Precision combustion analysis

Principles and Characteristics Combustion analysis is used primarily to determine C, H, N, O, S, P, and halogens in a variety of organic and inorganic materials (gas, liquid or solid) at trace to per cent level, e.g. for the determination of organic-bound halogens in epoxy moulding resins, halogenated hydrocarbons, brominated resins, phosphorous in flame-retardant materials, etc. Sample quantities are dependent upon the concentration level of the analyte. A precise assay can usually be obtained with a few mg of material. Combustions are performed under controlled conditions, usually in the presence of catalysts. Oxidative combustions are most common. The element of interest is converted into a reaction product, which is then determined by techniques such as GC, IC, ion-selective electrode, titrime-try, or colorimetric measurement. Various combustion techniques are commonly used. 

Basic Mechanisms. Finally, further work is necessary on fundamental mechanisms of individual fire retardants. These mechanisms are a function of the particular chemicals involved and the environmental conditions of the fire exposure. There is a need to establish common methods and conditions for determining these mechanisms in order to compare different treatments. This would give us a better understanding of how these compounds work in action and would provide a more efficient approach for formulating fire-retardant systems than a trial and error approach. Correlations also need to be established between rapid precise thermal analysis methods and standard combustion tests. Retardant formulations could be evaluated initially on smaller (research and development size) samples. The more promising treatments could be tested for flame-spread index, heat release rate, and toxic smoke production. 

Table 7-5 Accuracy and precision of combustion analysis of pure compounds" ... 

Since the carbon to hydrogen ratio of 1 2.67 is too far from a whole number to be attributed to experimental error, the empirical formula must be the lowest multiple to give a whole number ratio, and 2 x (1 2.67) gives 2 5.34, still not a whole number, but 3 x (1 2.67) gives 3 8.01, which is a whole number ratio within the experimental error. It is difficult to be precise about the experimental error that is acceptable in determinations of this kind, but the percentage of C, H, or N determined by combustion analysis is usually quoted to 0.1%. 

The precise knowledge of the chemicals, produced and stored in the factories, is a necessary condition for modeling of the accidents. However, it should be taken into account that in real conditions and especially when the accidents are accompanied with fire, new toxic compounds may be formed. Such is the case with the fire, which occurred on July 13, 1993 in the "Alen Mak" factory in the town of Plovdiv, Bulgaria. During the GC-MS analysis of the air, soil and various parts of the incident site, over 120 different chemical compounds were identified, some of which identical to the ones present on the premises before the fire, while others had obviously formed during the process of combustion. For a great number of these compounds, toxicological data was not found in the accessible information banks, such as IRIS of EPA and others [5],... 

The Dohrmann DX 20B system is based on combustion of the sample to produce the hydrogen halide, which is then swept into a microcoulometric cell and estimated. It is applicable at total halide concentrations up to lOOOpl-10 with a precision of 2% at the lOpg L-1 level. The detection limit is about 0.5pg L-1. Analysis can be performed in 5 min. A sample boat is available for carrying out analysis of solid samples. The instrument has been applied to waste waters, soils and sediments. 

The combustion infrared technique has been used for the analysis of diluted sludges [34], Schaffer et al. [34] have used a blender to prepare suspensions of samples of this type. After the sample was homogenized, a microlitre syringe was used to remove a 20pL aliquot from the blender. However, the method described by Schaffer et al. [34] does not necessarily allow for the isolation of a representative portion of the sample in a 20pL aliquot [35]. This was found to be particularly true if the sample contained large particles which settled rapidly, and it must be assumed that the precision... 

Compound-specific isotope analysis (CSIA) by GC-IRMS became possible in 1978 due to work of Mathews and Hayes [634], based on earlier low-precision work of Sano et al. [635]. The key innovation was the development of a catalytic combustion furnace based on Pt with CuO as oxygen source, placed between the GC exit and the mass spectrometer. The high pressure of helium (99.999% purity or better) ensures that all gas flows are viscous. After being dried in special traps avoiding formation of HC02 (i. e., interferes with 13C02) by ion-molecule reactions in the ion source, the C02 is transmitted to a device that regulates pressure and flow and then into the ion source [604]. 

Tobias et al. [665] have described a method in which the GC effluent is passed into a combustion furnace to convert the organic hydrogen content into water, which is then selectively reduced to hydrogen in a reduction furnace containing Ni metal. The final stream is transmitted to the IRMS via a heated Pd filter, which passes only hydrogen isotopes to the ion source. For a benzene sample a precision of < 5 %o was obtained for <52HSMOw> which approaches the performance of off-line techniques and the requirements for studies of natural variability. This already meets requirements for analysis of D-labeled compounds used in tracer studies [666,667]. 

A specific variant of El MS is isotope ratio (IR) MS [46]. It is based on electron impact ionization with maximized ionization probability. IR MS is limited to the analysis of gases of high volatility and low reactivity such as CO2, N2 or SO2. The analytes of interest thus have to be transformed into one of these gases before introduction into the IR MS. Information on the position of C labelings in the analyte can be only obtained, if all carbons are isolated position specific and subsequently combusted. In this context Corso and Brenna [47] showed position specific analysis by IR MS for methylpalmitate through pyrolytic fragmentation. IR MS exhibits an extremely high precision of 0.00001 % for the isotope ratio measurement and is optimal to quantify low label enrichments [48]. This is especially important for in vivo studies with ani-... 

Industrial analysis of hydrocarbon gases 25 years ago was limited almost to Orsat-type absorptions and combustion, resulting in crude approximations and inadequate qualitative information. The more precise method of Shepherd (56) was available but too tedious for frequent use. A great aid to the commercial development of hydrocarbon gas processes of separation and synthesis was the development and commercialization of high efficiency analytical gas distillation units by Podbielniak (50). In these the gaseous sample is liquefied by refrigeration, distilled through an efficient vertical packed column, the distillation fractions collected as gas and determined manometrically at constant volume. The operation was performed initially in manually operated units, more recently in substantially automatic assemblies. 

There are two main approaches to the oxidation of OC in water samples to C02 combustion in an oxidizing gas and UV-promoted or heat-catalyzed chemical oxidation. Other approaches are sometimes used, but are much less widespread.11 Carbon dioxide, which is released from the oxidized sample, can be detected in several ways, including conductivity detection, nondispersive infrared (NDIR) detection, or conversion to methane and measurement with a flame ionization detector (FID).1213 The limits of detection in TOC determination can be as low as 1 pg L 1, and the dynamic range can span many orders of magnitude. The precision of the method is usually very good, and the analysis can be completed in a few minutes. Another advantage is the very small amount of sample required—from 10 to 2000 pL. 

---