Some problems on adjustment of the equipment and on data evaluation, in view of the equations derived earlier for the calculation of specific heat and heat of transformation are discussed. It is shown that quantitative determinations are possible also if the adjustment is "non-ideal" and that both air and an inert material, respectively, may be used as reference material.Comprehensive information and detailed recommendations are to be found in the recent literature on how to select optimum experimental conditions and equipment for thermal analysis [1--3]. In this paper we wish merely to specify some points with respect to experimental technique and equipment design, required by the application of the equations derived earlier [4] for the calculation of specific heat and heat of phase transformation.

In thermal analysis of disperse materials, the gas filling the pores of the samples has an important effect on the determination of heats of phase transformations.Particularly significant errors, up to 25 ~, may arise in cases when thermal analysis is carried out in gas atmospheres having high thermal conductivities.Based on experimental data, a relationship expressing the dependence of peak area on the thermal conductivity of the gas and on the thermal conductivity and particle size of the studied material has been derived. This relationship allows to calculate the possible experimental error and hence to adopt measures for reducing its value.Owing to the inadequate elaboration of the quantitative thermal analysis of processes accompanied by the formation of a gaseous phase, in such cases researchers are often forced to apply relationships established for solid-phase processes.In our view, the particular features of heat exchange and mass exchange in a disperse material as a result of the appearance of gaseous products, the form in which they are reflected, the quantitative aspects of the effects, and hence wellfounded suggestions as to the necessity and means of eliminating or taking them into account, have not been discussed with satisfactory consistence in the thermal analysis literature [1][2][3][4][5][6][7].The fact that the presence of a gaseous phase in the pores of a disperse material is reflected in the magnitude of the peak areas corresponding to the phase transformations may be considered as reliably established. Gases with higher thermal conductivities increase, and gases with lower thermal conductivities decrease the peak areas [I-4].The quantitative aspect of this problem, however, i.e. the degree to which peak areas change as a function of the parameters of the disperse system and the gaseous medium, has been much less studied. Reports in the literature on the quantitative evaluation of the effect of the gaseous phase in the thermal analysis of disperse materials are insufficient for the establishment of general relationships. In many cases the data are contradictory, obviously because the experimental apparatus used by the different authors differed in design.The main objective of the present work was the quantitative evaluation of the effect of the gaseous phase on the quantitative characteristics of the thermal curves, and primarily on the phase-transition peak areas.

The effect of some experimental factors on quantitative determinations of heat of transformation are discussed. It is shown on the example of solid phase invariant processes that by using the equation derived earlier by the authors, a number of factors, viz. mass, bulk density, thermal conductivity and specific heat of the sample, as well as the position of the junction of the thermocouple, need not be taken into account.

Differential thermal analysis and simultaneous electrical conductivity and evolved gas detection measurements were used to determine temperature intervals and in a number of cases the nature of the recorded transformations of 21 sulphides. For example, along with the thermal dissociation of covellite interaction with copper sulphide impurity is also found. The result of polymorphic pentlandite transformation (610-620 ~ is vaesite; in the incongruent melting of tennantite (710--735~ lautite and sulphur are formed. Dissociation of sulpharsenides is as follows: cobaltite (885--905 ~ yields cattierite and sulphur; gersdorffite (800--860 ~ yields niccolite and pentlandite; and arsenopyrite (670-740 ~ yields loellingite. Endothermic transformation of pyrite (550--580 ~ results in destruction of its superficial oxidized film. A thermal change typical for each type of iron monosulphide has been established. A method for the quantitative estimation of sulphides is based upon measurement of the gas evolved during interaction of the sulphide with solid oxidants.Two lines of approach have been followed in studies on the thermal behaviour of sulphides. Some researchers investigated the processes taking place when the specimens were heated in air [1-6], while others worked in inert atmospheres [7][8][9][10]. In the first case the processes studied involve oxidation of the sulphides in air, and the subsequent dissociation of the products formed. On the other hand, studies of the thermal behaviour of sulphides under conditions excluding their oxidation allow investigation of transformations that are free from the superposition of additional processes.Several authors, e.g. [4,7,8], point out the feasibility of utilizing DTA results for the identification of sulphide minerals. However, data in the literature on the temperature ranges of sulphide oxidation in air are largely inconsistent, and demonstrate the poor reproducibility of the DTA results. This latter finding might be explained by the large number of not readily stabilizable factors that have an effect on the thermal analytical characteristics when sulphides are heated in oxidizing atmospheres. Hence, it appears difficult to solve the problem of thermoanalytical identification in this way. It is obvious that satisfactory reproducibility of thermal analytical data is a basic condition for their successful utilization in the identification of minerals. From this aspect, information on the thermal behaviour of sulphides in inert media appears more useful (in addition, such information is also very interesting from a theoretical viewpoint).

A highly simplified method for calculating heat of phase transitions from DTA data is presented. Two DTA curves are needed to calculate the heat of transition and the specific heat of the sample: one is for the original sample and one is for a sample prepared by mixing the original sample with some unreactive diluent the specific heat of which is known. The data of the DTA curves used in the calculations are the peak area, the rate of heating and the deviation of the DTA curve from the base line.The quantitative determination of thermal effects by differential thermal ana ysis (DTA) using normal apparatus is based on the well-known relation between the peak areas (S) and the heat (Q) absorbed or liberated by the substance investigated [1 -5]. This relation is usually expressed mathematically by the equationOwing to the extremely high number of factors determining the value of the coefficient K, it is practically impossible -at least at present -to give a full mathematical interpretation of K [6,7]. Attempts to take these factors into consideration empirically have led to much valuable experimental material in respect to the dependence of the peak area on various experimental conditions and on the characteristics of the substance investigated, i.e. heat conductivity [8], heat capacity [9, 10], shape [11] and degree of dispersion [12] of the sample, rate of heating [13], rate of heat exchange between the sample and the medium [14], position of the junction of the thermocouple in the sample [4, 9], etc. It seems reasonable to assume that all mentioned and unmentioned factors are reflected in the geometry of the thermal curves. The aim of the present paper is to find a method for their simplified mathematical analysis.Before passing to the mathematical treatment of thermal curves, it is necessary to discuss the effect of the thermophysical properties of the sample on the geometrical elements of the thermal curves.A number of limitations assumed by various authors attempting the mathematic interpretation of thermal curves [7,15] reflects that they accept the dependence of the geometry of thermal curves on the thermophysical properties of the sample.