. Anal. At. Spectrom., process of carbothermal reduction of metal oxides is a result
1991, 6, 25. of fundamental importance and deserves further and more
intensive investigation.
Paper 9/01850F
It is recognised that this study has been primarily a
