Article pubs.acs.org/ac
Solid-Sampling Electrothermal Vaporization Inductively Coupled Plasma Optical Emission Spectrometry for Direct Determination of Total Oxygen in Coal Thomas Vogt,*,†,‡ Daniela Bauer,†,‡,§ David Nennstiel,‡,⊥ and Matthias Otto‡,¶ †
German Centre for Energy Resources, Reiche Zeche, Fuchsmuehlenweg 9, 09596 Freiberg, Germany Institute of Analytical Chemistry, TU Bergakademie Freiberg, Leipziger Strasse 29, 09599 Freiberg, Germany
‡
S Supporting Information *
ABSTRACT: A new analytical method for direct determination of total oxygen contents in eight coal samples of the Argonne Premium Coal (APC) series and in the NIST SRM 1632d is presented. The development of a suitable calibration procedure, optimization of measurement conditions, and the application of a tailored data processing for handling of plasma effects and high blanks enable the quantification of oxygen simultaneously with other trace, minor, or major elements in whole coal samples by means of electrothermal vaporization inductively coupled plasma optical emission spectrometry (ETV-ICP OES). For comparison, the oxygen contents were determined by a direct oxygen analyzer. The obtained oxygen values of the APC and the reference material NIST SRM 1632d were compared to data in the literature. The precision of the ETV-ICP OES was within ±3.5%, and the recovery better than 92%. With this good accuracy, the developed direct solid sampling method ETV-ICP OES is well suited for the fast determination of oxygen in coals, varying in rank from lignite to semianthracite, in a content range of about 100 ppm up to 27% using 1.5 mg sample weight. This direct analysis method represents an accurate, advantageous alternative to currently used methods for estimation of total oxygen contents in coals.
D
due to the requirement of a suitable nuclear reactor and specialized equipment for radiochemical analysis. For optimization of the process efficiencies, maximization of the exploitation of resources, and ensuring the desired specifications of products, an essential requirement is the development of a procedure for rapid-process analysis. Ideally, the method for process analysis should enable the direct determination of oxygen simultaneously besides other relevant element contents, and the feasibility of element speciation should be provided. As previous studies5,6 of our research group demonstrated, the electrothermal vaporization inductively coupled plasma optical emission spectrometry (ETV-ICP OES) meets these requirements and is readily applicable for the fast quality control of coals. In contrast to classical ICP OES, it is theoretically possible to determine oxygen with ETV-ICP OES. Coupled to ETV for sample introduction, an oxygen-free and argon-purged closed system is created, which allows the measurement of the released oxygen. For analysis, the pulverized sample is placed in a graphite boat and is resistively heated up to a temperature of 2400 °C in the graphite furnace
ue to the high influence of the total oxygen content on the chemical behavior of coal during processing, a direct measuring method is particularly desirable for determining the oxygen content of coal. Currently, two methods are routinely used for direct analysis of the total quantity of oxygen in organic matter: the Schütze−Unterzaucher method1,2 or its modifications3 and fast neutron activation analysis (FNAA).4 By means of the oxygen analyzer, the modern version of the Schütze−Unterzaucher analysis method, the total oxygen content of solid samples is analyzed via melt extraction of the sample in graphite vessels at high temperatures of about 2400 °C, followed by infrared spectrometric detection. For quantification, the released oxygen-containing decomposition products are reduced by means of hot activated carbon. All the oxygen is thereby converted to carbon monoxide, which is transported by an oxygen-free carrier gas flow through a suitable nondispersive infrared detector. In opposite to the other direct analysis technique using fast neutron activation, the pyrolysis technique “oxygen analyzer (OA)” is not a multielement analysis method. Common to both techniques is the impossibility to provide information about the chemical species of an element in coal and on their temperature-dependent release. The great disadvantage of the FNAA method is the relatively long time of analysis and the expensive equipment, © 2015 American Chemical Society
Received: July 6, 2015 Accepted: September 15, 2015 Published: September 15, 2015 10414
DOI: 10.1021/acs.analchem.5b02530 Anal. Chem. 2015, 87, 10414−10420
Article
Analytical Chemistry Table 1. Sample Identification (ID), Seam, Rank, and Chemical Properties of Coal Samples Used11,12,14a,b ID
seam
rank
C
H
S
N
ash
OFNAA
POC UF PITT 1632d UT IL WV WY ND
Pocahontas Upper Freeport Pittsburgh Pittsburgh Blind Canyon Illinois Lewison Stockton Wyodak Anderson Beulap-Zap
low vol. bit. med. vol.bit. high vol. bit. bit. high vol. bit. high vol. bit. high vol. bit. sub bit. lignite
86.7 74.2 75.5 77.9 76.9 65.7 66.2 68.4 65.9
4.23 4.08 4.83 5.06 5.49 4.23 4.21 4.88 4.36
0.66 2.23 2.19 1.48 0.62 4.84 0.71 0.63 0.8
1.27 1.35 1.49 1.55 1.5 1.16 1.25 1.02 1.04
4.8 13.2 9.3 7.1 4.7 15.5 19.8 8.8 9.7
4.22 8.93 10.9 12 12.5 14.9 16.5 20.7 23.3
a
All contents are given in % on dry basis. bAbbreviations: ID, identifier; med., medium; vol., volatile; bit, bituminous; FNAA, fast neutron activation analysis.
with opened furnace door, no air enters the system, due to the argon stream, which flows in that case out of the furnace. By using the ETV for solid sample introduction, the samples were placed in graphite boats, which were automatically introduced in the furnace of the ETV system by an autosampler (AD50; Spectral Systems). To ensure the accurate regulation of the temperature, the ETV is equipped with an internal furnace pyrometer, which is regularly calibrated with an external pyrometer (PL31 AF4, Keller MSR) before each measuring run. Considering the durability of the graphite tube and boats, the solid samples were heated up to a maximal temperature of 2400 °C using an optimized temperature profile. The volatilization of carbide-forming elements was supported by adding a halogenated reaction gas (CCl2F2) to the inner gas flow of the furnace. This is basically necessary for the analysis of temperature stable oxides, for the volatilization of other elements in coals, and to prevent deposits in the graphite heating tube. The vaporized analytes and volatile halides were transported as a dry aerosol by an argon/reaction gas flow to the inductively coupled plasma of the atom emission spectrometer, where the analytes were atomized and respectively ionized. The excited outer electrons of the atoms and ions return to a lower energy level by emission of emitted electromagnetic radiation with specific wavelengths for each element.8 These emission lines can be detected simultaneously in the Rowland circle in the Paschen−Runge mount (130−770 nm) of the spectrometer. In the case of the used axial viewing ICP OES instrument, the plasma was operated in a horizontal orientation, and the analytical zone was observed from the end of the plasma (EOP). The argon-filled UV-PLUS Optic in the ICP enables the use of the typical spectral lines down to 130 nm for the detection of nonmetals like Cl, I, B, P, S, N, and O itself. For analyzing the total oxygen contents of the samples by means of ETV-ICP OES, a heating program with a relatively slow heating rate was used to prevent a detector overflow due to a high release of analytes. For best results with respect to the limit of detection (LOD), a fast heating rate would be advantageous, but this is not necessary for coal samples due to the partly high element concentrations. The contents of elements in the investigated samples vary over a wide concentration range (cf. Table 1). Due to the fact that LOD and working range are specific to the used spectral lines, the selection of appropriate wavelengths among the enormous variety of emission lines is very important. By the use of the solid-sampling introduction system (ETV), the input of sample material to the plasma of the ICP is very
of the ETV unit in argon atmosphere. During the pyrolysis process, the analytes are released and transported by an argon stream directly into the inductively coupled plasma of the atomic emission spectrometer.7 The atomic emission spectrometric analysis is based on the emission of light by atoms and ions returning to a lower energy level after excitation in the ICP of the spectrometer.8 This solid-sampling multielement method enables the direct determination of trace, minor, as well as major elements simultaneously in whole coal samples with a minimum effort of sample preparation and short analyzing times.6 The direct multielement spectrometric method is sensitive, can be automated to a large extent of samples, and is so applicable for process-accompanying analyses. Additionally, due to the working principle of this technique, different element compounds can be selectively evaporated, and so the possibility of element speciation by vaporization temperature is given by the controlled electrothermal vaporization of the powdered coal sample under argon atmosphere.5 This study is focused on the development of the ETV-ICP OES method to measure total oxygen contents in coals with the usage of Argonne Premium Coal (APC) samples, which cover the rank range from lignite to semianthracite and the Standard Reference Material (SRM) 1632d purchased from the National Institute of Standards and Technology (NIST). For the determination of oxygen in coals, the development of a suitable calibration procedure and data processing were key issues of investigation, because of the lack of availability of certified coal reference materials, plasma effects, and high blanks. Additionally, the heating rate of the ETV was optimized with regard to a high dynamic range. To verify the developed method, it is necessary to compare the obtained data with an established method. Therefore, in addition to the development of the ETVICP OES method, the analysis technique for oxygen determination via melt extraction had to be adapted for analyzing coal samples instead of metal or alloys.
■
EXPERIMENTAL SECTION Instrumentation. The measurements for determination of oxygen in the used samples were performed by means of an axial viewing ICP OES spectrometer (ARCOS EOP; SPECTRO Analytical Instruments) which was interfaced by PFA-transport tubing to a solid sampling ETV unit (ETV 4000c; Spectral Systems). In previous publications the ETVICP OES technique has been described in detail, so here the report about this direct analysis method is deliberately limited to a brief description. A basic requirement for oxygen analysis is the need of a completely leak-proof system, which is given by means of the ETV unite, as the solid sampling interface. Even 10415
DOI: 10.1021/acs.analchem.5b02530 Anal. Chem. 2015, 87, 10414−10420
Article
Analytical Chemistry
decomposition of calcium oxalate monohydrate be guaranteed, CaO (intermediate decomposition product) is thermally stable under normal conditions up to 2613 °C.10 By means of the calibrated oxygen analyzer and ETV-ICP OES method, the total set of coals from the Argonne Premium Coal sample program and NIST SRM 1632d has been examined. The APC, ranging in rank from lignite to semianthracite, were collected, prepared, and stored under controlled conditions and have been extensively studied worldwide. The selection of the eight APC was based on a cluster analysis of whole coal seam data. Criteria for selection were primarily the compositional characteristics carbon, hydrogen, oxygen, or sulfur, to cover a large degree of coalification or large range of sulfur contents. Particular attention has also been paid to coking properties and different maceral types.11 In summary, the element contents of interest are available for the APC11,12 and the SRM 1632c13 whereby the total oxygen contents were measured by FNAA. In Table 1 a compilation of the designations, geographic origins, rank, and analytical characteristics from the used samples is given, in order of increasing oxygen content. The element contents are given on the dry basis in percentage by mass. Sample Preparation. The reference materials or samples were analyzed as received with regard to the particle size (
